Post-implant site monitoring and medical device performance

Abstract

Techniques for remotely monitoring a patient and a corresponding medical device are described. The remote monitoring includes identifying a first set of images representing a particular location of a patient's body in which at least one component of an implantable medical device (IMD) coincides, determining a projection of a change characteristic of the particular location of the body, identifying a second set of images, determining a second set of change characteristics, comparing the second set of change characteristics to the projection, and identifying a potential abnormality at the particular location of the body.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to medical devices, and in some particular examples, to computing devices (e.g., mobile devices) configured to assess a patient’s recovery from implantation of a medical device and / or to assess performance of a medical device currently implanted in a patient. BACKGROUND

[0002] Medical devices can be used to treat a variety of medical conditions. Example medical devices include implantable medical devices (IMDs), such as cardiac or cardiovascular implantable electronic devices (CIEDs). An IMD, sometimes also referred to as an “implanted medical device,” can include a device implanted in a patient at a surgically or procedurally prepared implant site. An IMD can include a diagnostic device configured to diagnose a patient for various ailments, monitor a patient’s health status, and the like. Additionally or alternatively, an IMD can also be configured to deliver electrical stimulation therapy to a patient via electrodes, such as implanted electrodes, where the device can be configured to stimulate the heart, a nerve, a muscle, brain tissue, and the like. In any case, an IMD can include a battery-powered component in some cases, such as in the case of a reference to an implantable cardiac pacemaker, an implantable cardioverter-defibrillator (ICD), other electrical stimulators including spinal cord stimulators, deep brain stimulators, nerve stimulators, and muscle stimulators, infusion devices, cardiac and other physiological monitors, cochlear implants, and the like. In such cases, the battery-powered component of the IMD can be implanted, such as at a surgically or procedurally prepared implant site. Additionally, related devices, such as elongated medical electrical leads or drug delivery catheters, can extend from the IMD to other subcutaneously implanted sites, or in some cases, deeper into the body, such as to an organ or various other implant sites.

[0003] While preparation and implantation are performed in a sterile field, and IMD components are packaged in sterile containers or sterilized prior to introduction into the sterile field, there can still be a risk of introducing microorganisms into the site. Thus, implanting clinicians often apply a disinfectant or antiseptic to the skin at the surgical site prior to surgery, directly to the site prior to closure of the incision, and prescribe oral antibiotics for the patient to ingest during recovery. Despite these precautions, infections can still occur.

[0004] Accordingly, infections related to implanted medical devices remain a concern with respect to public health and economics. If an IMD-related infection occurs, device explantation is often the only appropriate course of action. Additionally, once a site is infected, the infection can migrate, for example, along a lead or catheter to the location where the lead and / or catheter is implanted. For example, removal of a long-term implanted lead or catheter in response to such an infection can be very difficult. Aggressive systemic medication is used to treat such infections. However, early detection of IMD-related infections can allow for early intervention, thereby reducing device explantation.

[0005] In some cases, patients who have implanted certain medical devices, such as CIEDs, require post-implant follow-up consultation visits by a healthcare professional (HCP). The HCP visits are typically conducted at any time from a few days to a few weeks after implantation. The HCP typically performs a wound check and can initiate a device interrogation session to determine performance metrics of the CIED. In most cases, these visits are not significant and are very brief, up to a level of 90% or more. However, the post-implant follow-up can still constitute a necessary checkpoint for the patient with the HCP. In some cases, such follow-up can find potential infections or device-related complications that can require clinical intervention. SUMMARY

[0006] While actual follow-up consultation visits by a healthcare professional (HCP) tend to be relatively short, the visits can still constitute a burden on the patient’s life, the doctor’s office, and the healthcare system as a whole. Aspects of the present disclosure relate to one or more computing devices having processing circuitry configured to facilitate or simulate a virtual check-in of a patient, such as a virtual follow-up or other health check, in which the patient and / or the doctor can check the status of a medical device, including an implantable medical device (IMD), and / or the patient themselves.

[0007] The processing circuitry of the one or more computing devices can implement various tools and techniques (hereinafter referred to as “tools,” processor systems, or processing circuitry) in order to provide virtual enrollment for a patient. The processing circuitry can be configured to provide the patient with an option to participate in an interactive session (e.g., a virtual enrollment process, an interactive enrollment session, an interactive reporting process, etc.) as part of a post-implant evaluation, for which the patient can conduct an interactive reporting session remotely, such as remotely from an official physician’s office environment or other HCP environment. In some examples, the processing circuitry of a mobile or portable computing device can be configured to manage the interactive session such that a user can efficiently navigate the interactive session from any remote location. Additionally, the processing circuitry can communicate patient and / or medical device data between various computing devices (e.g., end-user mobile devices, camera devices, wearable devices, edge devices, etc.). In some examples, the processing circuitry can be configured to manage the interactive session to ensure proper interaction with the enrollment tools such that the system can accurately determine whether the patient and / or IMD is experiencing any complications. The processing circuitry can communicate such information over a network and / or by implementing various communication protocols.

[0008] In some cases, the processing circuitry can implement the interactive session in a secure environment, such as on an authenticated mobile computing device and / or over a secure data network. In some cases, the processing circuitry can implement the interactive session on a mobile device that is configured to perform one or more of the various techniques of the present disclosure with or without the assistance of other devices (e.g., network devices). That is, the mobile device can operate various software tools configured to perform a virtual enrollment process. The patient can then conduct the entire session of the virtual interactive enrollment session (e.g., multiple sub-sessions of the interactive session) without network access or with limited network usage, such as when using wireless communication with an edge device (e.g., an IoT device having additional processing resources to assist or perform various enrollment functions). The mobile device can synchronize the virtual enrollment data with other network devices at a later point in time after the enrollment session (e.g., multiple sub-sessions of the interactive session) is complete. In this way, an HCP can access the data at the time, where it can not be important for the HCP to access such data in real-time or near real-time as the patient conducts a particular enrollment session (e.g., multiple sub-sessions of the interactive session).

[0009] According to the techniques of this disclosure, processing circuitry is configured to provide a comprehensive user interface (UI) to a user (e.g., a patient), where the UI is configured to guide the user through an interactive session and / or an assessment process. The processing circuitry of a computing device can provide the UI to the user by generating UI program data and / or accessing UI program data from memory. In some cases, the processing circuitry can access the UI program data from a separate computing device of a virtual enrollment computing system. In some cases, the processing circuitry of a computing device can tailor the UI program data for each particular user or user category. In some cases, the processing circuitry can tailor the UI program data by deploying various artificial intelligence (AI) algorithms and / or machine learning (ML) models configured to determine a UI reflective of a particular user or user category and a medical device corresponding to the particular user or user category. In any case, the processing circuitry can be configured to provide the UI to a user of a computing device operated by the user via a display device.

[0010] In some examples, the UI can include interactive UI enrollment elements (e.g., UI tiles) configured to systematically guide a patient through a virtual enrollment process. The UI enrollment elements can be configured to proactively assist a particular patient in providing specific patient inputs to the system. The patient inputs can include a specific type of information and a specific amount of information. In some examples, the processing circuitry can use the information to, for example, identify various conditions of the patient, such as a device pocket infection, or otherwise identify an abnormality of a medical device and / or the patient. In some examples, the processing circuitry can deploy AI and / or ML models in order to evaluate the inputs received through the UI to determine a health status of the patient and / or a status of one or more medical devices, including an IMD. In another example, the processing circuitry can communicate the inputs from the patient to an HCP. The processing circuitry can in turn receive inputs from the HCP. In this case, the processing circuitry can determine a health status of the patient and / or a status of one or more medical devices based on the HCP inputs. In any case, the processing circuitry can utilize the patient inputs and / or the HCP inputs to train the AI and / or ML models, where the processing circuitry can obtain the HCP inputs, which in some cases are based on the patient inputs (e.g., uploaded images of an implant site, ECG waveforms, etc.). Thus, the processing circuitry can utilize information from multiple sources to determine various conditions of the patient. While described with reference to a comprehensive UI, the techniques of this disclosure are not so limited, and it should be appreciated that the UI elements can be implemented as independent UIs that involve a subset of the UI elements of this disclosure. That is, the processing circuitry of a computing device can not include all of the UI elements in a single UI program, and in some cases can include additional UI elements configured to provide various other enrollment functions.

[0011] In some examples, the UI registration elements can include general patient status registration elements, physiological parameter registration elements, medical device registration elements, site check elements, such as wound check elements, etc. The processing circuitry can obtain information related to the various discrete registration elements and determine a health status of the patient, such as a status of the patient’s IMD. In some cases, the patient can perform such virtual checks periodically to check the status of the medical devices or implant sites. That is, the patient can perform health checks even after performing a number of virtual registration sessions (e.g., multiple sub-sessions of an interactive registration session) that are mandatory or recommended by the HCP after a surgical event for the patient.

[0012] The interactive reporting session can replace and / or in some cases supplement an HCP in-person visit. In some examples, the processing circuitry of the present disclosure can use the results of the implemented checks by the computing device to determine whether to provide an indication to the patient, and in certain cases to the physician or other HCP, as to whether an in-person visit is needed, or alternatively whether the IMD wound recovery for the patient is proceeding as expected. Additionally, the processing circuitry can implement one or more sub-sessions of the interactive session to determine whether the one or more medical devices are functioning within an acceptable range, etc. In some cases, the processing circuitry can obtain information from a user via one computing device, such as a mobile phone device, and evaluate the information via another computing device, such as an edge device or a network server. In such cases, the edge device can deploy AI and / or ML models trained according to medical device data, patient data, and / or heuristic data obtained from the network. In some cases, the processing circuitry can train the AI or ML models with data obtained from the user’s computing device and / or data obtained directly from the patient’s medical devices. In such cases, those medical devices can be configured to communicate with the edge device as well as the user computing device.

[0013] In one example, the present disclosure provides a method of imaging a body of a patient having an IMD, the method comprising: identifying, via processing circuitry of a computing device, a first set of images representing a particular location of the body in which at least one component of the IMD coincides; determining, from the first set of images, a projection of a change characteristic of the particular location of the body over time; identifying, via the processing circuitry, a second set of images representing the particular location of the body at a progression interval in time after the first set of images; determining, from the second set of images, a second set of change characteristics; comparing, via the processing circuitry, the second set of change characteristics to the projection; and identifying, based at least in part on the comparison, a potential abnormality at the particular location of the body.

[0014] In another example, this disclosure provides a system for imaging the body of a patient with IMD, the system comprising: a memory configured to store one or more frames of image data, the one or more frames representing at least a first set of images; and one or more processors communicating with the memory, the one or more processors being configured to: identify the first set of images, the first set of images representing a specific location of the body in which at least one component of the IMD coincides; determine, based on the first set of images, a projection of a change characteristic of the projection of the specific location of the body over time; identify a second set of images, the second set of images representing the specific location of the body at a progression interval relative to the first set of images; determine a second set of change characteristics based on the second set of images; compare the second set of change characteristics with the projection; and identify potential abnormalities at the specific location of the body based at least in part on the comparison.

[0015] This disclosure also provides a non-transitory computer-readable medium including instructions that cause a programmable processor to perform any of the techniques described herein. In one example, this disclosure provides a non-transitory computer-readable storage medium having instructions stored thereon that, when executed, cause one or more processors to: identify a first set of images representing a specific location of a body in which at least one component of an implantable medical device (IMD) overlaps; determine, based on the first set of images, a projection of a change characteristic of the specific location of the body over time; identify a second set of images representing the specific location of the body; determine a second set of change characteristics based on the second set of images; compare the second set of change characteristics with the projection; and identify, at least in part, a potential abnormality at the specific location of the body based on the comparison.

[0016] This disclosure also provides means for performing any of the techniques described herein.

[0017] The summary is intended to provide an overview of the subject matter described in this disclosure. It is not intended to provide an exclusive or exhaustive explanation of the systems, apparatus, and methods described in detail in the following drawings and specification. Further details of one or more examples of this disclosure are set forth in the drawings and the following detailed description. Other features, objectives, and advantages will be apparent from the specification, drawings, and claims. Attached Figure Description

[0018] Figure 1 An environment for an exemplary monitoring system in conjunction with a patient, based on one or more technologies disclosed herein, is shown.

[0019] Figure 2 This demonstrates one or more techniques disclosed herein. Figure 1 A functional block diagram of an example configuration of an example computing device.

[0020] Figure 3 is a block diagram illustrating an example network system of example computing devices, networks, edge devices, and servers incorporating Figure 1 or 2, in accordance with one or more techniques disclosed herein.

[0021] Figure 4 is a block diagram illustrating an example configuration of a medical device incorporating Figure 1 and / or Figure 3 in accordance with one or more techniques disclosed herein.

[0022] Figure 5 is an example user interface (UI) visualization of an example computing device incorporating Figure 1 , Figure 2 or Figure 3 in accordance with one or more techniques of the present disclosure.

[0023] Figure 6 is an example UI visualization of an example launch interactive session interface in accordance with one or more techniques of the present disclosure.

[0024] Figure 7 is an example UI visualization of an example menu interface in accordance with one or more techniques of the present disclosure.

[0025] Figure 8 is a flowchart illustrating an example method of utilizing UI input and data processing in accordance with one or more techniques of the present disclosure.

[0026] Figure 9 is an example UI visualization of an example patient status interface presented in accordance with a user’s request for a patient status interface incorporating Figure 7 in accordance with one or more techniques of the present disclosure.

[0027] Figure 10 is an example UI visualization of an example patient status interface in accordance with one or more techniques of the present disclosure.

[0028] Figure 11 is an example UI visualization of an example physiological parameter check interface in accordance with one or more techniques of the present disclosure.

[0029] Figure 12 is an example UI visualization of an example physiological parameter check interface in accordance with one or more techniques of the present disclosure.

[0030] Figure 13 is an example UI visualization of an example device check interface in accordance with one or more techniques of the present disclosure.

[0031] Figure 14 is an example UI visualization of an example device check interface in accordance with one or more techniques of the present disclosure.

[0032] Figure 15 is a UI visualization of an example site check interface according to one or more techniques of the present disclosure.

[0033] Figure 16 is a flowchart illustrating an example method of capturing images of a body over time according to one or more techniques of the present disclosure.

[0034] Figure 17 is a UI visualization of an example site check interface according to one or more techniques of the present disclosure.

[0035] Figure 18 is a UI visualization of an example site check interface according to one or more techniques of the present disclosure.

[0036] Figure 19 is a flowchart illustrating an example method of capturing images of a body over time according to one or more techniques of the present disclosure.

[0037] Figure 20 is a flowchart illustrating an example method of navigating a set of UI interfaces of a virtual enrollment process according to one or more techniques of the present disclosure.

[0038] Figure 21 is a UI visualization of an example full enrollment interface according to one or more techniques of the present disclosure.

[0039] Figure 22 is a UI visualization of an example full enrollment interface according to one or more techniques of the present disclosure.

[0040] Figure 23 is a flowchart illustrating an example method of determining instructions for a medical intervention regarding an IMD patient according to one or more techniques of the present disclosure.

[0041] Throughout the specification and drawings like reference numerals refer to like elements. DETAILED DESCRIPTION

[0042] In 2019, it is estimated that over 1.5 million implantable medical devices (IMDs) implants were implanted in patients worldwide. According to Heart Rhythm Society (HRS) and European Heart Rhythm Association (EHRA) guidelines, each of these implants should receive an in-person follow-up visit within two to twelve weeks. Patient compliance with post-implant visits improves mortality and patient outcomes. However, patient compliance with such visits ranges from 55% to 80%. In 93% to 99% of these visits, the patient did not present with an abnormality (e.g., an infection). That is, many of these visits could have been performed virtually.

[0043] The present disclosure presents systems and methods for remote IMD post-implant monitoring of implant site infections. It is estimated that approximately 0.5% of IMD implants and approximately 2% of IMD replacements will develop an implant infection. Early diagnosis of IMD infections can help drive effective antibiotic therapy or device explantation to treat the infection. This is accomplished in a hospital inpatient post-operative setting where the patient can be continuously monitored.

[0044] When a patient feels any infection-related symptoms such as pain or fever, a post-discharge infection diagnosis typically follows. However, for a portion of asymptomatic infection patients who are unable to self-report any implant site abnormalities, expert review is required. Since having all patients come in for a physical infection review can be cumbersome for both the patient and the caregiver, remote monitoring of the implant site is required.

[0045] Generally, the present disclosure relates to a browser interface or mobile device application that can be operated via various user-facing computing devices including, but not limited to, a smartphone, a tablet computer, a mobile device, a virtual reality (VR), an augmented reality (AR), or a mixed reality (MR) headset, and the like. The computing device can execute a software application that causes the computing device to perform the various functions described herein either locally using the computing resources of the computing device or via cloud computing such as by transmitting captured data to a backend system (e.g., a server system) that performs some or all of the analyses described herein via a network interface. Additionally, as described herein, some or all of the analyses can be performed through edge computing such as by transmitting captured data to an edge device (e.g., an IoT device or another computing device). In some examples, the edge device can include a user-facing device or a client device such as a smartphone, a tablet computer, a PDA, and other mobile computing devices. In any case, the backend system can or can not include certain edge devices as part of the backend system. In one example, the network can interface with one or more edge devices running at the edge of the backend system in order to mediate between the user’s computing device and various network servers. In such examples, the computing device can perform the various techniques of the present disclosure by utilizing edge computing, cloud computing, or a combination thereof. An example combination of edge computing and cloud computing can include a distributed computing or a distributed computing system. In any case, the computing device can perform the techniques of the present disclosure in the application, in the tablet computer, and / or in a cloud environment.

[0046] In cloud-based embodiments of the present disclosure, the mobile device application can receive various types of post-analysis data from the backend system and present the data to the user in the received form or after performing some additional processing or formatting the data locally.

[0047] A computing device executing an application (e.g., a virtual registration procedure) can perform various functions described below, whether via local computing resources provided by the computing device, or via a cloud-based backend system, or both. In some examples, the computing device can implement the application through a web browser. In some examples, the computing device can perform a device check. In such examples, the computing device can implement one or more interrogations of one or more medical devices (e.g., IMDs, CIEDs, etc.). Additionally, the computing device can analyze medical device settings, parameters, and performance metrics.

[0048] In some examples, the computing device can perform a physiological check. In such examples, the computing device can implement monitoring and / or analysis of physiological signals detected from a patient, including electrocardiogram (ECG) signals and other health signals that can be recorded (e.g., respiration, impedance, activity, stress, etc.).

[0049] Additionally, the computing device can perform a check of one or more surgical sites (e.g., an implant wound, an implant site from which an implant was removed, etc.). In some examples, the computing device can implement image processing regarding an area indicative of an implant site (e.g., a wound site). In some examples, the computing device can perform image processing using a camera of the computing device or otherwise communicatively coupled to the computing device to detect abnormalities, such as abnormalities in wound healing and / or by determining a potential infection at the implant site.

[0050] In some examples, the computing device can perform a patient status check. The application can implement an interactive log, journal, or diary function for the patient to answer several key questions regarding patient health information, medications, symptoms, physiological or anatomical metrics, or any relevant patient-reported information.

[0051] In some examples, the tools of the present disclosure can use artificial intelligence (AI) engines and / or machine learning (ML) models. In some examples, the AI engines can use cohort data to make individual checks. The cohort can include any number of cohorts, including a CIED cohort including CIED patients, an age cohort, a skin pigmentation cohort, an IMD type cohort, etc., or combinations thereof. Additionally, the cohort data can be used to train ML models to make individual checks. In a wound check example, the tools of the present disclosure can invoke an image recognition or image processing AI that leverages a library of wounds and infections that can be obtained from various sources (e.g., trained using the library of wounds and infections).

[0052] To perform an ECG check, the tools of the present disclosure can use arrhythmia classification Al with QRS models to classify normal rhythm and any potential arrhythmias relative to patient’s demographic characteristics and traits. It will be appreciated that QRS models generally refer to a QRS complex that contains a combination of various graphical deflections contained in a typical electrocardiogram (e.g., Q wave, R wave, S wave, etc.).

[0053] To perform the device check functionality of the present disclosure, the application can utilize various communication hardware components of the computing device to interrogate one or more medical devices (e.g., IMDs) to retrieve medical device parameters specific to the medical device family. The processing circuitry can be configured to compare the retrieved parameters to settings previously provided by the physician and comparable medical device families to perform a normal / orthodox deviation analysis.

[0054] The tools of the present disclosure can also provide a patient status check functionality using interactive sessions with the patient in which the patient provides elicited input to answer questions related to the patient’s health and current condition. In various examples, the tools of the present disclosure can enable the patient to input information (e.g., condition information or results) through text input, drop-down menus and / or radio buttons selecting from pre-populated responses, as shown in one or more of the accompanying drawings. In some non-limiting examples, the tools of the present disclosure can output questions to elicit patient responses by which the patient can input information such as medications, their dosages, and other prompting information.

[0055] At the completion of the session (e.g., interactive reporting session), the tools of the present disclosure can mark the results on the UI provided through the mobile device application with a date and time stamp. The tools of the present disclosure can enable the patient to have the ability to generate a report for the patient’s records (e.g., as a portable document format (PDF) or various other formats), transmit the report or select content of the report to a family member or physician (e.g., through a file transfer protocol (FTP)), etc. by email or other means.

[0056] In some non-limiting examples, the application of the present disclosure can enable a patient to have the ability to save results locally to a computing device (e.g., a smartphone or tablet) for future comparison, reference, or use as a standalone file. In some cases, the computing device can store the results locally to the computing device for use as a retrievable session within the application or other applications, or as part of a health kit implemented on the computing device. In some non-limiting examples, the application of the present disclosure can push reports to a proprietary network (e.g., via an online portal). By pushing reports or other data in this manner, the tools of the present disclosure enable HCPs to review a patient’s health information and enter the data into an electronic medical record (EMR) database or repository. The tools of the present disclosure can generate an acknowledgement of receipt to the HCP based on various criteria (and, in some examples, generate an indication of whether the report was reviewed by the HCP), and provide such communication to the patient through communication with the mobile device or other patient-accessible computing means. In some examples, the computing device can be configured to receive an indication of whether the report was reviewed by the HCP. In such examples, the computing device can be configured to provide a status of the report (e.g., reviewed by HCP, review in progress, etc.) to the user based at least in part on the indication.

[0057] If the tools of the present disclosure determine that any of the above-described checks or any combination of checks yields an abnormal result (or an abnormality outside of an acceptable normal range), the tools of the present disclosure can output a prompt to the patient using the mobile device application. In some examples, the prompt can indicate the abnormality. In some examples, the prompt can include a suggestion or indication to schedule a follow-up with an HCP.

[0058] In some examples, the tools of the present disclosure can store session information (e.g., sub-session information, combined session information, etc.) and results of previous checks (e.g., stored locally on the computing device, stored to a cloud storage resource, or stored to both). The computing device can do so in order to assist the patient and / or HCP in tracking progress or changes with respect to the patient’s wound recovery, functionality of a medical device, etc. In some examples, the tools of the present disclosure can implement a limited date counter with respect to the application in order to deter or prevent the patient from continuing to use the application after a follow-up time window has expired. Aspects of the present disclosure enable, for example, two-way communication for contacting a doctor, who can communicate with various messages such as “call my office,” etc.

[0059] As will be appreciated by those skilled in the art, remote medical device monitoring as disclosed herein represents a significant technical advance over existing implementations. In particular, the disclosed technology can identify and display a customized set of patient-related content elements and otherwise enhanced content, thereby improving the efficiency of quick access to relevant content and allowing a user to interact with multiple patient-related content elements (e.g., content tiles, sub-session interfaces, time-lapse representations, etc.). Further, the disclosed technology can yield health benefits by dynamically determining the presence of an anomaly using the particular tool deployed and implemented at a particular moment in time to achieve the highest precision of anomaly detection. Related control mechanisms can also be presented to the user that allow the user to intuitively manipulate the UI in order to capture the correct images (e.g., frequent images within a first time range and less frequent images within a second time range in order to determine a projection of healing or other changes at a particular body part of patient 4), the correct physiological parameters, the correct device interrogation data, and the correct patient status updates that will be used to achieve the highest accuracy of anomaly detection. As such, the examples described herein represent a significant improvement in this computer-related technology.

[0060] Further examples disclosed herein also represent improvements in computer-related technology. For example, a camera system can use enhanced overlays (e.g., wireframes) to allow a user to accurately align an implant site or other portion of a patient’s body so that images can be obtained in a consistent and reliable manner. Additionally, using such enhanced overlays allows a computing device to receive consistent images at a particular angle and / or with a particular implant site size in order to determine a patient’s healing progress over time based on particular analysis of the images over time. Another technical improvement includes using stored images of a patient’s body (e.g., images taken shortly after surgery) in order to develop wireframes that simulate an implant site (e.g., transparent augmented reality overlays) so that a patient can accurately align an implant site with the wireframes as a guide to capturing images at a particular angle and / or size or particular contrast, lighting, zoom, etc. Advantageously, when, for example, a healing scar at an implant site fits within a particular region of a wireframe, a computing device can initiate automatic still image capture of an image of the implant site. The computing device can utilize still image capture in order to determine a projection over time from a plurality of advantageously aligned images rather than from images that have not been controlled in any meaningful way with respect to their capture or alignment. Additionally, in some examples, the monitoring system disclosed herein can be a native system to other software applications and, as such, can use similar UI / UX elements to display tiles. Other example improvements include the ability to call external information sources to provide greater relevance and sophistication to post-operative reports and, in some examples, the ability to call APIs (e.g., language translation APIs, machine learning APIs, ECG waveform analysis APIs) to perform additional work within the monitoring system. In one example, a cloud-deployed API can be accessible as an endpoint for an ML model. Additionally, various ML models or AI engines can be deployed as so-called light versions that are configured to run efficiently on devices with very limited resources (e.g., mobile devices, tablet computers, etc.).

[0061] The ideas disclosed herein reside within the field of computer-related technology. For example, information display in a UI where related data does not necessarily reside in a storage device locally coupled to the related display device is not practically replicable outside the field of computer-related technology. That is, in some cases, related data (e.g., training sets, physiological parameter data, images, time-lapse sets, image overlays, etc.) can be stored on a cloud storage device, while in some examples, related data can be stored locally (e.g., on a patient’s mobile device) and / or synchronized with a cloud storage device or another device such as an HCP’s mobile device at times advantageous to the system in order to shift computational resources (e.g., processing, memory, power resources, etc.). In non-limiting examples, multi-faceted information representing, for example, static content tiles, dynamically enhanced content tiles, etc. can be obtained from different systems simultaneously or dynamically, in some cases identified by metadata, and presented in interactive session UIs and sub-session UIs simultaneously or dynamically. Additionally, the monitoring system can present data visually on appropriate displays through appropriate media and / or at appropriate times (e.g., on static schedules or dynamically updated schedules). Additionally, the technology of the present disclosure provides abnormal determination techniques that can utilize non-visual sensors in some examples, such as thermal imaging through a camera, to detect temperature changes at an implant site. Additionally, various data analysis techniques are described that utilize specific examples of synthesis of various data items (e.g., images of an implant site, device status information, etc.), where information sources can be optimized to provide specific data relevant to specific technical problems of post-implant patient monitoring. That is, in some examples, the synthesis of data provides robust algorithms for identifying abnormalities, although the algorithms described are used to obtain and analyze specific portions of data for potential abnormalities, such as image data.

[0062] Additionally, it has been noted that designing a computer UI that is "usable by humans and easy to learn is a non-trivial problem for software developers" (Dillon, A. (2003) User Interface Design. In MacMillan Encyclopedia of Cognitive Science, Vol. 4, London: MacMillan, 453-458). The various examples of interactive and dynamic UIs of the present disclosure are the result of substantial research, development, refinement, iteration, and testing, and in some examples, provide a particular way of obtaining, summarizing, and presenting information in an electronic device. This non-trivial development results in the UIs described herein that likely provide significant cognitive and ergonomic efficiencies and advantages over prior systems. The interactive and dynamic UIs include improved human-machine interactions that can provide reduced mental workload / burden, improved decision making, reduced work stress, etc. for users. For example, a UI with the interactive UIs described herein can provide optimized presentation of patient-specific information from various sources and can enable users to access, navigate, evaluate, and utilize such information more quickly than using prior systems, which can be slow, complex, and / or difficult to learn, especially for novice users. Thus, presenting concise and compact information on a particular UI corresponding to a particular patient facilitates efficient use of available information and optimized use of the medical devices and virtual exam functionality of the present disclosure.

[0063] Figure 1 An environment of an example monitoring and / or registration system 100 in conjunction with a patient 4 is shown. In some examples, the system 100 can implement various patient and medical device monitoring and anomaly detection techniques disclosed herein. The system 100 includes one or more medical devices 6 and one or more computing devices 2. While in some cases the medical devices 6 include IMDs, as shown, the techniques of the present disclosure are not so limited. However, for illustrative purposes, in some cases, the medical devices 6 can be referred to herein simply as IMDs 6 or one or more IMDs 6. Figure 1

[0064] The computing device 2 can be a computing device having a display that can be viewed by a user. The user can be a physician technician, a surgeon, an electrophysiologist, a clinician (e.g., an implanting clinician), or the patient 4. The patient 4 is typically a human, but is not necessarily a human. For example, the patient 4 can be an animal that requires continuous monitoring of various health conditions (e.g., cardiac conditions, spinal cord conditions, etc.). In this case, a human caregiver can operate aspects of the disclosed technology that utilize user input that can not be available from the animal patient 4.

[0065] ​In some cases, computing device 2 can be referred to herein as a plurality of“computing devices 2,” while in other cases, it can be referred to simply as“computing device 2” when appropriate. System 100 can be implemented in accordance with one or more techniques of the present disclosure in any environment in which at least one of computing devices 2 can interface with and / or monitor at least one of medical devices 6 and / or an implant site of medical devices 6. Computing devices 2 can interface with and / or monitor medical devices 6 in accordance with one or more techniques of the present disclosure, for example, by imaging an implant site of medical devices 6. Additionally, computing devices 2 can interrogate medical devices 6 to obtain data from medical devices 6, such as performance data, historical data stored to memory, battery strength of medical devices 6, impedance, pulse width, percentage of pacing, pulse amplitude, pacing mode, internal device temperature, etc. In some examples, computing devices 2 can perform interrogation sub-sessions with medical devices 6 by establishing wireless communication with one or more of medical devices 6. In some cases, medical devices 6 can or can not include an IMD. In one example, computing devices 2 interrogate memory of wearable medical devices 6 to determine device operating parameters as interrogation data. In another example, computing devices 2 can receive (e.g., obtain) physiological parameters from medical devices 6, such as waveforms of physiological parameters, parameter markers (e.g.,“detected abnormal ECG”), etc. In another example, computing devices 2 can obtain patient status input into computing devices 2 by a user (e.g., patient 4, caregiver of patient 4, etc.). In one example, a user can input patient status updates via a user interface of computing devices 2. In another example, a user can input patient status updates via another computing device 2, which can then communicate patient status data to one of computing devices 2 configured to run an interactive enrollment session. In such examples, network 10 or edge device 12 can facilitate data exchange between various computing devices 2, medical devices 6, etc., as referenced above. Figure 3 Further details are described in further detail.

[0066] In some examples, computing devices 2 can include one or more of a cellular telephone, a“smart phone,” a satellite telephone, a notebook computer, a tablet computer, a wearable device, a computer workstation, one or more servers, a personal digital assistant, a handheld computing device, a virtual reality headset, a wireless access point, a motion or presence sensor device, or any other computing device that can run an application that enables the computing device to interact with medical devices 6 or with another computing device that is configured to, in turn, interact with medical devices 6.

[0067] At least one of the computing devices 2 can be configured to communicate with the medical devices 6 and, optionally, other computing devices of the computing devices 2 through wired or wireless communication. For example, the computing devices 2 can communicate via near-field communication (NFC) technology (e.g., inductive coupling, NFC, or other communication technology that can operate within a range of less than 10 cm to 20 cm) and / or far-field communication technology (e.g., radio frequency (RF) telemetry in accordance with 802.11, The computing devices 2 can include an interface for providing input to the edge devices 12, the network 10, and / or the medical devices 6. For example, the computing devices 2 can include a user input mechanism, such as a touch screen, that allows a user to store images to a database. In such examples, one of the edge devices 12 can manage the database, which, in some cases, the computing devices 2 can access via the network 10 in order to perform one or more of the various techniques of the present disclosure.

[0068] The computing devices 2 can include a user interface (UI) 22. In some examples, the UI 22 can be a graphical UI (GUI), an interactive UI, or the like. In some examples, the UI 22 can also include a command line interface. In some examples, the computing devices 2 and / or the edge devices 12 can include a display system (not shown). In such examples, the display system can include system software for generating UI data to be presented for display and / or interaction. In some examples, a processing circuit, such as a processing circuit of the computing devices 2, can receive UI data from another device, such as from one of the edge devices 12 or the server 94 Figure 3 ) that the computing devices 2 can use to generate UI data to be presented for display and / or interaction.

[0069] The computing devices 2 can be configured to receive input from a user through the UI 22. The UI 22 can include, for example, a keypad and a display, which can be a liquid crystal display (LCD) or light emitting diode (LED) display, for example. In some examples, the display of the computing devices 2 can include a touch screen display, and the user can interact with the computing devices 2 through the display. It is noted that the user can also interact with the computing devices 2 remotely through a network computing device.

[0070] In some examples, the UI 22 can further include a keypad. In some examples, the UI 22 can include a keypad and a display. The keypad can take the form of an alphanumeric keypad or a reduced set of keys associated with particular functions. The computing device 2 can additionally or alternatively include a peripheral pointing device, such as a mouse, through which a user can interact with the UI 22. In some cases, the UI 22 can include a UI that utilizes a virtual reality (VR), augmented reality (AR), or mixed reality (MR) UI, such as those that can be implemented via a VR, AR, or MR headset.

[0071] In some examples, processing circuitry (e.g., processing circuitry 20 of the computing device 2 Figure 2 ), processing circuitry 64 of the edge device 12 Figure 3 ), processing circuitry 98 of the server 94 Figure 3 ) can determine identification data, such as authentication data, of the patient 4. In one example, the processing circuitry 20 can identify IMD information corresponding to the IMD 6 based at least in part on the identification data. As such, the processing circuitry 20 can determine an interactive session program defining one or more parameters for imaging an implant site, such as an implant site of the patient 4, based at least in part on the IMD information. In one example, the interactive session program can define one or more parameters for imaging one or more implant sites of the patient 4. The interactive session program can include a site check sub-session application, a patient status sub-session application, and the like (e.g., a mobile application installed from an application store). In some examples, the interactive session program can also include a sub-session procedure customized for a user of the computing device 2 (e.g., the patient 4, an HCP, and the like) (e.g., a site check sub-session imaging program, a patient status sub-session program, a physiological parameter sub-session program, and the like).

[0072] In an illustrative example, the system 100 includes a system for monitoring the patient 4. The system includes a processor and a storage device system, such as those elements described with reference to Figures 2-4 In some examples, the storage device can include a memory, such as a memory device of the computing device 2, where the memory can be configured to store image data (e.g., image data frames, images of implant sites, and the like). Additionally, the storage device can be configured to store additional data items, such as physiological parameter values, patient status updates, device interrogation data, and the like. A processor system, which can include one or more processors, can be in communication with at least one of the storage devices configured to store image data.

[0073] In some examples, a computing device 2 may include one or more processors in a processor system implemented in a circuit, wherein the one or more processors may be configured to provide interactive sessions, such as virtual registration interactive sessions, via computing device 2. In such examples, the virtual registration interactive session may be configured to allow a user (e.g., via computing device 2) to navigate multiple sub-sessions. The multiple sub-sessions may include at least two sub-sessions as part of an interactive session. In one example, the multiple sub-sessions may include at least a first sub-session and a second sub-session. In such examples, the second sub-session is distinct from the first sub-session. Additionally, the first sub-session may include a sub-session involving the capture of image data via one or more cameras. That is, the first sub-session may include a site examination sub-session. The second sub-session may include a device examination sub-session, a physiological parameter examination sub-session, a patient status sub-session, or other sub-sessions configured to elicit information about patient 4 that is useful in monitoring patient 4 (e.g., an IMD patient). In some cases, the interactive session may include a third and / or fourth sub-session, which includes one or more of a device examination sub-session, a physiological parameter examination sub-session, a patient status sub-session, or other sub-sessions, as described in the specific exemplary sub-sessions herein.

[0074] In an exemplary example, computing device 2 may determine a first set of data items based on a first sub-session of an interactive session. In such an example, the first set of data items includes image data, such as image data frames. That is, computing device 2 may determine the first set of data items based on a camera (e.g., ...). Figure 2 The camera 32) captures image data frames and can store the frames in memory. In one example, the image data may include frames of still images of the implantation site of the medical device 6. In another example, the image data may include frames of still images of a region of the patient 4's body near the implantation site, such as where the lead may be routed under the patient 4's skin. In some cases, the computing device 2 may process the image data before storing it as a first set of data items. In one example, the computing device 2 may deploy an AI engine and / or ML model trained to identify anomalies in images of the patient 4's implantation site or other body parts. The AI ​​engine and / or ML model may determine anomaly metrics, such as the probability of an anomaly at the implantation site (e.g., confidence value), the type of anomaly, etc. In any case, the computing device 2 may determine that the first set of data items includes the image data (e.g., anomaly determination, etc.).

[0075] Additionally, the computing device 2 can determine a second set of data items from a second sub-session of the interactive session. In this case, the second set of data items can be different from the first set of data items. This is because the second sub-session contains a sub-session configured to obtain complementary or supplemental data related to the first sub-session, rather than serving as a duplicate of the first sub-session. To illustrate, the second set of data items can include one or more of: interrogational data obtained from the medical device 17 (e.g., the IMD 6), one or more physiological parameters of the patient 4, and / or user input data. In such examples, the computing device 2 can determine an abnormality corresponding to at least one of the patient or the IMD based at least in part on the first set of data items and the second set of data items. In one example, the first set of data items and the second set of data items can indicate various abnormality states. The abnormality states can include device migration, potential infection, healing abnormality, physiological parameter abnormality, device parameter abnormality, patient input indicative of a perception abnormality, etc. Additionally, the abnormality states can include healing granulation around an edge of an implant site, discharge from an implant site, inflammation at an implant site, tissue erosion at or around an implant site, etc. Thus, the abnormality corresponding to the patient 4 or the IMD 6 can include an abnormality determination based on various abnormality states observed from various data items.

[0076] In one example, the computing device 2 can determine an ECG change indicative of device migration and thus increase a likelihood of detecting a potential abnormality from the image data from the physiological parameters obtained through the second sub-session. In such cases, the computing device 2 can analyze the image data using a bias for detecting abnormalities, or can include an increased likelihood (e.g., probability, confidence interval) of a potential abnormality based on a likelihood determined from the first set of data items and the second set of data items in the post-implant report. The computing device 2 can output a post-implant report of the interactive session. In illustrative examples, the post-implant report can include an indication of the abnormality and / or an indication of an amount of time that has occurred since the date of implanting the IMD.

[0077] In some examples, the computing device 2 can contain a programming head or paddle (not shown). In such examples, the computing device 2 can interface with the medical device 6 through the programming head. The programming head can be placed near the medical device 6 (e.g., near an implant site of the IMD 6) proximate to the body of the patient 4. The computing device 2 can contain the programming head in order to improve a quality or security of communication between the computing device 2 and the medical device 6. Additionally, the computing device 2 can contain the programming head in order to improve a quality or security of communication between the computing device 2, the medical device 6, and / or the edge device 12.

[0078] In Figure 1 In illustrative and non-limiting examples, the medical device 6 includes at least one IMD. In such examples, the at least one IMD can be implanted outside of a thoracic cavity of the patient 4 (e.g., subcutaneously implanted inFigure 1 The medical device 6 can be positioned in proximity to the pectoral muscle (e.g., at least partially within the cardiac silhouette) at or just below the heart level of the patient 4 (e.g., in the pectoral muscle position shown in FIG. 1). In some examples, the medical device 6 can be positioned in proximity to the pectoral muscle near or just below the heart level of the patient 4, e.g., at least partially within the cardiac silhouette. As used herein, an IMD can include, be, or be part of, various devices or integrated systems such as, but not limited to, implantable cardiac monitors (ICMs), implantable pacemakers, including those that deliver cardiac resynchronization therapy (CRT), implantable cardioverter-defibrillators (ICDs), diagnostic devices, cardiac devices, etc. In some examples, the tools of the present disclosure can be configured to monitor the functioning of implants other than CIEDs or the patient’s acclimation to the implants, such as spinal stimulators, deep brain stimulators, gastric stimulators, urinary system stimulators, other neural stimulators, orthopedic implants, respiratory monitoring implants, etc.

[0079] In some examples, the medical device 6 can include one or more CIEDs. In some examples, the patient 4 can interact with multiple medical devices 6 simultaneously. In an illustrative example, the patient 4 can have multiple IMDs implanted within the patient’s 4 body. In another example, the medical device 6 can include a combination of one or more implantable and / or non-implantable medical devices. Examples of non-implantable medical devices include wearable devices (e.g., a monitoring watch, a wearable defibrillator, etc.) or any other external medical device configured to obtain physiological data of the patient 4.

[0080] In some examples, the medical device 6 can include a diagnostic medical device. In one example, the medical device 6 can include a device that predicts a heart failure event or detects a worsening of heart failure in the patient 4. In a non-limiting and illustrative example, the system 100 can be configured to measure impedance fluctuations in the patient 4 and process the impedance data to accumulate evidence of a worsening of heart failure. In any case, the medical device 6 can be configured to determine a health condition related to the patient 4. The medical device 6 can transmit diagnostic data or a health status as interrogation data to the computing device 2 so that the computing device 2 can correlate the interrogation data with the image data to determine whether an abnormality (e.g., an infection at an implant site) exists in the particular one of the medical device 6 (e.g., an IMD) or the patient 4.

[0081] In some examples, the medical device 6 can operate as a therapy delivery device. For example, the medical device can deliver electrical signals to the heart of the patient 4, such as an implantable pacemaker, a cardiac cardioverter and / or defibrillator, a drug delivery device that delivers a therapeutic substance through one or more catheters to the patient 4, or a combination therapy device that delivers both electrical signals and therapeutic substances. As described herein, the computing device 2 can determine various interactive session procedures or at least some aspects of the interactive session procedures (e.g., imaging procedures, UI procedures, etc.) based on the type of medical device implanted in the patient 4 or based on identifying information of the patient 4.

[0082] It should be noted that while certain example medical devices 6 are described as being configured to monitor cardiovascular health, the techniques of the present disclosure are not so limited, and one of skill in the art will appreciate that the techniques of the present disclosure can be implemented in other contexts (e.g., neurology, orthopedics, etc.). In some examples, one or more of the medical devices 6 can be configured to perform deep brain stimulation (DBS), spinal cord stimulation (SCS), pelvic stimulation, peripheral nerve stimulation, muscle stimulation, etc.

[0083] Additionally, while certain example medical devices 6 are described as insertable or implantable devices, the techniques of the present disclosure are not so limited, and one of skill in the art will appreciate that the techniques of the present disclosure can be implemented with medical devices 6 that are not configured to be insertable or implantable, such as wearable devices or other external medical devices. In non-limiting examples, the medical devices 6 can include wearable devices (e.g., smartwatches, head-mounted devices, etc.) configured to obtain physiological data (e.g., activity data, heart rate, etc.) and communicate such data to the computing device 2, the network 10, the edge device 12, etc. for subsequent utilization in accordance with one or more of the various techniques of the present disclosure.

[0084] Furthermore, while certain example medical devices 6 are described as electrical devices or electrically-powered devices, the techniques of the present disclosure are not so limited, and one of skill in the art will appreciate that, in some examples, the medical devices 6 can include non-electrical or non-electrically-powered devices (e.g., orthopedic implants, etc.). In any case, the medical devices 6 can be configured to communicate medical data to the computing device 2, such as via telemetry protocols, radio frequency identification (RFID) transmissions, etc. Thus, any medical device and / or computing device configured to communicate medical data can be configured to implement the techniques of the present disclosure.

[0085] In some examples, the medical devices 6 can be implanted subcutaneously to the patient 4. Further, in some examples, the computing device 2 can monitor subcutaneous impedance values obtained from the medical devices 6. In some examples, at least one of the medical devices 6 takes the form of a Reveal LINQ® TM insertable cardiac monitor (ICM) or a similar device to the LINQ® ICM developed by Medtronic, Inc. of Minneapolis, MN, for example. In such examples, the medical devices 6 can facilitate relatively long-term monitoring of the patient during normal daily activities. TM a version or modification of the ICM or another ICM. In such examples, the medical devices 6 can facilitate relatively long-term monitoring of the patient during normal daily activities.

[0086] In non-limiting examples, the medical device 6 can include an IMD configured to operate as a pacemaker, cardioverter, and / or defibrillator or otherwise monitor electrical activity of the heart of the patient 4. In some examples, the medical device 6 can provide pacing pulses to the heart of the patient 4 based on electrical signals sensed within the heart of the patient 4.

[0087] In some examples, the medical device 6 can also provide defibrillation therapy and / or cardioversion therapy through electrodes positioned on at least one lead and / or housing electrodes. The medical device 6 can detect an arrhythmia, such as ventricular fibrillation, of the heart of the patient 4 and deliver defibrillation therapy to the heart of the patient 4 in the form of electrical pulses. In some examples, the medical device 6 can be configured to deliver a series of therapies, e.g., pulses of increasing energy levels, until the fibrillation of the heart of the patient 4 stops. In such examples, the medical device 6 can employ one or more fibrillation detection techniques known in the art to detect fibrillation.

[0088] In some examples, the system 100 can be implemented in an environment that includes the network 10 and / or the edge devices 12. That is, in some examples, the system 100 can operate under the environment of the network 10 and / or include one or more edge devices 12. In some cases, the network 10 can include the edge devices 12. Similarly, the computing device 2 can include the functionality of the edge devices 12 and thus can also function as one of the edge devices 12.

[0089] In some examples, the edge devices 12 include modems, routers, Internet of Things (IoT) devices or systems, smart speakers, screen-enhanced smart speakers, personal assistant devices, and the like. Additionally, the edge devices 12 can include user-facing devices or client devices, such as smartphones, tablet computers, personal digital assistants (PDAs), and other mobile computing devices.

[0090] In examples involving the network 10 and / or the edge devices 12, the system 100 can be implemented in a home environment, a hospital environment, or any environment that includes the network 10 and / or the edge devices 12. The example techniques can be used with the medical device 6, which can be in wireless communication with one or more of the edge devices 12 and other devices (e.g., network servers) not depicted in FIG. 1. Figure 1

[0091] In some examples, the computing device 2 can be configured to communicate with one or more of the medical device 6, the edge devices 12, or the network 10, which operates a network service such as the Medtronic CareLink® Network, developed by Medtronic, Inc. of Minneapolis, MN. In some examples, the medical device 6 can be in wireless communication with the network 10 through the computing device 2. In some examples, the medical device 6 can be in wireless communication with the network 10 through the computing device 2. ​In communication with computing device 2. In some cases, network 10 can include one or more of edge devices 12. Network 10 can be and / or include any appropriate network, including a private network, a personal area network, an intranet, a local area network (LAN), a wide area network, a wired network, a satellite network, a cellular network, a point-to-point network, a global network (e.g., the Internet), a cloud network, an edge network, a device network, or the like, or a combination thereof, some or all of which can or can not be connected to and / or from the Internet. That is, in some examples, network 10 includes the Internet. In illustrative examples, computing device 2 can periodically transmit and / or receive various data items to and / or from one of medical devices 6 and / or edge devices 12 via network 10.

[0092] Additionally, computing device 2 can be configured to connect to cellular base transceiver stations (e.g., for 3G, 4G, LTE, and / or 5G cellular network access) and Wi-Fi TM access points, as these connections are available. In some examples, the cellular base transceiver stations can have connections to network 10. These various cellular and Wi-Fi TM network connections can be managed by different third-party entities, which are referred to herein as “carriers.”

[0093] In some examples, system 100 can include one or more databases (e.g., storage device 96) that store various medical data records, cohort data, image data. In such examples, server 94 (e.g., the one or more databases) can be managed or controlled by one or more separate entities (e.g., an Internet service provider (ISP), etc.).

[0094] The computing device 2 and / or the edge device 12 can be used to configure operational parameters of the medical device 6. In some examples, a user can use the computing device 2 as a programmer to program measurement parameters, stimulation programs, etc. The computing device 2 can also be configured to program therapy progress, select electrodes to deliver defibrillation pulses, select waveforms for defibrillation pulses or stimulation pulse sequences, select or configure fibrillation detection algorithms, etc. The user can also use the computing device 2 to program aspects of other therapies that can be provided by the medical device 6, such as cardioversion or pacing therapy. In some examples, the user can activate certain features of the medical device 6 by entering a single command via the UI 22 of the computing device 2, such as pressing a single key or combination of keys of a keypad or a single point selection action with a pointing device. Additionally, the computing device 2 can operate an interactive session of the present disclosure, where the interactive session is loaded with programming parameters. The computing device 2 can utilize such data to determine physiological parameters and / or interrogate data that deviates from expected values of the programming parameters. The computing device 2 can utilize an AI engine and / or an ML model to determine such deviations (e.g., anomalies), where the AI engine and / or the ML model can be trained from the programming parameters and the anomaly data to determine correlations between such data items.

[0095] In some examples, the computing device 2 can be configured to retrieve data from the medical device 6. The retrieved data can include values of physiological parameters measured by the medical device 6, indications of cardiac arrhythmias or other disease episodes detected by the medical device 6, and physiological signals obtained by the medical device 6. In some examples, the computing device 2 can retrieve a cardiac EGM segment recorded by the computing device 2, for example, as a result of the computing device 2 determining that a cardiac arrhythmia or another disease episode occurred during the segment, or in response to a request from the patient 4 or another user to record the segment.

[0096] In some examples, the user can also use the computing device 2 to retrieve information from the medical device 6 regarding other sensed physiological parameters of the patient 4, such as activity or posture. In some examples, the edge device 12 can interact with the medical device 6 in a similar manner to the computing device 2, for example, to program the medical device 6 and / or retrieve data from the medical device 6.

[0097] The processing circuitry of the system 100, e.g., the medical device 6, the computing device 2, the edge device 12, and / or one or more other computing devices (e.g., a remote server), can be configured to perform example techniques of the present disclosure to determine an abnormal state of the patient 4 and / or an IMD 6 component. In some cases, the processing circuitry can be referred to herein as a processor system or processing circuitry. In some examples, the processing circuitry of the system 100 obtains physiological parameters, images, medical device diagnostics, etc. to determine whether to provide an alert to the patient 4 and / or an HCP.

[0098] In some examples, the processing circuitry of system 100, e.g., computing device 2, can determine a projection of a change characteristic of a particular location of the body over time based on one or more images, including but not limited to images before and / or after implantation or explantation procedures. The processing circuitry can compare the second set of change characteristics to the projection determined from the second set of images in order to identify potential abnormalities (e.g., infection, drainage, device migration, erosion, etc.) at the particular location of the body.

[0099] In some cases, when patient health data (e.g., implant site images, ECG parameters, etc.), medical device diagnostic data combine to indicate the onset of an abnormality, system 100, e.g., processing circuitry of computing device 2, provides an alert to patient 4 and / or other users and indicates the onset of the abnormality. The alert can be an audible alert generated by medical device 6 and / or computing device 2, a visual alert generated by computing device 2, such as a text prompt or a flashing button or screen, or a haptic alert generated by medical device 6 and / or computing device 2, such as a vibration or vibration pattern. In addition, the alert can be provided to other devices, e.g., over network 10. Several different levels of alerts can be used based on the severity of the potential infection detected according to one or more of the various techniques disclosed herein.

[0100] In some cases, when patient health data (e.g., implant site images, ECG parameters, etc.), medical device diagnostic data combine to indicate the onset of an abnormality, system 100, e.g., processing circuitry of computing device 2, provides an alert to patient 4 and / or other users and indicates the onset of the abnormality. The process of determining when to issue an alert to patient 4 involves measuring an abnormality (e.g., a severity or probability level) against one or more thresholds and is described in more detail below. The alert can be an audible alert generated by medical device 6 and / or computing device 2, a visual alert generated by computing device 2, such as a text prompt or a flashing button or screen, or a haptic alert generated by medical device 6 and / or computing device 2, such as a vibration or vibration pattern. In addition, the alert can be provided to other devices, e.g., over network 10. Several different levels of alerts can be used based on the severity of the potential abnormality detected by the techniques disclosed herein.

[0101] Figure 2 is a block diagram showing an example configuration of components of at least one computing device 2. In Figure 2 In an example of the at least one computing device 2, the at least one computing device 2 includes processing circuitry 20, communication circuitry 26, storage 24, and UI 22.

[0102] The processing circuitry 20 can include one or more processors configured to implement functionality and / or process instructions for execution within the computing device 2. For example, the processing circuitry 20 can process instructions stored in the storage device 24. The processing circuit 20 can include, for example, microprocessors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), complex programmable logic devices (CPLDs), or equivalent integrated or discrete logic circuitry, or combinations of any of the foregoing. Accordingly, the processing circuitry 20 can include any suitable structure, whether in hardware, software, firmware, or any combination thereof, to execute the functionality described herein as being attributed to the processing circuitry 20.

[0103] The trained ML model 30 and / or AI engine 28 can be configured to process and analyze user inputs (e.g., images of implant sites, patient status data, etc.), device parameters (e.g., accelerometer data), historical data of medical devices (e.g., medical device 6), and / or physiological parameters in accordance with certain examples in which ML models are considered advantageous in the present disclosure (e.g., predictive modeling, inferential detection, context matching, natural language processing, etc.). Examples of ML models and / or AI engines that can be configured to perform aspects of the present disclosure include classifiers and non-classifier ML models, artificial neural networks (“NN”), linear regression models, logistic regression models, decision trees, support vector machines (“SVM”), naive or non-naive Bayesian networks, k-nearest neighbor (“KNN”) models, deep learning (DL) models, k-means models, clustering models, random forest models, or any combination thereof. Depending on the implementation, the ML models can be supervised, unsupervised, or in some cases a hybrid combination (e.g., semi-supervised). These models can be trained based on data indicative of how a user (e.g., patient 4) interacts with the computing device 2. For example, for illustrative purposes only, certain aspects of the present disclosure will be described using events or actions (such as clicks, views, or watches) related to items (e.g., wound images, cameras, videos, physiological parameters, etc.). In another example, these models and engines can be trained to synthesize data to identify anomalies of the patient 4 or medical device 17 and identify anomalies of the patient 4 or medical device 17 from individual data items.

[0104] In a non-limiting example, patient 4 can have difficulty capturing images of the implant site from various angles. These useful data can be shared in a health monitoring or computing network such that optimal results can be presented to more than one user based on similar queries and user reactions to those queries. For brevity, these aspects can not be described in relation to events or behaviors of the subject (e.g., data objects such as search strings). In some examples, processing circuitry 40 can use ML algorithms (e.g., DL algorithms) to, for example, monitor the progress of a healing wound or predict a potential infection that is occurring, for example, for an implant site of one of medical devices 17. In an illustrative and non-limiting example, AI engine 28 and / or ML model 30 can utilize a deep neural network to locate an implant site in an image and classify abnormal states. In another example, AI engine 28 and / or ML model 30 can utilize Naive Bayes and / or decision trees to synthesize (e.g., combine) data items and their analysis (e.g., image analysis and ECG analysis) in order to obtain a comprehensive abnormality determination for patient 4 and include such comprehensive determinations in reports such as for patient 4.

[0105] In another example, AI engine 28 and / or ML model 30 can be loaded with and trained on group parameters (e.g., age group, IMD type group, skin pigmentation group, etc.) and combinations of group parameters. In such examples, AI engine 28 and / or ML model 30 can utilize historical reference queries for comparison in determining deviations from baseline parameters.

[0106] Additionally, computing device 2 can utilize different image processing algorithms and / or data synthesis algorithms, such as AI, ML, DL, digital signal processing, neural networks, and / or other techniques, depending on various different environments and other various different resource constraints (e.g., processing power, network access, battery life, etc.).

[0107] In some examples, AI engine 28 can be trained to analyze a patient’s gait with an orthopedic implant in place (e.g., by comparing video data representing the patient to group gait information). In such examples, computing device 2 can deploy AI engine 28 and / or ML model 30 to analyze patient activity to determine the presence of potential abnormalities of patient 4 and / or medical device 17 (e.g., IMD 6), such as when a patient’s gait appears abnormal to AI engine 28 and / or ML model 30, and / or when a patient’s gait changes over time to represent a developing or sudden abnormality of patient 4.

[0108] Additionally, the trained AI engine 28 can be used to learn about the patient 4 over time and learn about the implant site of the patient 4. In this way, the AI engine 28 can provide a personalized detection algorithm for detected abnormalities at a particular implant site. In such examples, the AI engine 28 can be loaded with and trained on group parameters using historical reference interrogations or images for comparison. In this way, the computing device 2 can use different algorithmic approaches (e.g., AI, ML, DL, digital signal processing, neural networks, and / or other techniques) to provide a solution in order to personalize the site check process for the patient 4.

[0109] A user, such as a clinician or the patient 4, can interact with one or more of the computing devices 2 through a UI 22. The UI 22 includes a display (not shown), such as a liquid crystal display (LCD) or light emitting diode (LED) display or other type of screen, through which the processing circuitry 20 can present health or device related information, e.g., a cardiac EGM, an indication of impedance change detection, a temperature change, etc. Additionally, the UI 22 can include input mechanisms for receiving input from the user. The input mechanisms can include, for example, buttons, a keypad (e.g., an alphanumeric keypad), a peripheral pointing device, a touchscreen, or another input mechanism that allows the user to navigate through and provide input to the UI 22 presented by the processing circuitry 20 of the computing device 2. In one example, the UI 22 can allow the user to rotate images and adjust zoom levels (e.g., pinch-to-zoom, gestures, gaze tracking, etc.) using a touchscreen of the computing device 2. Additionally, the UI 22 can allow the user to control camera parameters, such as selecting a front or rear camera, lighting, zoom level, focus level, contrast, etc., such that the user can capture images according to the camera parameters. In another example, the computing device 2 can automatically adjust such parameters without initial input from the user.

[0110] In some examples, the computing device 2 can include an imaging device. In an illustrative example, the computing device 2 can include a camera 32 or multiple cameras 32 (e.g., a digital camera) as an example imaging device. As shown, the camera 32 can refer to a collection of one or more image sensors 34, one or more lenses 36, and one or more camera processors 38. In some examples, the processing circuitry 20 can incorporate the camera processor 38. Figure 2

[0111] ​In some examples, the computing device 2 can include an imaging device, such as a camera 32. The camera 32 can be a digital camera built into the computing device 2. In some examples, the camera 32 can be separate from the computing device 2. In such examples, the camera 32 can communicate imaging data to the computing device 2 and / or other computing devices (e.g., the edge device 2). In some examples, the computing device 2 can include a charge-coupled device (CCD) chip. The CCD chip can be configured to operate in a spectral response analysis mode using a flash as an excitation (e.g., white light) light source. In such examples, the computing device 2 can employ the CCD chip to analyze images of the implant site. In one example, the computing device 2 can employ the CCD chip to provide separate color filtering analysis at different light wavelengths to better detect redness and swelling in different skin tones. That is, the computing device 2 can perform color filtering on the images to identify abnormalities from the one or more images. In another example, the computing device 2 can perform color filtering on the images to identify abnormalities from the one or more images. In an illustrative example, the computing device 2 can perform a region comparison of the body region to distinguish the implant site from another region of the body and for different skin types (e.g., skin tones) in accordance with one or more of the various techniques of the present disclosure.

[0112] In some examples, the computing device 2 can image the implant site of the patient 4 using the camera 32. In such examples, the storage device 24 of the computing device 2 can store image data (e.g., still images, etc.). In this case, processing circuitry, such as the processing circuitry 20 of the computing device 2 and / or the processing circuitry 64 of the edge device 12 Figure 3 may communicate with the storage device 24.

[0113] In some examples, the processing circuitry (e.g., the processing circuitry 20) can determine, in some cases, a projection of a change characteristic of a particular location of the body of the patient 4 over time based on one or more images, including but not limited to images before and / or after an implant or explant procedure. The processing circuitry 20 can compare the second set of change characteristics to the projection determined from the second set of images in order to identify potential abnormalities (e.g., infection, drainage, device migration, erosion, etc.) at the particular location of the body of the patient 4. The images can be captured by one or more cameras 32, which can or can not be separate from the computing device 32. In some cases, the processing circuitry 20 can include one or more of the camera processors 38 (e.g., image signal processors). In another example, the camera processors 38 can be separate from the processing circuitry 20 and / or included in a separate computing device 2, server 94, edge device 12, etc.

[0114] In some examples, the multiple cameras 32 can be contained in a single computing device (e.g., a mobile phone having one or more front-facing cameras and one or more rear-facing cameras) in computing device 2. In some examples, computing device 2 can include a first camera 32 having one or more image sensors 34 and one or more lenses 36, and a second camera 32 having one or more image sensors 34 and one or more lenses 36, etc. It should be noted that while some example techniques herein can be discussed with reference to frames received from a single camera (e.g., from a single image sensor), the techniques of the present disclosure are not so limited, and one of skill in the art will appreciate that the techniques of the present disclosure can be implemented for any type of camera 32 and combination of cameras 32, such as combinations of cameras 32 that can be included in computing device 2 or otherwise communicatively coupled to computing device 2. In some examples, image sensor 34 represents one or more image sensors 34 that can include image sensor processing circuitry. In some examples, image sensor 34 contains an array of pixel sensors (e.g., pixels) for capturing a representation of light.

[0115] While shown (e.g., via dashed lines) as optionally included in computing device 2, the techniques of the present disclosure are not so limited, and in some cases, camera 32 can be separate from computing device 2, such as a standalone camera device or a separate camera system. In any case, camera 32 can be configured to capture images of an implantation site and transmit image data to processing circuitry 20 via camera processor 38. In some examples, camera 32 can be configured to implement various zoom levels. In one example, camera 32 can be configured to perform cropping and / or zooming techniques to implement a particular zoom level. In some examples, camera 32 can be configured to manipulate output from image sensor 34 and / or manipulate lens 36 in order to implement a particular zoom level.

[0116] In some examples, the computing device 2 can include audio circuitry (not shown) for providing audible notifications, instructions, or other sounds to a user and receiving voice commands from the user, or both. In some examples, the computing device 2 can provide audible notifications to the user indicating directions through the UI 22 and / or the virtual enrollment process. Additionally, the computing device 2 can provide audible notifications indicating when a task is complete, such as an authentication task (e.g., successful login), an ECG measurement task is complete, or any other task is complete. In another example, the computing device 2 can provide different audible notifications depending on a particular outcome. In one example, when the outcome of the interactive session is that no follow-up appointment is needed (e.g., no underlying infection, device operation is correct, physiological parameters are good, etc.), the computing device 2 can provide a static notification, while when the outcome of the interactive session indicates that a follow-up appointment is recommended, the computing device 2 can provide a louder notification. In one example, the computing device 2 can provide an audible notification when a likelihood of one or more abnormalities is identified. In another example, the computing device 2 can provide an audible notification upon identifying a collective abnormality (e.g., after data synthesis). In an illustrative example, the computing device 2 can determine a collective abnormality based on underlying abnormalities identified at the implant site and based on abnormalities in physiological parameters (e.g., ECG abnormalities). In some cases, when no abnormality (e.g., implant site abnormality) is identified in one sub-session but an abnormality (e.g., IMD abnormality) is identified in another sub-session, the computing device 2 can still determine that an abnormality that is potentially threatening to health exists such that the analysis of the at least two sub-sessions indicates the presence or absence of an abnormality that requires a follow-up appointment recommendation. Additionally, the computing device 2 can provide different audible notifications after each sub-session and then provide a different audible notification again after all prohibited sub-sessions (e.g., two sub-sessions, three sub-sessions, etc.) are complete.

[0117] The communication circuitry 26 can include any suitable hardware, firmware, software, or any combination thereof for communicating with another device, such as the medical device 6. Under the control of the processing circuitry 20, the communication circuitry 26 can receive downlink telemetry from and send uplink telemetry to the medical device 6 or another device. The communication circuitry 26 can be configured to transmit or receive signals via inductive coupling, electromagnetic coupling, NFC, RF communication, Wi-Fi TM or other proprietary or non-proprietary wireless communication schemes. The communication circuitry 26 can also be configured to communicate with devices other than the medical device 6 through any of a variety of forms of wired and / or wireless communication and / or network protocols. In some examples, the computing device 2 can perform telemetry selection by scanning (e.g., TelC, TelB, TelM, etc.).

[0118] The storage device 24 can be configured to store information within the computing device 2 during operation. The storage device 24 can include a computer-readable storage medium or computer-readable storage device. In some examples, the storage device 24 includes one or more of a short-term memory or a long-term memory. The storage device 24 can include, for example, read-only memory (ROM), random-access memory (RAM), non-volatile RAM (NVRAM), dynamic RAM (DRAM), static RAM (SRAM), magnetic disk, optical disk, flash memory, various forms of electrically erasable programmable ROM (EEPROM), or erasable programmable ROM (EPROM), or any other digital media.

[0119] In some examples, the storage device 24 is used to store data indicative of instructions executed by the processing circuitry 20. Additionally, the storage device 24 can store image data and / or supplemental image data. In some examples, the storage device 24 can store frames of image data. That is, the storage device 24 can store one or more images. In some examples, the storage device 24 can store one or more images of the patient’s 4 body (e.g., an implant site image, a skin image for skin color analysis, a skin surface proximate to lead wiring, etc.), a physiological parameter image (e.g., an ECG image), etc. The storage device 24 can be used by software or applications running on the computing device 2 to temporarily store information during program execution. The storage device 24 can also store historical medical device data, historical patient data, timing information (e.g., days since implant of an IMD, days since a particular physiological parameter exceeded a certain threshold, etc.), AI and / or ML training sets, image data, etc.

[0120] Data exchanged between the computing device 2, the edge device 12, the network 10, and the medical device 6 can include operational parameters of the medical device 6. The computing device 2 can transmit data including computer-readable instructions to the medical device 6. The medical device 6 can receive and implement the computer-readable instructions. In some examples, when implemented by the medical device 6, the computer-readable instructions can control the medical device 6 to change one or more operational parameters, export collected data, etc. In an illustrative example, the processing circuitry 20 can transmit instructions to the medical device 6 requesting that the medical device 6 export collected data (e.g., ECG, impedance values, etc.) to the computing device 2, the edge device 12, and / or the network 10. The computing device 2, the edge device 12, and / or the network 10 can in turn receive the collected data from the medical device 6 and store the collected data in, for example, the storage device 24. Additionally, the processing circuitry 20 can transmit interrogation instructions to the medical device 6 requesting that the medical device 6 output operational parameters (e.g., battery, impedance, pulse width, percentage of pacing, etc.).

[0121] In some examples, the computing device 2 may be coupled to an external electrode or to an implantable electrode via a percutaneous lead. In such examples, according to one or more techniques disclosed herein, the computing device 2 can receive and monitor physiological parameters, ECG, etc., from the medical device 6.

[0122] exist Figure 2 In the examples shown, the processing circuitry 20 is configured to perform various techniques described herein, such as those described in the references. Figures 5-22 The described techniques. To avoid confusion, the processing circuit system 20 is described as performing various processing techniques prohibited by the computing device 2, but it should be understood that at least some of these techniques can also be performed by other processing circuit systems (e.g., the processing circuit system 40 of the medical device 17). Figure 4 The processing circuitry system 20 (such as the processing circuitry system 98 of server 94, the processing circuitry system 64 of edge device 12, etc.) performs the operation. In an exemplary example, the processing circuitry system 20 may capture an image of the implantation site, obtain other data items, output the image and / or other data items to an analysis platform for infection detection, determine the presence of potential abnormalities (e.g., infection) based on the image and other data items, output a summary of the implantation status containing information on potential abnormalities, and in some cases, transmit the summary of the implantation status containing information on potential infection to another device. In this example, the analysis platform may be a separate program executed by the processing circuitry system 20. In another example, the analysis platform may be a separate program executed alternatively by other processing circuitry systems from the processor system.

[0123] Figure 3 This is a block diagram illustrating an example system 300 comprising one or more example computing devices 2, one or more medical devices 17, a network 10, one or more edge devices 12, and one or more servers 94, according to one or more technologies disclosed herein. In some examples, system 300 is referenced... Figure 1 An example of system 100 is described. In another example, system 300 illustrates an exemplary network system hosting monitoring system 100. In some examples, medical device 17 may be... Figure 1 Example of medical device 6. That is, medical device 17 may include IMD, CIED, etc., similar to the reference. Figure 1 As described. Additionally, medical device 17 may include non-implantable medical devices, including wearable medical devices (e.g., smartwatches, wearable defibrillators, etc.).

[0124] The medical devices 17 can be configured to transmit data, such as sensed, measured, and / or determined values of physiological parameters (e.g., heart rate, impedance measurements, fluid index, respiration rate, activity data, electrocardiogram (EGM), historical physiological data, blood pressure values, etc.) to the edge devices 12, computing devices 2, and / or access points (e.g., gateways). In some examples, the medical devices 17 can be configured to determine a plurality of physiological parameters. For example, the medical devices 17 can include medical devices 6 (e.g., IMDs) configured to determine respiration rate values, subcutaneous tissue impedance values, EGM values, etc. The edge devices 12 and / or access point devices can then communicate the retrieved data to the server 94 via the network 10.

[0125] In some examples, the medical devices 17 can transmit data to the server 94, edge devices 12, or computing devices 2 through wired or wireless connections. For example, the server 94 can receive data from the medical devices 17 (e.g., IMDs 6, wearable devices, etc.) or from the edge devices 12. In another example, the edge devices 12 can receive data from the server 94 through the network 10. In some examples, the edge devices 12 can receive data from the medical devices 17 through the network 10 or through wired or wireless connections. In such examples, the edge devices 12 can determine data received from the server 94, medical devices 17, or computing devices 2. In some examples, the edge devices 12 can store data to storage 62 internal to the edge devices 12. The processing circuitry 64 of the edge devices 12 can also include AI engines and / or ML models, as described with reference to Figure 2 the AI engines and ML models described.

[0126] In this example, the medical devices 17 can communicate with one of the edge devices 12 using the communication circuitry 42 over a first wireless connection. In some examples, the medical devices 17 can communicate with an access point using the communication circuitry 42 over a second wireless connection. The access point can include a device that connects to the network 10 through any of a variety of connections, such as a telephone dial-up, a digital subscriber line (DSL), or a cable modem connection. In some examples, the access point can be coupled to the network 10 through different forms of connections including wired or wireless connections. In some examples, one of the computing devices 2 can function as an access point for the network 10. For example, a user device such as a tablet computer or a smart phone can be located in the same location as the patient 4 and can be configured to function as an access point. In any case, the computing devices 2, edge devices 12, and server 94 are interconnected and can communicate with each other through the network 10.

[0127] The medical devices 17, edge devices 12, and / or computing device 2 can be configured to communicate with remote computing resources (e.g., servers 94) on the network 10 via various connections. A data network (e.g., network 10) can be implemented by the servers 94 (e.g., data servers, deletion servers, analysis servers, etc.). In one example, the servers 94 can include a data server configured to store data and / or perform computations based on the data. In another example, the servers 94 can include a data server configured to store data (e.g., a database) and send data to another one of the servers 94 for data analysis, image processing, or other data computations in accordance with one or more of the various techniques of the present disclosure. In some examples, the servers 94 can be implemented on one or more host devices, such as a blade server, a mid-size computing device, a mainframe computer, a desktop computer, or any other computing device configured to provide computing services and resources. Protocols and components for communicating over the Internet or any of the other above-described types of communication networks are known to those skilled in the computer communications art and, as such, need not be more specifically described herein.

[0128] In some examples, one or more of the medical devices 17 can function as or include the servers 94. That is, the medical devices 17 can include storage capacity or processing power sufficient to perform the techniques disclosed herein on a single one of the medical devices 17 or on a network of medical devices 17 coordinating tasks over the network 10 (e.g., on a private or closed network). In some examples, one of the medical devices 17 can include at least one of the servers 94. For example, a portable / bedside patient monitor can be configured to function as one of the servers 94 and as one of the medical devices 17 configured to obtain physiological parameter values from the patient 4.

[0129] In some examples, the servers 94 can communicate with each of the medical devices 17 over a wired or wireless connection to receive physiological parameter values and / or device interrogation data from the medical devices 17. In a non-limiting example, the physiological parameter values and / or device interrogation data can be communicated from the medical devices 17 to the servers 94 and / or edge devices 12. The servers 94 and / or edge devices 12 can perform analysis on the data to determine the presence of an abnormality at an implant site (e.g., a location in the body of the patient 4 where one or more IMD components are located). The servers 94 and / or edge devices 12 can transmit the results of the analysis to the computing device 2 over the communication circuitry for display and / or further processing.

[0130] In some examples, the server 94 can be configured to provide a secure storage site for data that has been collected from the medical device 17, the edge device 12, and / or the computing device 2. In some cases, the server 94 can include a database that stores medical and health-related data. For example, the server 94 can include a cloud server or other remote server that stores data collected from the medical device 17, the edge device 12, and / or the computing device 2. In some cases, the server 94 can compile data through the computing device 2 in a web page or other document for viewing by a trained professional, such as a clinician. In Figure 3 In the illustrated example, the server 94 includes a storage device 96 (e.g., for storing data retrieved from the medical device 17) and processing circuitry 98. As described with reference to Figure 2 The computing device 2 can similarly include a storage device and processing circuitry, as described.

[0131] The processing circuitry 98 can include one or more processors configured to implement functionality and / or process instructions for execution within the server 94. For example, the processing circuitry 98 can process instructions stored by the storage device 96. The processing circuitry 98 can include, for example, microprocessors, DSPs, ASICs, FPGAs, or equivalent integrated or discrete logic circuitry, or combinations of either, or any of the foregoing. Accordingly, the processing circuitry 98 can include any suitable structure, whether in hardware, software, firmware, or any combination thereof, to

[0132] The storage device 96 can include computer-readable storage media or computer-readable storage devices. In some examples, the storage device 96 includes one or more of short-term memory or long-term memory. The storage device 96 can include, for example, ROM, RAM, NVRAM, DRAM, SRAM, magnetic disk, optical disk, flash memory, various forms of EEPROM, or EPROM, or any other digital media. In some examples, the storage device 96 is used to store data indicative of instructions executed by the processing circuitry 98.

[0133] In some examples, one or more of the computing devices 2 can be a tablet computer or other smart device placed by a clinician (or other HCP) through which the clinician can program the medical device 17, receive alerts from the medical device, and / or interrogate the medical device. For example, the clinician can access data collected by the medical device 17 through the computing device 2, such as to check the status of a medical condition when the patient 4 is between clinician visits. In some examples, the computing device 2 can transmit data regarding images, infection or other abnormality indications, training sets, ML models, IMD information, patient identification data, authentication data, etc. to one or more other computing devices 2, edge devices 12, and / or servers 94. Likewise, the computing device 2 can receive similar information.

[0134] In further examples, the computing device 2 can generate an alert to the patient 4 (or forward an alert determined by the medical device 17, edge device 12, or server 94) based on an abnormality determined from a combination of data items, which can cause the patient 4 to proactively seek medical assistance before receiving instructions for medical intervention. In some examples, the computing device 2 can transmit the alert to the patient 4 via a text message, email, or other message. In some examples, the computing device 2 can transmit the alert to the patient 4 via a phone call or other communication. Figure 3 In the illustrated example, the server 94 includes a storage device 96 (e.g., for storing data retrieved from the medical device 17) and processing circuitry 98. As described with reference to the computing device 2, the server 94 can include a display device 100 and input device 102. The server 94 can be configured to receive data from the medical device 17 via the network 10, and to transmit data to the medical device 17 via the network 10. The server 94 can be configured to receive data from the computing device 2 via the network 10, and to transmit data to the computing device 2 via the network 10. The server 94 can be configured to receive data from the edge device 12 via the network 10, and to transmit data to the edge device 12 via the network 10. Figure 2 As described with reference to the computing device 2, the server 94 can be configured to receive data from the medical device 17 via the network 10, and to transmit data to the medical device 17 via the network 10. The server 94 can be configured to receive data from the computing device 2 via the network 10, and to transmit data to the computing device 2 via the network 10. The server 94 can be configured to receive data from the edge device 12 via the network 10, and to transmit data to the edge device 12 via the network 10.

[0135] In some examples, a clinician can input instructions for medical intervention for the patient 4 into an application executed by the computing device 2, such as based on a status of the patient condition determined by another of the computing device 2, medical device 17, edge device 12, server 94, or any combination thereof, or based on other patient data known to the clinician. The patient condition can include an implant status, such as a potential infection at an implant site of the implant. The user’s computing device 2 can receive the instructions over the network 10. The computing device 2 in turn can display a message on a display device indicating a message for medical intervention.

[0136] In some examples, one of the computing devices 2 can transmit instructions for medical intervention to another one of the computing devices 2 positioned by the patient 4 or a caregiver of the patient 4. For example, such instructions for medical intervention can include instructions to change a medication dosage, timing, or selection, instructions to schedule a clinician visit, or instructions to seek medical attention. In this way, the patient 4 can be empowered to take action as needed to address their medical condition, which can help improve clinical outcomes for the patient 4.

[0137] Figure 4is a functional block diagram that illustrates an example configuration of one or more of the medical devices 17 in accordance with one or more techniques disclosed herein. In the illustrated example, the medical device 17 includes processing circuitry 40, a storage device 50, and communication circuitry 42. Additionally, in some examples, the medical device 17 can include one or more electrodes 16, an antenna 48, sensing circuitry 52, switching circuitry 58, a sensor 62, and a power source 56. As previously noted, the medical device 17 can be one of the medical devices 6 Figure 1 illustrated and described with reference to Figure 1 In another example, the medical device 17 can comprise a non-implantable medical device, such as a wearable medical device, a medical workstation cart, etc.

[0138] In some examples, one of the medical devices 17 can be a medical device implanted within the patient 4, while another of the medical devices 17 can include the camera 32, and thus can perform one or more of the various techniques of the present disclosure. That is, in accordance with one or more of the various techniques of the present disclosure, one of the medical devices 17 can capture images of the patient’s 4 body (e.g., the patient’s 4 implant site) via the camera 32.

[0139] The processing circuitry 40 can include fixed function circuitry and / or programmable processing circuitry. The processing circuitry 40 can include any one or more of a microprocessor, a controller, a DSP, an ASIC, an FPGA, or equivalent discrete or analog logic circuitry. In some examples, the processing circuitry 40 can include multiple components, such as any combination of one or more microprocessors, one or more controllers, one or more DSPs, one or more ASICs, or one or more FPGAs, as well as other discrete or integrated logic circuitry. The functions attributed to the processing circuitry 40 herein can be embodied as software, firmware, hardware or any combination thereof.

[0140] In some examples, the processing circuitry 40 can include an AI engine 44 and / or an ML model 46. The AI engine 44 and the ML model 46 can be similar to those described with reference to Figure 2 In one example, the ML model 46 can include one or more DL models that are trained, for example, on various physiological parameter data such as ECG data.

[0141] In some examples, the processing circuitry 40 can include an AI engine 44 and / or an ML model 46. The AI engine 44 and the ML model 46 can be similar to those described with reference to Figure 4In the illustrated, non-limiting example, the medical device 17 includes a plurality of electrodes 16A-16N (collectively, "electrodes 16"). In some cases, the electrodes 16 can be referred to herein as a plurality of "electrodes 16," while in other cases, they can simply be referred to as "electrodes 16" as appropriate. The electrodes 16 can be disposed within one body layer of the patient 4, while at least one other electrode 16 can be disposed within another body layer of the patient 4. In some examples, the medical device 17 can sense electrical signals attendant to the depolarization and repolarization of the heart of the patient 4 through the electrodes 16.

[0142] In some examples, the electrodes 16 can be configured to be implanted external to the chest of the patient 4. In some examples, the housing of the medical device 17 can serve as an electrode in combination with electrodes positioned on a lead. In some examples, the medical device 17 can be configured to measure impedance changes within the interstitial fluid of the patient 4, ECG morphology changes, etc. For example, the medical device 17 can be configured to receive one or more signals indicative of subcutaneous tissue impedance. In some examples, the computing device 2 can utilize such information to determine abnormalities of the IMD 6, such as device migration abnormalities, the computing device 2 can use the information to determine a likelihood of an abnormality (e.g., an infection) at the implant site based on the obtained images of the implant site.

[0143] One or more of the electrodes 16 can be coupled to at least one lead. In some examples, the medical device 17 can employ the electrodes 16 to provide sensing and / or pacing functionality. The configuration of the electrodes 16 can be unipolar or bipolar. The sensing circuitry 52 can selectively couple to the electrodes 16 through switching circuitry 58, for example, to select the electrodes 16 and the polarity, referred to as the sensing vector, controlled by the processing circuitry 40 for sensing impedance and / or cardiac signals. The sensing circuitry 52 can sense signals from the electrodes 16, for example, to produce a cardiac EGM or subcutaneous ECG to facilitate monitoring of the post-implant status of the IMD 6. The sensing circuitry 52 can also monitor signals from sensors 54, for example, which can include one or more accelerometers, pressure sensors, temperature sensors, and / or optical sensors. In some examples, the sensing circuitry 52 can include one or more filters and amplifiers for filtering and amplifying signals received from the electrodes 16 and / or sensors 54. In an illustrative example, the computing device 2 can obtain temperature sensor data from the IMD 6 to determine a likelihood of device pocket infection based on the obtained images of the implant site of the IMD 6, as a temperature sensor elevation at the IMD 6 that occurs before a temperature elevation at other locations of the patient 4 can indicate a device pocket infection. In some examples, the computing device 2 can obtain thermal images of the implant site and / or areas proximate to the implant site to compare the thermal images to temperature data and determine the presence of an anomaly at the implant site, such as a temperature elevation at the IMD 6 that causes a temperature elevation at other external locations of the patient 4.

[0144] In some examples, the processing circuitry 40 can use the switching circuitry 58 to select which of the available electrodes to use to obtain various measurements, for example, through a data / address bus. The switching circuitry 58 can include an array of switches, a matrix of switches, multiplexers, an array of transistors, microelectromechanical switches, or any other type of switching device suitable to selectively couple the sensing circuitry 58 to selected electrodes. In some examples, the sensing circuitry 52 includes one or more sensing channels, each of which can include an amplifier. In response to signals from the processing circuitry 40, the switching circuitry 58 can couple the output from selected electrodes to one of the sensing channels.

[0145] In some examples, one or more channels of sensing circuitry 52 can include an R-wave amplifier that receives signals from electrodes 16. In some examples, the R-wave amplifier can take the form of an automatic gain control amplifier that provides an adjustable sensing threshold as a function of the measured R-wave amplitude. Additionally, in some examples, one or more channels of sensing circuitry 52 can include a P-wave amplifier that receives signals from electrodes 16. Sensing circuitry 52 can use the received signals for pacing and sensing in the heart of patient 4. In some examples, the P-wave amplifier can take the form of an automatic gain control amplifier that provides an adjustable sensing threshold as a function of the measured P-wave amplitude. Other amplifiers can also be used. In some examples, sensing circuitry 52 includes a channel that includes an amplifier having a relatively wider passband than the R-wave or P-wave amplifiers. Signals from selected sensing electrodes selected for coupling to this wideband amplifier can be provided to a multiplexer and then converted to a multi-bit digital signal by an analog-to-digital converter (ADC) for storage in storage device 50. Processing circuitry 40 can employ digital signal analysis techniques to characterize the digitized signals stored in storage device 50. In some examples, processing circuitry 40 can detect and classify cardiac arrhythmias from the digitized electrical signals. In some examples, computing device 2 can obtain physiological parameters (e.g., arrhythmia data) via a physiological parameter sub-session as part of the set of data items. Additionally, computing device 2 can obtain device performance parameters, such as amplifier performance, ADC performance, etc., via a device check sub-session as part of the set of data items including interrogatory data items. Computing device 2 can utilize such data items to elucidate and / or inform analysis of implant site images according to one or more of the various techniques of the present disclosure.

[0146] In some examples, medical device 17 can include a measurement circuit having an amplifier design configured to switch between a plurality of different measurement parameters in real-time and continuously. Additionally, medical device 17 can enable sensing circuitry 52 and / or switching circuitry 58 for short periods of time in order to conserve power. In one example, medical device 17 can use an amplifier circuit, such as a chopper-stabilized instrumentation amplifier, according to certain techniques described in U.S. Application No. 12 / 872,552, entitled "CHOPPER-STABILIZED INSTRUMENTATION AMPLIFIER FOR IMPEDANCE MEASUREMENT," to Denison et al., filed August 31, 2010.

[0147] In some examples, the medical devices 17 can operate as therapy delivery devices. In such examples, the medical devices 17 can include leads. The leads can extend to any location within or near the heart or chest of the patient 4. In illustrative and non-limiting examples, one of the medical devices 17 can include a single lead extending from the one of the medical devices 17 into the right atrium or the right ventricle, or two leads extending into the right atrium and the right ventricle, respectively.

[0148] In some examples, one of the medical devices 17 can be configured to include sensing circuitry, such as sensing circuitry 52, and one or more sensors, such as sensors 54. Additionally, in some examples, one of the computing devices 2 can be configured to also include sensing circuitry, such as sensing circuitry 52, and one or more sensors, such as sensors 54. In one example, one of the computing devices 2 and / or one of the medical devices 17 can include a heart rate sensor, a pulse sensor, a photoplethysmogram (PPG) sensor, a blood oxygen saturation (Sp02) sensor, or the like.

[0149] The sensing circuitry 52 can be implemented in one or more processors, such as in one or more processors of the processing circuitry 40 of the medical devices 17 or the processing circuitry 20 of the computing devices 2. In Figure 4 In the example of FIG. 1, the sensing circuitry 52 is shown in combination with the sensors 54. Similar to the processing circuitry 20, 98, 40, 64 and other circuitry described herein, the sensing circuitry 52 can be embodied as one or more hardware modules, software modules, firmware modules, or any combination thereof.

[0150] In some examples, at least one of the medical devices 17 can include a sensor device, such as an activity sensor, a heart rate sensor, a wearable device worn by the patient 4, a temperature sensor, a chemical sensor, an impedance sensor, etc. In some examples, the one or more other medical devices 17 can be external devices relative to the patient’s 4 body or relative to the medical devices 17 implanted in the patient’s 4 body. In any case, the medical devices 17 can interact with each other via the communication circuitry 42, and in some examples can interact in similar fashion with the computing device 2, the edge device 12, etc. In an illustrative and non-limiting example, the computing device 2 can obtain temperature sensor data obtained from a temperature sensor on an IMD 6 board, and can correlate such data with image data to predict potential abnormalities. In one example, the computing device 2 can correlate images with an increased body temperature of the patient 4 or the IMD 6 to determine, for example, the presence of an infection at the implant site. In another example, the computing device 2 can obtain chemical data from a chemical sensor, which can indicate lactic acid formation and pH changes, which are primary indicators of abnormalities such as infections at the IMD 6 implant site. The computing device 2 can correlate such data based on, for example, baseline values obtained prior to implanting the IMD 6. In one example, impedance monitoring can be useful in detecting changes in device migration, but a number of days, such as 10 days, can be useful to wait to allow impedance to stabilize after an implant event. In some examples, the computing device 2 can determine that impedance has stabilized to a baseline value before relying on impedance data to determine the presence of potential abnormalities.

[0151] In one example, the computing device 2 can utilize orientation information of the medical devices 17, such as indicated by accelerometer data, to adjust, for example, image processing parameters of the computing device 2. In one example, shadows can be formed based on the orientation of the medical devices 17 that affect the image processing techniques of the present disclosure, such that the image processing algorithms can adjust the processing techniques based on information about where the medical devices 17 are positioned within the patient 4. In this case, the computing device 2 can receive information such as temperature data and / or orientation data from the medical devices 2, and utilize such information during analysis of image data. That is, the computing device 2 can analyze image data received via the camera 32 based on information received from the medical devices 2 in order to accurately characterize potential abnormalities. In a non-limiting and illustrative example, the computing device 2 can analyze image data based on a reference set of images of the implant site uploaded immediately after implantation, where the reference set of images can no longer align with the position state of the medical devices 2 due to movement of the medical devices 2, and thus, the computing device 2 can adjust the image processing techniques in order to maintain a degree of accuracy while the computing device 2 or another device ultimately performs abnormality analysis.

[0152] In one example, the computing device 2 can determine an ECG change indicative of device migration and thus increase the likelihood of detecting a potential abnormality from the image data from the physiological parameters obtained through the second sub-session. In such cases, the computing device 2 can analyze the image data using a bias for detecting abnormalities, or can include in the post-implant report an increased likelihood (e.g., probability, confidence interval) of a potential abnormality based on the likelihood determined from the first and second sets of data items.

[0153] In another example, the computing device 2 can determine an ECG signal and utilize the ECG signal to map to an orientation of the IMD 6. In some examples, the computing device 2 can utilize ECG morphology data to map to an orientation of the IMD 6 by identifying deflections (e.g., PQRST data) in the ECG or by comparing the ECG morphology data to a population-wide or cohort database. When the ECG morphology indicates a deviation from a baseline ECG of the patient 4, then the computing device 2 can indicate a device migration abnormality. The device migration abnormality can indicate a potential infection abnormality. In such cases, when the computing device 2 does not determine whether there is a potential infection from the set of images, the computing device 2 can determine to bias towards identification of an abnormality based on the ECG morphology data. In another example, the computing device 2 can obtain lead impedance information from one or more of the medical devices 17 (e.g., the IMD 6). Similar to the above, the computing device 2 can utilize the IMD information (e.g., lead impedance information) as additional input for abnormality detection and / or prediction.

[0154] The communication circuitry 42 can include any suitable hardware, firmware, software, or any combination thereof for communicating with another device, such as the edge device 12, a network computing device (such as a server), other medical devices 17, or sensors and / or the computing device 2. Under the control of the processing circuitry 40, the communication circuitry 42 can receive downlink telemetry from the edge device 12 or another device, and send uplink telemetry to it, with the aid of an internal or external antenna, such as the antenna 48. In addition, the processing circuitry 40 can communicate with the network computing device via the network 10, such as the Medtronic Carelink® network. The antenna 48 and communication circuitry 42 can be configured to communicate via inductive coupling, electromagnetic coupling, NFC, RF communication, Bluetooth®, Bluetooth® Low Energy, Wi-Fi, cellular communication, or any combination thereof. The network 10 can be any suitable network, such as the Medtronic Carelink® network. The antenna 48 and communication circuitry 42 can be configured to communicate via inductive coupling, electromagnetic coupling, NFC, RF communication, Wi-Fi TMor other proprietary or non-proprietary wireless communication schemes to transmit and / or receive signals. In one example, processing circuitry 40 can provide data to be transmitted upstream through communication circuitry 42 to edge device 12, computing device 2, and / or other devices through network 10. In an illustrative example, computing device 2 can receive a signal (e.g., uplink data) from a particular medical device of medical devices 17 (e.g., IMD 6). In such an example, computing device 2 can determine information related to patient 4 and / or the particular medical device of medical devices 17 (e.g., IMD 6) from the signal. In one example, the information can include device information (e.g., IMD information) corresponding to the particular medical device of medical devices 17 or corresponding to a set of medical devices 17, where the set can include other paired medical devices 17 (e.g., wearable devices, etc.) in addition to one of IMD 6. In such an example, computing device 2 can initiate a device interrogation session in which computing device 2 receives a signal from the particular medical device of medical devices 17 (e.g., IMD 6) indicating device interrogation data (e.g., battery health, operational parameters, etc.).

[0155] In some examples, processing circuitry 40 can provide control signals using an address / data bus. In some examples, communication circuitry 42 can provide data to processing circuitry 40 via a multiplexer, where the data is received externally via antenna 48. In some examples, medical devices 17 can transmit data to another device using a wired connection, such as a universal serial bus (USB) connection, an Ethernet connection (e.g., LAN) over network 10, etc.

[0156] In some examples, processing circuitry 40 can send temperature data or other device interrogation data to edge device 12 through communication circuitry 42. For example, medical devices 17 can send internal temperature measurements to edge device 12, which are then analyzed by edge device 12. In such examples, edge device 12 performs the processing techniques described. Alternatively, medical devices 17 can perform the processing techniques and transmit abnormal results to edge device 12 for reporting purposes, such as for providing an alert to patient 4 or another user.

[0157] In some examples, the storage 50 includes computer-readable instructions that, when executed by the processing circuitry 40, cause the medical device 17, including the processing circuitry 40, to perform various functions attributed herein to the medical device 17 and the processing circuitry 40. The storage 50 can include any volatile, non-volatile, magnetic, optical, or electrical media. For example, the storage 50 can include ROM, RAM, NVRAM, DRAM, SRAM, magnetic discs, optical discs, flash memory, forms of EEPROM or EPROM, or any other digital media. As an example, the storage 50 can store programmed values of one or more operating parameters of the medical device 17 and / or data collected by the medical device 17 for transmission to another device using the communication circuitry 42. As an example, data stored by the storage 50 and transmitted by the communication circuitry 42 to one or more other devices can include electrocardiograms, cardiac EGMS (e.g., digitized EGMS), and / or impedance values.

[0158] Various components of the medical device 17 are coupled to a power source 56, which can include a rechargeable or non-rechargeable battery. Non-rechargeable batteries can be capable of holding a charge for years and rechargeable batteries can be charged from an electrical outlet or other external charging devices (e.g., inductive charging). In examples involving a rechargeable battery, the rechargeable battery can be charged, for example, once a day, once a week, or once a year. In some examples, the power source 56 can be separate from the medical device 17 and implanted in a separate implant site of the patient 4 that can monitor abnormalities in accordance with one or more of the various techniques of the present disclosure.

[0159] As described herein, the medical device 17 can include the medical device 6 (e.g., the IMD 6). In such examples, the medical device 17 can have a geometry and size designed for ease of implantation and patient comfort. The volume of the examples of the medical device 17 described in the present disclosure can be 3 cubic centimeters (cm 3 ) or less, 1.5 cm 3 or less, or any volume therebetween. Additionally, the medical device 17 can include a proximal end and a distal end that are rounded to reduce discomfort and irritation to surrounding tissue after implantation under the skin of the patient 4. Example configurations of the medical device 17 are described in, for example, U.S. Patent Publication No. 2016 / 0310031. The computing device 2 can receive IMD information (e.g., IMD size, whether the ends are rounded, length of the elongated lead, etc., if any) including such configuration details of the medical device 17. As described herein, the computing device 2 can utilize the IMD information, such as the IMD configuration, to train the AI engine 28 and / or the ML model 30 to provide IMD-tailored abnormality assessment in accordance with one or more of the various techniques of the present disclosure.

[0160] Figure 5 is an example UI visualization of an example computing device 502 in accordance with one or more techniques of the present disclosure. The computing device 502 can be an example of one of the computing devices 2 described with reference to Figure 1 , 2 or 3. In accordance with one or more various techniques of the present disclosure, the computing device 502 can include one or more cameras 32. As described herein, in some examples, the one or more cameras 32 can be separate from the computing device 502. In Figure 5 exemplary and non-limiting examples of the computing device 502 and other examples showing the computing device 502, the computing device 502 can comprise a mobile handheld device (e.g., a tablet computer, a smartphone, etc.). While shown as a mobile handheld device, the techniques of the present disclosure are not so limited. It will be appreciated that UI visualization elements can be employed using various other devices and in various other environments, such as in a virtual reality, augmented reality, or mixed reality environment. That is, a user can use the camera 32 of an augmented reality headset in order to image an implant site of a patient 4, as well as navigate one or more of the UIs of the present disclosure.

[0161] In some examples, a UI visualization or interface as described herein can comprise a UI page or screen of a related UI (e.g., the UI 22). In some examples, a user can navigate laterally (e.g., forward, backward, etc.) through various UI pages or screens using navigation buttons (e.g., back, next, scroll buttons, etc.). In some cases, a UI visualization can comprise a virtual reality and / or augmented reality visualization, such as when the computing device 502 comprises a VR, AR, or MR headset. That is, a UI visualization can be presented as an immersive UI program that a user can interact with in a virtual reality space and / or an augmented reality space. Those skilled in the art will appreciate that, although shown as different pages of a UI, instead of UI pages on a screen of a handheld mobile device, the computing device 502 can similarly present VR UI elements that a user can navigate through in order to conduct an interactive session in accordance with, or at least not contrary to, one or more of the various techniques of the present disclosure.

[0162] In some examples, the computing device 502 can present an interface 504 (e.g., via the UI 22) that includes various login options. In one example, the interface 504 can include a login tile 506 or other login element (e.g., a camera icon 508). In accordance with one or more techniques of the present disclosure, the computing device 502 can receive a user input in order to invoke and / or initiate a virtual check-in interactive session as a result of a user input relating to the login tile 506.

[0163] In illustrative examples, interface 504 can include a camera icon that first initializes camera 32 for login purposes. Computing device 502 can receive an image from camera 32 and authenticate the user from the received image. In some examples, computing device 502 can perform facial recognition or implant site (e.g., wound) characteristic recognition. In some examples, patient 4 can tap computing device 502 to the implant site for NFC or RFID authentication. That is, computing device 502 can receive an NFC or RFID indication. Computing device 502 can use such indication to identify and / or authenticate a particular user. In any case, computing device 502 can load an interactive session from storage 24 by or based on a reading or scanning of a barcode (e.g., a one-dimensional barcode, a two-dimensional barcode, a quick response (QR) code, a matrix barcode, etc.). In some cases, the barcode can be included in a handout given to patient 4 at implantation or shortly after implantation surgery. TM

[0164] In some examples, computing device 502 can use one of medical devices 17 to authenticate a user. In such cases, medical devices 17 can include a wearable device capable of verifying a user (e.g., patient 4, an HCP, etc.). In some cases, the wearable device can include a wristband that includes a barcode, an RFID chip, etc. In such cases, the user can tap medical device 17 to computing device 502 or computing device 502 can otherwise scan medical device 17. In some cases, the tap can be a non-contact tap, such as an air tap that maintains a small air gap between the devices. In any case, in non-limiting examples, computing device 502 can receive authentication information from medical device 17 and authenticate the user, such as by allowing the user to proceed to the next interface of the interactive session.

[0165] In some examples, UI 22 can include a button 600 (e.g., a soft key, a hard key, etc.) that functions as a login element. That is, button 600 can include a multifunction button that provides various login options. In one example, button 600 can include a fingerprint scanner or other biometric scanner. Button 600 can be on the front, back, or any other portion of computing device 502.

[0166] In some examples, computing device 502 can not present interface 504, such as after the user first logs in. In such examples, the user can select “remember device,” “remember me,” and / or “I am the only user associated with this device.” Computing device 502 can receive the user input and forego the login interface for future login events accordingly. In some cases, the user can want to re-authenticate for each login, such as in cases where multiple IMD patients use the same computing device 502 for implant site or other IMD monitoring. ​

[0167] Figure 6 This is a UI visualization of a launch session interface 602 according to one or more technologies of this disclosure. In some cases, the launch session interface 602 may include a launch session icon 606 and / or a camera icon 608. The camera icon 608 may be similar to camera icon 508, except that once authenticated, the camera icon 608 can be used as a shortcut to launch a site examination sub-session or other sub-sessions of the virtual registration process (e.g., interactive reporting and / or registration super-sessions). In such examples, the computing device 502 may receive user input for the camera icon 608, and the computing device 502 may then initialize the imaging device (e.g., camera 32) and launch a site examination sub-session (e.g., wound examination sub-session), as referenced. Figures 15 to 19 As described.

[0168] In some cases, the initiation session interface 602 may include a configuration file interface 604. In some examples, the configuration file interface 604 may include an image of patient 4, images of one or more relevant implantation sites of patient 4, or both. This allows patient 4 to enjoy a more personalized experience by using the virtual registration UI. In such examples, computing device 502 may access the images from storage device 24 or from another storage device such as via network 10. In some examples, the images may be from a previous site examination sub-session.

[0169] In some examples, virtual checks in an interactive session can be pushed to patient 4's device, such as via a cloud solution. In this case, computing device 502 can receive push notifications from edge device 12 or from another device (e.g., server 94 via network 10). In some cases, push notifications may originate from another user's computing device 2, such as computing device 2 originating from HCP. In some cases, computing device 502 may receive notifications prompting a virtual registration session based on pre-determined scheduling reminders. In such cases, HCP may program a calendar timer to cause computing device 502 to provide push notifications and / or automatically initiate virtual registration sessions (e.g., interactive reporting sessions) upon expiration. In some examples, computing device 502 may provide the user with instructions to open (e.g., initiate) a virtual registration interactive session when a scheduling trigger or push notification (e.g., an HCP push) is determined. In some cases, computing device 502 may automatically initiate a virtual registration interactive session when a scheduling trigger or push notification is determined. That is, computing device 502 may automatically present... Figure 5 504 error on the interface Figure 6 The interface 602, or in some cases, can be presented by default and referenced. Figures 7-21The described interface is similar to the interface described above. In some cases, the computing device 502 can determine whether the user needs to authenticate first before presenting various other interfaces.

[0170] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can be configured to determine identification data of patient 4. In one example, processing circuitry 20 can determine identification data of patient 4. The identification data can include one or more of the following: biometric data, RFID data, NFC data, telemetry protocol data, user login information, authentication data, patient name or username data, etc. In any case, the user can input the identification data via computing device 2 and / or via camera 32. In one example, computing device 2 can include a computing device 502 that functions as a user check-in device. In this case, computing device 2 can determine the identification data of patient 4 through input received via elements of interface 504 (e.g., via button 600, where button 600 includes a biometric scanner).

[0171] In some examples, patient 4 can provide authentication data in order to gain access to perform an interactive session. In some examples, patient 4 can provide biometric data (e.g., facial data, fingerprint data, iris data, voice data, characteristic implant site data, etc.). In one example, computing device 2 can include a biometric scanner (e.g., camera 32, a fingerprint scanner, etc.) that patient 4 or another user can use to provide authentication / biometric data. In some cases, camera 32 can image an implant site to identify patient 4, where processing circuitry 20 includes an AI engine 28 and / or a ML model 30 trained to identify patient 4 to a certain degree of certainty based on one or more initial images of the implant site. Processing circuitry 20 can identify patient 4 from unique characteristics of the implant site of patient 4. In another example, processing circuitry 20 can authenticate the user based on wireless communication with IMD 6 of patient 4 (e.g., NFC data, telemetry protocol data, RFID data, etc.).

[0172] In some examples, patient 4 can or can not be the primary user of computing device 2 for purposes of participating in and / or navigating the interactive session. In illustrative examples, a user separate from patient 4 can be the user of computing device 2 for purposes of participating in and / or navigating the interactive session. That is, the user can coordinate with patient 4 while navigating the interactive session on behalf of patient 4. In illustrative examples, a user separate from patient 4 can log in to access the interactive session. In such cases, upon authenticating to access the interactive session, the user can then identify patient 4. In some examples, the user can identify the patient by taking a photo of the patient, entering the patient’s name, scanning a barcode of the patient, imaging a wound site, etc. In illustrative examples, the user can take a photo of patient 4’s face or a photo of the implant site. In any case, computing device 2 can perform facial detection or wound characteristic detection to determine patient 4.

[0173] Processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can deploy image processing tools (e.g., AI engine 28 and / or ML model 30) in order to perform the authentication process. In one example, processing circuitry 20 can train the image processing tools with patient data, implant site data, etc. In one example, the imaging processing tools can learn, for example, how an implant site heals over time (e.g., healing trends) when the implant site is scanned. Generally speaking, the characteristics of an implant site can change over time, and thus processing circuitry 20 can adjust the relevant portions of the identification algorithm accordingly over time. This is particularly useful in cases where the identification algorithm of processing circuitry 20 utilizes, for example, images of the implant site to identify and / or verify the user. In such cases, processing circuitry 20 can still be able to accurately identify patient 4 even if a period of time has passed between enrollment sessions.

[0174] In some cases, when attempting to identify the patient 4 via the camera 32, the processing circuitry 20 can detect a potential anomaly at this stage. In such cases, the computing device 2 can request additional identification data to correctly and / or accurately identify the patient 4. In some cases, the processing circuitry 20 can store the authentication data (e.g., two-step authentication data) to the storage 24 in order to simplify the authentication and identification process of the patient 4 in subsequent sessions. In another example, the processing circuitry 20 can query a database (e.g., storage 96 of the remote server 94) that maintains patient data (e.g., username, password, implant site characteristic data, implant site images, etc.) over the network 10. In such cases, the processing circuitry 20 can identify the patient 4 from the patient data upon receiving the search results from the database. In an illustrative example, the processing circuitry 20 can receive user input via the UI 22 indicating the patient name of the patient 4. The processing circuitry 20 can query the database and identify the patient 4 as a known patient to the system 100 (e.g., system 300) based on the results of the query.

[0175] The processing circuitry 20 can reference the patient data (e.g., patient identifier) such that the same computing device 2 (or same algorithm library) can be shared among multiple patients (e.g., in a clinic). As described herein, the computing device 2 can adjust the site check algorithm library based on the patient data in order to customize the process and UI visualizations to each respective patient. In another example, a generic site check algorithm can be deployed to accommodate all patients of a certain category (e.g., nursing home patients, patients of a particular nursing home, etc.). In this way, the site check algorithm can maintain and provide a certain level of consistency for various users of the interactive session, where these users can be part of a common category.

[0176] In some examples, when operating the interactive session concurrently, multiple computing devices 2 can capture images of the implant site of a single patient 4 and determine other data items (e.g., interrogation data). That is, when operating the imaging procedure, the computing device 2 of the patient 4 as well as the computing device 2 of the caregiver can perform the techniques of this operation. In such examples, the computing devices 2 can synchronize the data and / or analysis results in real-time, or in some cases, at a later point in time, such as by waiting until a wireless connection between the computing devices 2 is available or a network connection becomes available (e.g., connection via the network 10). In some cases, until the computing devices 2 determine to perform the data synchronization process, the computing devices 2 can store the data locally, such as to the storage 24, or in some cases, to the storage 62 of the edge device 12.

[0177] In some examples, upon receiving authentication data identifying a user of the imaging procedure, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can determine identification data of patient 4. In such examples, the user of the interactive session can include an HCP, a family member of patient 4, patient 4, etc. To illustrate, processing circuitry 20 can authenticate the user and authorize the user to access the interactive session. Subsequently, processing circuitry 20 can receive, via UI 22, input data, such as user input, that effectively identifies patient 4. In one example, the input data includes barcode scan data, selection of patient 4 from a drop-down menu (e.g., via a keyword search), manually entered patient information, etc. In any case, computing device 2 can then determine the identification data of patient 4 from the input, such as by determining a name or ID of patient 4.

[0178] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can determine IMD information corresponding to one of medical devices 17 (e.g., IMD 6) of patient 4. In one example, processing circuitry 20 can identify IMD information corresponding to a particular one of medical devices 17 based at least in part on identifying data. In some examples, the IMD information can include IMD implant site information, such as a location at which the implant site is positioned on the body of patient 4, a size and shape of the implant site, and other information about the implant site. In another example, the IMD information can include historical data related to the implant site of the IMD and / or historical data related to the IMD. In some examples, the historical data can include images of the implant site after implant surgery, wound characteristics, shape and size of the implant site (e.g., wound size), incision information, any complication history during surgery, implant date, etc. In some examples, the IMD information can also include IMD type information, IMD communication protocol information, an estimated orientation of medical device 17 (e.g., IMD 6), data about one or more HCPs (e.g., surgeons, clinicians) responsible for implanting (e.g., inserting) medical device 17 (e.g., IMD 6), information related to one or more methods employed by the one or more HCPs in sealing the implant site, images of the implant site over time, etc. Exemplary implant methods and instruments are described in, for example, U.S. Patent Publication No. 2014 / 0276928. In any case, computing device 2 can determine a presence of a potential abnormality (e.g., a likelihood of an actual abnormality) at the implant site based on the implant method and instruments used during the implant surgery. In one example, computing device 2 can compare image data of the implant site to reference image data from a library of reference images, where the reference images in the library can have been tagged with various details such as implant method and instrument (e.g., library metadata). Computing device 2 can determine an abnormality at the implant site at least in part by referencing a corresponding image from the library of reference images that includes similar implant method and instrument attributes to the implant method and instruments of the implant site that computing device 2 is imaging.

[0179] In non-limiting and illustrative examples, computing device 502 can initiate an interactive session (e.g., a virtual check-in) upon a prompt by patient 4, at a specific date scheduled after implantation, and / or at or near a clinic visit. In some examples, computing device 502 can expire an interactive session program for a user (e.g., patient 4) after a given time. In another example, the interactive session can not include an explicit expiration date (e.g., a so-called “evergreen” application).

[0180] In illustrative examples, the computing device 502 can identify a follow-up schedule for the patient when providing the interactive session. In one example, the computing device 502 can receive the follow-up schedule, for example, from one of the computing devices 2 of the HCPs, or can access the follow-up schedule from a database via the network 10. The follow-up schedule can define one or more time periods in which the computing device 502 is configured to prompt the user for the interactive session. In some examples, the one or more time periods can include at least one time period corresponding to a predetermined amount of time from the day of implantation of the IMD 6. That is, a first time period or at least one time period of the one or more time periods can be a predetermined time from the day of implantation of the IMD (e.g., days from implantation of the IMD 6).

[0181] In illustrative examples, the follow-up schedule can include a first time period in which the computing device 502 provides a prompt, for example, 15 days after implantation or removal of the medical device 17 (e.g., the IMD 6). In another example, the computing device 502 can include a variable time period that the AI engine 28 and / or the ML model 30 can determine for the patient 4, such that different patients can have different enrollment schedules based on a variety of different criteria. In such examples, the computing device 502 can provide the interactive session according to the follow-up schedule. In another example, the computing device 502 can provide a prompt to the mobile device user, for example, to use the computing device 502 to capture image data at a predetermined time after implantation that results in an implant-related wound. In some examples, the computing device 502 identifies the follow-up schedule by receiving a push notification from another device via the communication circuit 26 and determining a first time period of the follow-up schedule based on the push notification. In such examples, the push notification can include an HCP push received from one of the computing devices 2 of the HCPs, such as the HCP corresponding to the patient 4. In some examples, the computing device 502 can identify the follow-up schedule by receiving a physiological parameter indicative of an abnormality (e.g., an ECG abnormality) and determining a first time period of the follow-up schedule based on the physiological parameter abnormality. That is, the computing device 2 can determine a triggering event for identifying a follow-up schedule that includes the trigger based on an amount of time elapsed, a particular signal received from one of the medical devices 17 (e.g., the IMD 6), such as an activity level, an ECG, or based on a trigger received via the network 10 (e.g., a computing device 2 of the HCP).

[0182] Figure 7UI visualizations of the menu interface 702 in accordance with one or more techniques of the present disclosure. Upon initiation of an interactive session, the computing device 502 can output an interactive UI via the UI 22 in accordance with the virtual check-in interactive session. In some examples, the computing device 502 can provide a top-level interface including at least one first interface tile. The first interface tile can correspond to a first sub-session. In such cases, the first sub-session can include a first sub-interface level at a lower level relative to the level of the top-level interface. The top-level UI interface can present a number of tiles from which a user can select. In Figure 7 In examples, the application of the present disclosure outputs a UI program containing one or more graphical UI visualizations elements (collectively referred to as UI elements or "tiles").

[0183] While described as being provided as part of a hierarchical structure having hierarchical levels, it should be understood that the techniques of the present disclosure are not so limited, and that the sub-session interfaces can be separate from the first interactive session interface in terms of what the user can consider to be the default when initializing the receiving session interface. In examples involving a mixed reality environment, the computing device 502 can present all of the interfaces at a single level, where the user can access each individual sub-session interface, for example, by shaking the head of the patient 4 around the virtual reality user interface. In illustrative and non-limiting examples, the user can access and / or initiate a sub-session from a virtual reality user interface, where the virtual reality interface can in some cases transition to an augmented reality interface, such as when the computing device 502 detects a selection of a first sub-session (e.g., a site check sub-session). In such cases, in accordance with one or more of the various techniques of the present disclosure, the computing device 502 can transition to an augmented reality mode in which the user can image the implanted site of the patient 4, for example, via a camera 32 (e.g., a head-mounted camera or another camera communicatively coupled to the computing device 502). Once the first sub-session is complete, the computing device 502 can revert to a separate user interface (e.g., a top-level menu interface or a second sub-session interface), which in some cases is again presented in a virtual reality environment. While a number of examples can be described herein with respect to user interaction with the interactive sub-sessions of the present disclosure, it will be understood that the interactive sub-sessions and sub-sessions can be provided in a variety of environments, which are not necessarily described herein for the sake of brevity.

[0184] As Figure 7As shown in the illustrative example of FIG. 7, the UI elements of interface 702 can include a patient status tile 706, a physiologic parameter analysis tile 708, a device check tile 710, and / or a site check tile 712. A user (e.g., patient 4, a caregiver, a physician) can select any of these tiles (e.g., by touch input) to utilize the functionality associated with the description of each such tile. In some examples, the UI elements can include interactive graphical elements that can be presented to the user via UI 22. In some examples, selecting a camera icon can cause the computing device 502 to automatically initiate a site check sub-session in order to provide a shortcut to a site check page that is accessible via the site check tile 712.

[0185] Additionally, any of the interfaces described herein can include a session ID tracker tile 704. The session ID tracker tile 704 can provide session and / or sub-session tracking information, such as date stamps, time stamps, session ID stamps and / or sub-session ID stamps, user information, etc. The session ID tracker can also include historical data regarding prior sessions (e.g., historical data regarding one or more prior sub-sessions), such as a summary of the results of prior reports (e.g., session reports, reports detailing the results of one or more sub-sessions, etc.). In any case, the computing device 502 can include such session tracking information when generating a new report for each particular interactive session, each particular sub-session (e.g., a sub-report of a sub-session), etc. The top-level interface 702 can include multiple session ID tracker tiles 704, such as one for each sub-session tile.

[0186] Additionally, the individual interface for one or more sub-sessions can contain tracker tiles that, when the computing device 502 detects selection of a particular tracker tile, the computing device 502 can retrieve historical information regarding each sub-session and prior sub-session results (e.g., reports, etc.). The computing device 502 can provide a pop-up interface that provides such information through the UI 22, or in some cases, can automatically navigate the user to the report and / or historical interface for further review through the UI 22. Additionally, the computing device 502 can export such data (e.g., reports, history, etc.) from the interactive session interface to another interface and / or to another device (e.g., one of the computing devices 2 of the HCP, one of the servers 94, an edge device 12, etc.). In one example, in response to detecting selection of one of the sub-session tracker tiles or the interactive session tracker tile 704 by the user, the computing device 502 can export a report in response to detecting selection of an export report button (e.g., soft key). The exported report can contain multiple reports (e.g., sub-session reports) or a compiled report (e.g., an interactive session report) detailing analysis of multiple data items from multiple sub-sessions. In an illustrative example, the computing device 502 can detect selection of a sub-session tracker tile through the sub-session interface and, in response, can generate and / or export a report corresponding to a particular sub-session of a particular interface of an interactive session.

[0187] In another example, the computing device 502 can detect selection of an interactive session tracker tile (e.g., tracker tile 704) and, in response, can generate and / or export a report corresponding to the interactive session, a report corresponding to a particular sub-session that has been performed so far in the case of not all sub-sessions, and / or a comprehensive report corresponding to the entire interactive session, as well as any summary report of the individual sub-sessions. In another example, the computing device 502 can generate a historical report that includes an aggregation of any one or more of these reports, such that in response to detecting selection of the interactive session tracker tile or other tracker tile, the computing device 502 can retrieve the historical report and compile and / or summarize past reports in order to produce a single post-implantation historical report for export and / or display (e.g., via a pop-up interface). In such cases, the computing device 2 of the HCP can receive the post-implantation report via the network 10, such that the HCP can review the historical record and / or a summary of the historical record of the patient 4 (e.g., for the medical device 17 corresponding to the patient 4).

[0188] Figure 8 is a flowchart illustrating an example method of implementing sub-session techniques of a virtual registration interactive session in accordance with one or more techniques of the present disclosure. The sub-session techniques of the present disclosure can be used to obtain various data items and make comprehensive determinations of abnormalities at a body site of a patient 4 based on combinations of the data items.

[0189] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can obtain image data related to a body of patient 4 and determine a current state of an implant-related wound on the body of patient 4. In such examples, the image data includes one or more frames representing images of the body of patient 4. Additionally, in some examples, a portion of the body of patient 4 includes an implant site of IMD 6, such that the frames represent images of the implant site. In an illustrative example, processing circuitry 20 can prompt a mobile device user (e.g., patient 4) to capture image data using a mobile device. In such examples, processing circuitry 20 can prompt the mobile device user at a predetermined time after implantation that results in an implant-related wound. As described herein, processing circuitry 20 can further obtain data related to a function of one of medical devices 17 implanted within the body of patient 4. Processing circuitry 20 can further determine a performance metric of medical device 17 based on the obtained data (e.g., interrogation data, diagnostic data). In some examples, processing circuitry 20 can determine a performance metric of a medical device based on captured image data, such as images indicative of device migration that in turn can affect various performance metrics of, for example, IMD 6.

[0190] In an illustrative example, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can monitor patient 4. In one example, processing circuitry 20 can provide an interactive session configured to allow a user to navigate a plurality of sub-sessions, the plurality of sub-sessions including at least a first sub-session and a second sub-session different from the first sub-session, where the first sub-session includes capturing image data by one or more cameras 32 (802). In some cases, processing circuitry 20 can provide the interactive session in response to being pushed to patient 4 by a cloud solution. In one example, the interactive session can include a mobile application pushed to patient by a cloud solution. In another example, processing circuitry 20 can provide the interactive session based on scanning a QR code from a flyer given at or shortly before / after implantation.

[0191] In some cases, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can provide an interactive session (e.g., a first sub-session, a site check sub-session, etc.) by training a session generator with one or more of cohort parameters or IMD information (e.g., via processing circuitry 20), deploying the session generator via the computing device to generate the interactive session, and providing the interactive session via the session generator that is at least partially personalized to the user as a result of the training. Additionally, the session generator can reference historical interrogation data for comparison when personalizing the interactive session. The session generator can include AI engine 28 and / or ML model 30.

[0192] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can determine a first set of data items according to a first sub-session of the interactive session (804). In one example, processing circuitry 20 can determine a first set of data items including image data as part of a site check sub-session.

[0193] In some examples, processing circuitry 20 can determine a second set of data items according to a second sub-session of the interactive session, the second set of data items being different from the first set of data items and including one or more of: data obtained from a medical device 17 (e.g., IMD 6), at least one physiological parameter of patient 4, or user input data (806). In some examples, processing circuitry 20 of computing device 2 or computing device 502 can perform a pairing session (e.g., a pairing process) with one or more medical devices 17 (e.g., IMD 6), with another computing device 2, and / or with an edge device 12 (e.g., an IoT device at home). The pairing session is configured to pair the computing device with the respective device (e.g., IMD 6). In some examples, the pairing session can include an authentication process as described herein. In an illustrative example, a mobile computing device can initialize to communicate with IMD 6, including authentication (e.g., 2-step authentication), such that only authorized mobile computing devices can interact with IMD 6. In such examples, in a first instance, IMD 6 can transmit a signal to computing device 2 in response to the computing device attempting to pair with IMD 6.

[0194] The user can then self-authenticate, and if appropriate, IMD 6 can then pair with computing device 2 and initiate bidirectional or unidirectional communication between the devices. IMD 6 can then remember the unique ID of computing device 2 for future authentication (e.g., through a hashed key, neural network, etc.). In any case, computing device 2 can receive information about IMD during the pairing session. In this case, computing device 2 can receive interrogation data about IMD 6 (e.g., from IMD 6) through the pairing process. In some examples, computing device 2 can store the interrogation data as historical interrogation data for subsequent reference and / or use as a training set for AI engine 28 and / or ML model 30. That is, in some cases, IMD information can incorporate device interrogation data (e.g., historical interrogation data).

[0195] According to one or more of the various techniques of the present disclosure, processing circuitry, e.g., processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can determine an abnormality corresponding to at least one of patient 4 and / or IMD 6 based at least in part on the first set of data items and the second set of data items (808). In one example, processing circuitry 20 can output a post-implant report of the interactive session, where the post-implant report includes an indication of the abnormality (810). Processing circuitry, e.g., processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can cause patient 4 to have the ability to generate a report (e.g., in PDF format or various other formats) for their records, send to a family member or physician via email or FTP, etc. In some examples, the post-implant report can also include an indication of an amount of time that has occurred since the day of implantation of IMD 6. In some examples, processing circuitry 20 can determine the post-implant report by determining an abnormality (e.g., an ECG abnormality) from the second set of data items and determining the post-implant report based at least in part on the abnormality and the image. In any case, processing circuitry 20 can determine a way to provide feedback to patient 4, such as through a graphical simplified icon or a complex result (e.g., depending on HCP preference). At the end of the interactive session, processing circuitry, e.g., processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, in some cases can stamp the results provided via the mobile device application on UI 22 with a date and time stamp.

[0196] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can output the post-implant report by outputting the post-implant report to a device of the HCP via network 10 (e.g., via a bidirectional communication protocol). In one example, processing circuitry 20 can output the post-implant report to UI 22 and / or another device of a user of the interactive session via network 10, where the user can or can not necessarily be patient 4.

[0197] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can determine that the interactive session is complete and provide a notification of the completion of the interactive session, such as by providing a confirmation receipt or by marking tiles of the UI 22 interface as having a threshold number of completed sub-sessions. In some examples, processing circuitry 20 can output results of the post-implant report for display (via a display of computing device 2 or another device (e.g., one of medical devices 17 that includes a display)). In an illustrative example, when processing circuitry 20 determines that any one or any combination of the sub-sessions produced an abnormal result (or an abnormality outside of an acceptable normal range), then processing circuitry 20 can output a prompt to patient 4 using the mobile device application. In some examples, the prompt can indicate the abnormality. In other examples, the prompt can contain a recommendation or instruction to schedule a follow-up by an HCP.

[0198] In some examples, the first set of data items can include image overlays (e.g., augmented reality overlays). The image overlays can be used to augment the set of preview frames for a particular angle and size reference. In such examples, processing circuitry 20 can output the image overlays at the first sub-interface level. Additionally, when providing the first sub-session, processing circuitry 20 can detect a selection of the first interface tile and provide the first sub-session of the interactive session. In another example, processing circuitry 20 can provide a prompt to a user (e.g., patient 4, an HCP, etc.) to capture image data with one or more cameras 32 when determining the first set of data items, and determine at least a portion of the first set of data items (e.g., the image data or at least a portion of the image data) after the prompt. In any case, processing circuitry 20 can prompt the mobile device user to capture image data using the mobile device at a predetermined time after implantation that results in an implant-related wound.

[0199] In some examples, processing circuitry 20 can provide a second sub-session of the interactive session through the second sub-session interface and can in turn modify the second interface tile to indicate (e.g., with a checkmark) completion of the second sub-session. Additionally, processing circuitry 20 can provide the first sub-session of the interactive session after the second sub-session to obtain image data and can in turn modify the first interface tile to indicate completion of the first sub-session. In some examples, to facilitate user navigation, processing circuitry 20 can facilitate navigation from the second sub-session interface directly to the first sub-session interface without having to involve the top-level interface. In any case, processing circuitry 20 can also provide a third sub-session of the interactive session after the first sub-session or the second sub-session and determine a third set of data items from the third sub-session of the interactive session, where the third set of data items is different from the first set of data items. In such examples, determining the third set of data items can include processing circuitry 20 receiving a physiological parameter signal and determining the third set of data items from the physiological parameter signal. It should be noted that the sub-sessions described in this disclosure can be provided in any order, accessed in any order, and in some cases, processing circuitry 20 can step one or more sub-sessions down over a particular time period, such as a predetermined time from implantation, a predetermined time from a user completing a particular interactive session, a result of a particular interactive session (e.g., such as an implant site status), and the like. In one example, processing circuitry 20 can disable access to an imaging sub-session after a predetermined time from implantation or after a predetermined time from a health status update (e.g., a wound is scheduled to heal according to an expected timeline). In such examples, processing circuitry 20 can generate the interactive session to include one of the remaining set of sub-sessions that are included with the interactive session at the expected time. In some examples, processing circuitry 20 can generate the interactive session to include a disabled or stepped-down sub-session (e.g., a site check sub-session) such that the disabled sub-session is hidden or otherwise inaccessible to the user. Processing circuitry 20 can provide the interactive session to include a sub-session even though some sub-sessions appear to be inaccessible to the user.

[0200] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of data server 94, or processing circuitry 40 of medical device 17, can identify an abnormality from various data items. In one example, processing circuitry 20 can determine that the abnormality includes a first abnormality. In such examples, processing circuitry 20 can determine a post-implant report by outputting, via communication circuitry 26, a second set of data items to another computing device in computing device 2. Processing circuitry 20 can in turn receive, via communication circuitry 26, a result of an analysis of any one or more of the sets of data items (e.g., the second set of data items). The result can indicate that the second set of data items does not indicate a presence of a second abnormality. In such examples, processing circuitry 20 can determine the post-implant report based at least in part on the first abnormality and the second set of data items.

[0201] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of data server 94, or processing circuitry 40 of medical device 17, can provide an interactive session including a third sub-session and a fourth sub-session. In one example, processing circuitry 20 can provide a third sub-session including a physiological parameter sub-session and can provide a fourth sub-session including a device check sub-session. In such cases, processing circuitry 20 can determine an abnormality by determining a third set of data items from the third sub-session, where the third set of data items includes at least one physiological parameter of the patient, and determining a fourth set of data items from the fourth sub-session, where the fourth set of data items includes interrogation data. As such, processing circuitry 20 can further determine the abnormality based at least in part on the third set of data items and / or the fourth set of data items.

[0202] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of data server 94, or processing circuitry 40 of medical device 17, can initiate a second sub-session after the first sub-session. In one example, processing circuitry 20 can determine the first set of data items by receiving the first set of data items via communication circuitry 26, and can determine the second set of data items by receiving the second set of data items via computing device 2. In some examples, processing circuitry 20 can determine the second set of data items including information indicative of an abnormality of at least one physiological parameter of the patient or an abnormality of a device parameter corresponding to one or more medical devices 17. In such examples, processing circuitry 20 can output at least one of the first set of data items or the second set of data items to an abnormality determiner for abnormality analysis upon determining an indication of an abnormality, and determine a result of the abnormality analysis, where the result is indicative of the abnormality. In this case, the abnormality determiner can include at least one of AI engine 28 and / or ML model 30. In one example, one of computing devices 2 can include and deploy the abnormality determiner to determine an abnormality at a body site of patient 4. That is, processing circuitry 20 can deploy the abnormality determiner to determine an abnormality, where the abnormality determiner can be trained to identify various abnormalities based on image data and other data items as described herein.

[0203] In some examples, processing circuitry 20 can determine the presence of a potential infection at the implant site as a potential abnormality. That is, processing circuitry 20 can determine the presence of a potential infection at the implant site upon identifying an abnormality at the implant site. In some examples, processing circuitry 20 can detect a potential abnormality at the implant site based on analysis of at least one frame. In another example, processing circuitry 20 can determine a potential abnormality by transmitting images and other data items to edge device 12 via communication circuitry 26. That is, computing device 2 can transmit one or more frames of image data to edge device 12, where the frames include one or more images of the implant site. Additionally, computing device 2 can transmit images and / or other data items to edge device 12, medical devices 17 (e.g., wearable devices, bedside workstations), and / or server 94 via network 10.

[0204] In some examples, the computing device 2 can transmit the image and / or other data items to another device, such as the server 94, via the network 10, in which case the server 94 can transmit the image and / or other data items to the edge device 12 for further analysis. That is, in some examples, the computing device 2 can indirectly transmit data, such as image or video data, to the edge device 12 and / or the server 94 via the network 10. In one example, the computing device 2 can transmit data to the edge device 12, which in turn performs processing of the data and / or transmits the data (e.g., edge-processed data) to the server 94 for further analysis. In such cases, the edge device 12 and / or the server 94 can determine the presence of a potential abnormality based on data (e.g., image data, video data, etc.) received from the computing device 2 via the communication circuit 26. In any case, the computing device 2 can determine the presence of a potential abnormality at the implant site upon receiving potential abnormality information from another device (e.g., the edge device 12, the server 94, etc.).

[0205] In illustrative examples, the edge device 12 and / or the server 94 can receive the image and / or other data items (e.g., the second set of data items, the third set of data items, the fourth set of data items, etc.) from the computing device 2. In some cases, the edge device 12 and / or the server 94 can perform image processing analysis prior to transmitting results of the image processing analysis to the computing device 2 and / or the edge device 12. In some examples, the edge device 12 can identify the presence of a potential abnormality based on analysis of the image data and / or other data items of the other sub-session. In some examples, the edge device 12 can deploy an image processing engine (e.g., by the AI engine 28 and / or the ML model 30) to determine the presence of a potential abnormality. In another example, the edge device 12 can perform some or all of the analysis with the aid of the server 94. That is, in some examples, the server 94 or the edge device 12 can include an image processing engine or various analysis tools configured to aid in detecting potential abnormalities. The server 94 and / or the edge device 12 can include an image processing engine, such as the AI engine 28 or the ML model 30 described with reference to Figure 2 In one example, the server 94 or the edge device 12 can include a training set for training one or more imaging processing engines. The server 94 and / or the edge device 12 can perform training of the image processing engine, or in some cases, can aid the computing device 2 in training the image processing engine. In another example, the server and / or the edge device 12 can transmit the training set to the computing device 2. In such cases, the computing device 2 can train the image processing algorithm (e.g., by the AI engine 28 and / or the ML model 30). In any case, the computing device 2 can determine the presence of a potential infection at the implant site from the image based on analysis of the image.

[0206] Additionally, when an anomaly is identified, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can determine the likelihood that a potential anomaly is an actual anomaly (e.g., severity measure, probability measure, etc.). In one example, processing circuitry system 20 can determine the likelihood that a potential anomaly is an actual anomaly based on information defining the potential anomaly (e.g., anomaly category, anomaly characteristics, IMD type, etc.). In some cases, processing circuitry system 20 can determine the likelihood by receiving data from another device (e.g., edge device 12, server 94, etc.) indicating the likelihood that a potential anomaly is an actual anomaly, or, in some cases, data indicating that a potential anomaly poses a health risk (whether severe or not). In a non-limiting example, processing circuitry 20 can determine the likelihood that a potential infection identified based on a set of images represents an actual infection.

[0207] As those skilled in the art will understand, an actual infection (e.g., a genuine infection) can be demonstrated by the presence of an infectious agent that is confirmed or verified at the implantation site. In other words, a potential infection determination is an unconfirmed determination that the implantation site has been infected by an infectious agent. That is, the computing device 2 can determine a potential infection as a possible infection based on data available to the computing device 2, but the computing device 2 may not be able to diagnose an actual infection without additional input indicating that the infection is actually an actual infection. In any case, the computing device 2 can determine the likelihood that a potential infection at the implantation site is an actual infection based at least in part on images and one or more other data items. In addition, actual abnormalities such as healing abnormalities can be demonstrated by actual abnormalities in the healing process, such as confirmation or verification by another source (e.g., HCP).

[0208] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, may use weighting factors (e.g., built-in bias, error range, etc.) when determining the likelihood that a potential anomaly is an actual anomaly. In one example, processing circuitry system 20 may train AI engine 28 and / or ML model 30 based on a data pool containing patient data, implantation site data, false alarm rate, error range information, etc., to determine the accurate likelihood that a potential anomaly is an actual anomaly. In one example, built-in bias may include a bias that tends to identify a potential anomaly as highly probable when the IMD belongs to a specific type or a specific number of images indicating a potential anomaly.

[0209] Figure 9 It is based on the user's Figure 7 The UI visualization of the patient status interface 902 is presented in response to a request for the patient status interface shown. In other words,Figure 9 It shows the user's... Figure 1 The screenshot shows a UI visualization of the selection of a patient status tile, invoking and launching the user-facing mobile device application UI of this disclosure. Various UI visualizations of this disclosure may include navigation buttons 906A-906N. Navigation buttons may include a back button 906A, a home button 906B, a next button 906C, an end button 906N, etc. In some cases, button 600 may function as a home button 906B, which allows the user to return to the home page of computing device 502 or the home page of an application, such as interface 602 or interface 702. Interface 902 may include a start session tile 904, which, when selected, can launch a patient status sub-session configured to trigger or request patient input. In some examples, the patient status interface 902 may serve as a health information hub for patient 4. In one example, in addition to health information received via computing device 502, computing device 502 may also receive health information from other devices, wearable devices, or other software applications.

[0210] Figure 10 This is a UI visualization of an example patient status interface 1002 based on one or more technologies disclosed herein. Figure 10 Non-limiting examples of patient status questionnaires that can be generated by the backend system or mobile device application of this disclosure are shown. The patient status questionnaire can be used to assess information about patient 4's overall health status, implant recovery of specific symptoms, medications recently taken or to be taken by the patient, etc. The patient status interface 1002 can be a UI page that allows patient 4 or another user to input information about patient 4, for example via free text, drop-down menus, radio buttons, etc. In some examples, computing device 2 can utilize patient health information to train AI engine 28 and / or ML model 30 to more accurately determine abnormal information at the implantation site (e.g., the severity of the abnormality type, etc.). In one example, computing device 2 can receive user input indicating soreness, redness, etc., at the implantation site. In this case, AI engine 28 and / or ML model 30 can utilize such information when identifying potential abnormalities. In some cases, computing device 502 can receive patient health information as audio data, in which case computing device 502 can train AI engine 28 and / or ML model 30 based on the audio data.

[0211] As described herein, in some examples, the second sub-session can include a patient status sub-session. In such examples, processing circuitry 20 can determine a set of data items via UI 22 in accordance with the patient status sub-session. In one example, processing circuitry 20 can receive user input data in accordance with the second sub-session of the interactive session and determine a second set of data items in accordance with the user input data. The user input data can include one or more of: patient entered data, medication information, symptom information, physiological metrics, or anatomical metrics. In an illustrative example, the user input data can include a patient’s level of pain, level of soreness, opinion of redness at or near the implant site, etc. Processing circuitry 20 can utilize such information in order to determine whether an abnormality exists at or near the implant site in accordance with the image data. In one example, processing circuitry 20 can receive an image in which a user points to a particular area of a body part of patient 4 and can indicate via user input that the indicated area is sore at a particular level of soreness. In some examples, processing circuitry 20 can obtain one or more additional images of the body part in accordance with the site check sub-session and perform image processing and analysis based on the user input and gesture indication (e.g., pointing) in order to determine with specificity and confidence that a particular area of the body part includes a potential abnormality at a particular confidence interval. In some cases, processing circuitry 20 can indicate a potential abnormality of a different area of the body part (e.g., above the implant site) even though the patient indicated that another area (e.g., below the implant site) is sore, where processing circuitry 20 identifies that a potential abnormality is more likely associated with a certain area (e.g., above the implant site) than another area (e.g., below the implant site) in accordance with the image data and / or other data items.

[0212] In some examples, processing circuitry 20 can input the user input data into a risk calculator (e.g., AI engine 28 and / or ML model 30) that is configured to control a frequency at which patient 4 receives follow-up notifications to image a particular body part of patient 4. In one example, when the symptom risk score is higher, processing circuitry 20 can prompt patient 4 to image a body part of patient 4 more frequently.

[0213] Figure 11 A UI visualization of an example physiological parameter check interface 1102 in accordance with one or more techniques of the present disclosure. The physiological parameter check interface 1102 can include buttons including start physiological parameter analysis 1104, previous parameter 1106, next parameter 1108, and physiological parameter menu. In some cases, the physiological parameter check interface 1102 shows an interface invoked or initialized by a user selection of a physiological parameter analysis tile (e.g., tile 708) as shown. Figure 7

[0214] Figure 12 ​A UI visualization of an example physiological parameter review interface 1202 in accordance with one or more techniques of the present disclosure. In the illustrated example, the physiological parameter review interface 1202 includes an ECG analysis page. The physiological parameter review interface 1202 illustrates two results as non-limiting examples, a normal ECG analysis 1204 and an abnormal ECG analysis 1208 (e.g., as indicated by an abnormality 1210 observed in the ECG). The ECG can be received from one of the medical devices 6, the medical device 17 (e.g., an IMD), or another device such as a watch, fitness tracker, or other wearable device configured to collect ECG data. In any case, the computing device 2 can obtain the ECG data for analysis. In the non-limiting and illustrated example, the analysis can include a normal result 1206 or an abnormal result 1212, in the case of an abnormal result an alert to contact a clinic can be required.

[0215] In some examples, the sub-session can include a physiological parameter sub-session. In such examples, the processing circuitry 20 can determine a set of data items in accordance with the physiological parameter sub-session. In one example, the processing circuitry 20 can receive at least one physiological parameter corresponding to the patient 4 in accordance with the second sub-session and determine the set of data items in accordance with the at least one physiological parameter. In some examples, the processing circuitry 20 can receive the at least one physiological parameter from one or more medical devices 17 including the IMD 6 via the communication circuitry 26. The at least one physiological parameter includes at least one of an electrocardiogram (ECG) parameter, a respiration parameter, an impedance parameter, a core body temperature, a skin surface temperature, an activity parameter, a blood pressure, a vital sign, a blood glucose level, a rhythm data, and / or a stress parameter. In some examples, the ECG parameter represents an abnormal ECG. In this case, the processing circuitry 20 can determine an abnormality at a body part of the patient 4 based at least in part on the abnormal ECG and in some cases in combination with the image data. The physiological parameter can include a health signal retrieved from the IMD 6 or from one or more other medical devices 17 such as another IMD or a wearable device including one or more sensors.

[0216] Examples of ECG collection include ECGs collected directly from one of the medical devices 17 (e.g., medical devices 6). In some examples, the medical devices 17 can include wearable devices, such as activity trackers, heart rate monitors, pulse monitors, pulse oximetry monitors, temperature monitors (e.g., core temperature monitors, surface temperature monitors). Additionally, the computing device 2 can receive ECGs collected from wearable devices or other ECG devices (e.g., medical devices 17). In some examples, the computing device 2 can provide a programmed connection (e.g., via a drop-down menu) with the medical devices 17. The computing device 2 can receive input from the patient 4 or HCP indicating a target programmed connection (e.g., wireless connection) to one or more of the medical devices 17 through the drop-down menu. In some examples, the virtual enrollment application of the present disclosure includes a device-aware application. That is, the computing device 2 can store information about what devices the computing device 2 is communicating with. In some cases, the computing device 2 can determine such information through a pairing process. In another example, the computing device 2 can receive information as downloaded by a physician. In some examples, the computing device 2 can determine such information from user selection in a drop-down menu (e.g., a device drop-down menu).

[0217] In some examples, the computing device 2 can include device-aware aspects based on interrogation of one of the medical devices 17. In another example, the computing device 2 can receive device parameters and information pushed through another device, such as through a push notification. In some examples, the computing device 2 can receive device information from another of the computing devices 2 operated by an HCP, where the HCP can fill in the correct information via the UI 22.

[0218] In some cases, the computing device 2 can receive physiological parameters or physiological parameter analysis results directly from another device or indirectly from another device, such as through the network 10. In an illustrative example, the computing device can receive an ECG or an image of an ECG from another device. That is, a scanner program can scan an ECG and upload the scan in any suitable document format. Likewise, a loader program can upload an image of an ECG for use by the computing device 2. In some cases, the computing device 2 can include a scanner program and / or a loader program. In such cases, the computing device 2 can upload physiological parameter information to another device or store the parameter information to an internal storage device (e.g., storage device 24).

[0219] In some examples, the circuitry, e.g., processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of data server 94, or processing circuitry 40 of medical device 17, can detect selection of the second interface tile prior to providing the second sub-session, where the first interface tile corresponding to the first sub-session and the second interface tile corresponding to the second sub-session are different from one another. In one example, processing circuitry 20 can provide the second sub-session of the interactive session in response to selection of the second interface tile.

[0220] In some examples, the top-level interface of computing device 502 can include the second interface tile. In another example, the interface of the first sub-session can include the second interface tile, such that a user can navigate from the first sub-session to the second sub-session without having to return to the top-level interface.

[0221] Figure 13 UI visualization of an example device check interface 1302 according to one or more techniques of the present disclosure. Device check interface 1302 shows an example interface that can be invoked or initialized according to user selection of a device check tile (e.g., tile 710) as shown. Figure 7

[0222] Figure 14 UI visualization of an example device check interface 1402 according to one or more techniques of the present disclosure. Invocation of device check interface 1402 can cause computing device 2 to interrogate medical device 17 (e.g., IMD 6). In one example, computing device 2 can interrogate medical device 17 through various short-range wireless communication protocols and telemetry. Computing device 2 can populate device check interface 1402 with various parameters of the device check section, as shown. In various examples, a backend system of a cloud-based implementation can push interrogation information to the mobile device, or the mobile device application can initiate interrogation locally at the mobile device. In any case, a user can review performance of medical device 17 (e.g., IMD 6) by invoking the application and sending the statistical data, as shown. Figure 14 Figure 14 In examples, the monitored statistical data includes, but is not limited to, battery strength, impedance, pulse width, percent pacing, pulse amplitude, and pacing mode. In some cases, computing device 2 can automatically schedule a routine device check or follow-up based on the results of the interrogation.

[0223] ​​In such examples, sub-sessions of interfaces 1302 and 1402 include device inspection sub-sessions, where processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can determine a second set of data items. In one example, processing circuitry system 20 can perform an inquiry to one or more medical devices among the medical devices 17 corresponding to the patient via communication circuitry system 26. The one or more medical devices 17 may include an IMD 6 and, in some cases, other medical devices 17, such as wearable monitoring devices. Processing circuitry system 20 can perform the inquiry by establishing a wireless communication connection with the corresponding medical device 17. That is, processing circuitry system 20 can receive the second set of data items via a computing network. In some examples, the computing device can receive medical device inquiry data directly from one or more medical devices 17, from one or more edge devices 12, or any combination thereof via network 10. In any case, processing circuitry system 20 can determine the second set of data items based on the inquiry, wherein the second set of data items contains device inquiry data (e.g., battery, impedance, pulse width, etc.). In some examples, processing circuitry 20 may determine a post-implantation report upon completion of the device inspection sub-session. That is, processing circuitry 20 may determine that the medical device 17 meets one or more performance thresholds based on the set of data items obtained via the device inspection sub-session. In one example, processing circuitry system 20 may compare query data with data read during implantation. In such examples, processing circuitry 20 may determine the post-implantation report based at least in part on data items from the device inspection sub-session. Examples of post-implantation reports and the generation of such reports are described herein (e.g., see references...). Figures 21 to 23 ).

[0224] Figure 15 This is a UI visualization of an example part inspection interface 1502 based on one or more techniques of this disclosure. Part inspection interface 1502 displays the user's... Figure 7 The selection of the shown part inspection block (e.g., block 712) invokes or initiates a part inspection sub-session.

[0225] The site inspection interface 1502 may include a Start Inspection 1504 button and a camera icon 608 button. The camera icon 608 automatically starts the site inspection sub-session. However, in some examples, the user may want to adjust camera parameters before starting. Although not shown, the site inspection interface 1502 may include additional options for adjusting various camera parameters. Additionally, the user can adjust camera parameters in real-time while capturing images of the implantation site using the camera. In some examples, camera parameters include adjustments between a front-facing camera and another camera, adjustments to illumination (e.g., infrared, thermal imaging, flash, etc.), zoom level, focus, contrast, etc. Once ready, the processing circuitry 20 can receive the user's selection to Start Inspection 1504 via the UI 22 to proceed to the next page (e.g., the interface) of the site inspection sub-session (e.g., the next graphical interface of the site inspection UI).

[0226] Figure 16 This is a flowchart illustrating an exemplary method utilizing image acquisition, recognition, and / or image processing techniques according to one or more technologies of this disclosure. For example, refer to... Figure 15 , Figure 17 and Figure 18 The interface is described, and the processing circuitry, such as the processing circuitry 20 of computing device 2, the processing circuitry 64 of edge device 12, the processing circuitry 98 of server 94, or the processing circuitry 40 of medical device 17, can receive UI commands to begin a site examination sub-session (e.g., a wound examination sub-session). In one example, the processing circuitry system 20 can receive an instruction from a user to select the start examination icon 1504.

[0227] In an exemplary example, a method for imaging the implantation site (e.g., an image capture process) is described. Camera 32 captures images (1612) of the implantation site or adjacent to one of the medical devices 17 (e.g., IMD 6). Processing circuitry system 20 can capture images at specific time points after implantation of the medical device 17 (e.g., IMD 6). In some examples, a first computing device in computing device 2 can store various images to storage device 24 (1614). Alternatively or additionally, computing device 2 can transmit various images to a secure backend system. Exemplary backend systems include edge device 12, server 94, and / or network 10 (e.g., Medtronic). Other components of the network. The backend system's processing circuitry (e.g., processing circuitry 64, processing circuitry 98, etc.) can process various images and / or route images to various other devices. In some examples, processing circuitry 20 can store images to storage device 50 of medical device 17. Similarly, processing circuitry 20 can store such data to storage device 62 of edge device 12 or storage device 96 of server 94.

[0228] The processing circuitry 20 can guide the user by providing predetermined reminders. The reminders can be pushed to the user’s computing device 2. Additionally, when the implant site shows no signs of healing over time, the HCP (e.g., the prescribing physician) can access an interactive session interface (e.g., a site check sub-session interface) to follow up with the patient 4. That is, the processing circuit 20 can determine an abnormality, such as an implant site that shows no signs of healing over time, and thus can transmit a notification to the computing device 2 of the HCP so that the HCP can access the interactive session interface (e.g., the site check sub-session interface) via the UI 22, including images, physiological parameters, etc.

[0229] In some examples, the processing circuit, such as the processing circuit 20 of the computing device 2, the processing circuit 64 of the edge device 12, the processing circuit 98 of the server 94, or the processing circuit 40 of the medical device 17, can deploy an ML model (e.g., a DL model) and / or an Al engine that analyzes each captured image from the site check sub-session for abnormalities as part of the site check sub-session. In an illustrative and non-limiting example, the site check sub-session can provide three levels of detection: “an abnormality exists,” “an abnormality does not exist,” “it is uncertain.” In one example, in the event that the processing circuit 20 detects an abnormality with a high likelihood in any of the images via the site check sub-session, then the processing circuit 20 can determine that an abnormality exists at that time. In the event that the processing circuitry 20 determines that no abnormality with a high likelihood exists in all of the images, then the processing circuitry 20 can determine that no abnormality exists at that time. In some examples, the processing circuit 20 can transmit the images from the site check sub-session to another computing device in the computing device 2 (e.g., the computing device 2 of a technician or HCP) for review by a human expert. The “expert” can be a trained professional that can review the images of the patient’s 4 body and identify some potential abnormalities and / or can capture images of the patient’s implant site with relative clarity. If the human expert detects an abnormality, then the patient can be required to have a follow-up with the prescribing physician.

[0230] In some examples, the processing circuit, such as the processing circuit 20 of the computing device 2, the processing circuit 64 of the edge device 12, the processing circuit 98 of the server 94, or the processing circuit 40 of the medical device 17, can determine a threshold for the presence or absence of an abnormality. In some examples, the processing circuit, such as the processing circuit 20 of the computing device 2, the processing circuit 64 of the edge device 12, the processing circuit 98 of the server 94, or the processing circuit 40 of the medical device 17, can determine parameters for the site check sub-session, including the threshold for abnormality detection, based on whether the implant site is a first-time implant or an already replaced implant, the IMD type (e.g., CIED type), the skin tone, and the age of the patient 4 being monitored, etc.

[0231] In an illustrative example, processing circuit 20 can determine a sensitivity level for a site check sub-session (e.g., an abnormality detection algorithm, etc.). To determine the sensitivity level, processing circuit 20 can determine medical device information (e.g., IMD information), physiological parameters (e.g., ECG), etc. about medical device 17. In such examples, processing circuit 20 can determine a sensitivity level based on the determined medical device information, physiological parameters, etc. In an illustrative and non-limiting example, processing circuit 20 can determine a prevalence factor that defines a prevalence of a particular type of IMD defined abnormality (e.g., infection). In another example, processing circuitry 20 can determine an impact factor that defines an impact that an abnormality (e.g., device failure, infection, etc.) can have on patient 4 and medical device 17 (e.g., IMD 6). Processing circuitry 20 can determine a sensitivity level that results in a more conservative algorithm that errs on the side of over-detection of abnormalities rather than under-detection of abnormalities from such information or other information. That is, it can be important to be more conservative and not miss a potential abnormality (e.g., infection, failure of IMD, etc.) because doing so can provide a life-saving therapy of high energy. In one example, IMD 6 can fail to the extent that temperature or migration data of IMD 6 cannot be determined such that when a potential abnormality is detected from an image, processing circuitry 20 can determine not to rely on data received from IMD 6 to identify the likelihood that an abnormality is an actual abnormality according to an adjusted sensitivity level.

[0232] In another example, when processing circuitry 20 determines that, for example, a stimulation generator (not shown) and a lead (not shown) are from different manufacturers but are combined into a single IMD 6, processing circuitry 20 can determine a higher (e.g., more conservative) sensitivity level relative to other less conservative sensitivity levels. That is, where items from different manufacturers are combined into a single device, the likelihood of a potential abnormality can be higher and, therefore, processing circuitry 20 can adjust the sensitivity level to be at a higher sensitivity level in such cases. In some examples, processing circuit 20 can adjust the sensitivity level used to identify a particular abnormality based on a duration of time that IMD 6 has been implanted (e.g., built-in computational bias). This is because certain abnormalities (e.g., pocket infection) can be most common during, for example, the first year after implantation and, therefore, a higher sensitivity level can be used for a particular time range (e.g., first year) and / or a lower sensitivity level can be used for another time range (e.g., after the first year).

[0233] In another example, medical device 17 (e.g., IMD 6) can include LINQ TMICM. In this case, processing circuitry 20 can employ a different sensitivity level for LINQ TM The ICM has a detection algorithm with different sensitivity levels. That is, the detection algorithm can employ medical device information indicating that the medical device was implanted as part of a particular type of procedure (e.g., an outpatient procedure) with a particular sensitivity level. Additionally, processing circuitry 20 can determine from the IMD information that a schedule of virtual enrollments or other types of enrollments is scheduled at particular intervals (e.g., regular intervals, irregular intervals, frequent intervals, infrequent intervals, etc.). In another example, processing circuitry 20 can determine a frequency at which a virtual monitoring service receives and / or monitors medical device diagnostics, such as diagnostics from medical devices 17 (e.g., IMD 6, a wearable heart rate monitor, and / or an activity monitor, etc.).

[0234] In illustrative and non-limiting examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can employ a first sensitivity level in determining abnormalities for a first type of implant (e.g., a pacemaker implant) and, conversely, processing circuitry can employ a second sensitivity level in determining abnormalities for a second type of implant (e.g., an ICM). In one example, the first sensitivity level can be higher than the second sensitivity level because processing circuitry 20 can have determined that the implantation procedure for the second type of implant was performed outside of an office or as an outpatient; that there is one or more regularly scheduled patient follow-ups via system 100 and / or system 300; and that processing circuitry 20 is continuously or at least semi-continuously monitoring medical device diagnostic device data, such as a result of regular data transmissions from medical device 17 to computing device 2 and / or other devices of system 300 (e.g., edge device 12, etc.).

[0235] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can determine parameters of an imaging procedure, including thresholds for abnormality detection, based on whether the implant site is a first time implant or an implant that has been replaced, the IMD type (e.g., CIED type), the skin tone, and the age of the patient 4 being monitored, among others. In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can determine parameters of an imaging procedure of camera 32 based on information about various abnormality control procedures. In an illustrative example, the abnormality control procedure can contain HCPs (e.g., surgeons or implanting clinicians) using Medtronic, Inc.’s TYRXTM absorbable antibacterial envelope or other similar element. It will be appreciated that TYRX TM is a mesh envelope that houses an implantable cardiac device, implantable neurostimulator, or other IMD. Processing circuitry 20 can determine a bias factor based on the presence of such a control program. This is because TYRX TM is designed to stabilize the device after implantation while releasing the antibacterial agent, minocycline, and rifampin for at least seven days, and thus, has a lower likelihood of an abnormality occurring during that time compared to implants that do not include such a control program. In other words, patients with TYRX TM envelopes as part of the implant generally have a lower chance of infection than patients without TYRX TM envelopes. In any case, various abnormality detection algorithms of the present disclosure can be configured to be more sensitive and / or have a lower specificity (e.g., imaging program parameters) for non-TYRX TM patients compared to imaging program parameters used for TYRX TM patients.

[0236] In some examples, an AI engine and / or ML model, such as AI engine 28 of computing device 2, AI engine 44 of medical device 17, ML model 30 of computing device 2, or ML model 46 of medical device 17, can determine a sensitivity level based on training of the AI engine and / or ML model. In one example, to determine the sensitivity level, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can train the AI engine and / or ML model with a dataset including prevalence data (e.g., infection rates for a particular type of medical device 17), severity data, data regarding likelihood of an abnormality, potential or actual impact of a particular type of abnormality (e.g., device malfunction and infection abnormalities, etc.), IMD information (e.g., device manufacturing information, implantation surgery information), etc. In training with such data, the AI engine and / or ML model can determine a sensitivity level for each individual monitoring event (e.g., device environment) or each class of monitoring event corresponding to patient 4.

[0237] Advantages of applying such levels of sensitivity include allowing the monitoring system 100 to efficiently allocate processing, memory, and / or power resources by scaling the sensitivity level according to the needs of the individual. Additionally, the sensitivity level can govern the rate at which the computing device 2 receives data from other devices (e.g., transmission rate, etc.). In this way, the computing device 2 can receive more data, for example, in situations where there is a higher likelihood of an anomaly occurring, while receiving less data in other situations. This selective use of sensitivity levels also aids in bandwidth considerations, such as by limiting the amount of communication data that can otherwise consume significant bandwidth of the system 100 and / or system 300.

[0238] In some examples, the site check sub-session including the AI engine 28 and / or the ML model 30 can be trained with a plurality of images / videos that have been labeled as corresponding to whether an anomaly (e.g., whether infected). In one example, the video can provide evidence of the patient’s 4 gait. In such examples, the processing circuitry 20 can include a site check sub-session executed by the AI engine 28 and / or the ML model 30. The AI engine 28 and / or the ML model 30 can be trained with data that has been labeled based on an improvement in the implant site or a worsening of the implant site. In one example, the AI engine 28 and / or the ML model 30 can detect over time that the implant site “improved” or the implant site “did not improve” as the implant status.

[0239] In some examples, the processing circuitry 20 can acquire images of the implant site from a plurality of site check sub-sessions over time after implantation. The processing circuitry 20 can analyze the time aging of the implant site with different thresholds in chronological order via the site check sub-sessions to detect an improvement or no improvement in the implant site. The prescribing physician can use the system to follow up only on patients whose implant site shows no signs of healing. In some examples, the processing circuitry 20 can operate according to a delay mode. During the delay, the processing circuitry 20 can collect images, for example, in chronological order (e.g., on a daily schedule) and merge the pictures together within 2-12 weeks after implantation. This would allow a time series presentation and time series analysis of the images. In such examples, the processing circuitry 20 can train the AI engine 28 and / or the ML model 30 according to a rate of change of the implant site healing (e.g., infection spread or growth). In some examples, the AI engine 28 and / or the ML model 30 can instead analyze the progress of the implant site healing rather than assessing the implant site itself (e.g., tracking a record of the difference between images over a period of weeks and incremental between successive images that are overlaid). In some cases, the AI engine 28 and / or the ML model 30 can track the difference between still images to determine an increment of the implant site over time (e.g., an increment between healing of the implant site in the delay). The processing circuitry 20 can then determine an anomaly based on an analysis of the increment over time at the implant site.

[0240] In some examples, processing circuitry 20 can determine the presence of a potential abnormality based on the images (1616). In one example, the backend system can route the various images to a computing device 2 of an expert technician trained to identify abnormalities from images of implant sites. In another example, the backend system can route the various images to a second computing device of the computing device 2 having a particular configuration of AI engine 28 and / or ML model 46 trained to identify abnormalities from images of implant sites. In some examples, when an abnormality is identified, processing circuitry 20 can determine a likelihood of the abnormality (e.g., a severity of the abnormality) based on the images of the implant site (e.g., the captured images of the implant site). In some cases, processing circuitry 20 can deploy a probabilistic model to determine the likelihood of the abnormality. In one example, processing circuitry 20 can determine a potential severity of the potential abnormality based on the analysis of the captured images.

[0241] In some examples, processing circuitry 20 can output a summary of the implant status containing the potential abnormality information (1618). In one example, the backend system can transmit a report back to the first computing device of the computing device 2 indicating the results of the image analysis performed by the backend system. In an illustrative example, processing circuitry 20 can generate a summary report identifying the abnormality based on the likelihood of the abnormality, where the summary can contain information about the determined potential infection. In some examples, processing circuitry 20 can store all of the image data and the labels (e.g., expert labels) in a database and computing system to continually improve and deploy the automated abnormality detection system (e.g., AI engine 28 and / or ML model 30). In any case, processing circuitry 20 can output the summary report. In some examples, the summary report contains one or more frames of image data. In such examples, processing circuitry 20 can generate the summary report (e.g., post-implant report) including the one or more frames of image data, where the image data is received prior to the determination of the abnormality and used at least in part to determine the abnormality determination. In some cases, the report contains frames corresponding to one or more camera angles (e.g., left versus right) in which the potential abnormality was identified. In some examples, when an abnormality is detected, processing circuitry 20 can output one or more of the various images of the implant site from the sub-session for review by a human expert.

[0242] In some examples, computing device 2 can determine whether to configure the interactive session to include a walk-through mode in which the patient 4 does not need to come in for a consultation (e.g., for a post-implant infection consultation). In such examples, computing device 2 can provide the HCP with access to the various images (e.g., still images). Thus, in some examples, the HCP can decide whether the implant site is healing as expected or whether an in-person follow-up is needed.

[0243] In some examples, the HCP can have programmed the interactive session program and transmitted it to an application database (e.g., program data store). In another example, the HCP can have uploaded the interactive session program directly to the computing device 2. That is, the computing device 2 can access the interactive session program from the application database. In one example, the interactive session program can include an imaging program for conducting a site check sub-session. That is, the imaging program can be part of the interactive session program, while in some cases, the imaging program that executes the site check sub-session can be separate from the interactive session program. Additionally, some aspects of the site check sub-session program can be controlled by a separate programming application, such as a camera application native to the operating software of one of the computing devices 2 (e.g., an OEM camera application), while other aspects can be controlled by the interactive session programming application, such as augmented overlays, camera parameter control (e.g., zoom, contrast, focus, etc.), such that the interactive session software can work in tandem with other software applications (e.g., camera applications, augmented reality applications, etc.) or devices on which such software applications run (e.g., augmented reality headsets, etc.).

[0244] The HCP can configure the interactive session to operate in a pass-through mode in which the patient can forego the post-implant consultation entirely until execution of the interactive session results in identification of a potential abnormality, such as an abnormality that meets a pre-defined threshold. In such examples, in any case, the HCP has access to the images to determine whether an abnormality exists, regardless of whether the processing circuitry 20 determines the presence of a potential abnormality. In this way, the HCP (e.g., physician, nurse, etc.) can independently review the images independent of the automated image analysis of the site check sub-session in order to independently determine the status of the implant, including the implant site healing status.

[0245] Figure 17 is a UI visualization of an example site check interface 1702 in accordance with one or more techniques of the present disclosure. The site check interface 1702 includes a photo icon 608 configured to cause the camera 32 to capture an image of the implant site 1704.

[0246] In some cases, the user can point the camera at the implant site 1704, in which case the processing circuitry 20 can automatically perform abnormality detection regardless of whether a command to capture an image has been received. Additionally, when the processing circuitry 20 detects an abnormality, the processing circuitry 20 can cause the camera 32 to automatically capture an image of the implant site. Once the photo is captured, an implant status indicator 1706 appears that indicates the status of the implant, including the implant site 1704. In some cases, the implant status indicator 1706 can include a textual description of the implant status, such as “Healing Well” or “Healing Poorly.” In other cases, the implant status indicator 1706 can include a graphical representation of the implant status, such as a green checkmark or a red X, respectively. Figure 17In the example of FIG. 17, the implant site is indicated as “normal.” In some examples, processing circuitry 20 can perform such detection by comparing the first image (captured automatically or manually) to one or more baseline images (e.g., a set of first images captured during a predetermined time period measured from implantation or from when the first image was taken). In one example, processing circuitry 20 can apply a sliding window to a set of historical images taken over time to filter images (e.g., old images) from the one or more baseline images and compare the first image to the one or more filtered baseline images. Additionally, processing circuitry 20 can determine a projection from the one or more filtered baseline images and determine a reference (e.g., a reference change characteristic) at an expected time that can be compared to the first image to determine the presence of an abnormality at the image site.

[0247] In some examples, processing circuitry 20 can implement a region comparison zone (e.g., Dx) for differential diagnosis. Processing circuitry 20 can perform gradient analysis on implant site 1704 and, for example, a different skin-colored sternum region to determine differential diagnosis. Processing circuitry 20 can reference the differential diagnosis to determine whether a potential abnormality is present at implant site 1704. AI engine 28 and / or ML model 30 also use various measurements in the user-provided picture to determine whether the implant site is within a threshold of a “normal” state. In some cases, processing circuitry 20 can overlay a ruler or other augmentation or overlay on the image to help the user capture an image that is useful for the measurements. That is, processing circuitry 20 can provide an augmentation or another frame that the user can use to obtain the correct distance and / or perspective of the implant site. In an illustrative example, processing circuitry 20 can direct the user to image the implant site with camera 32 at a target distance from the implant site and, in some cases, a second target distance from the implant site. Processing circuitry 20 can further direct the user to image the implant site at a target angle relative to the implant site or relative to a reference plane of camera 32 (e.g., a starting position of computing device 2). In such examples, processing circuitry 20 can use an overlay, augmentation ruler, etc., as in an augmented reality implementation, in order to capture an image at a particular angle (e.g., a particular view of the implant site) with the implant site having a particular relative size in the image data frame, etc.

[0248] In some examples, processing circuitry 20 can train AI engine 28 and / or ML model 30 to discern relative measurements regarding color and markings. As such, processing circuitry 20 can compensate for different appearances of wounds on different skin tones, across different patient demographic groups, and so on. In some examples, the image processing AI can use intra-picture region comparisons to distinguish between the implant site and areas of patient 4’s skin that are unaffected by the implant (e.g., to derive relative or delta information). The image processing AI can also be trained to compensate for differences between systemic infections and infections caused by implant sites such as at the pocket or incision. In some examples, processing circuitry 20 can deploy AI engine 28 and / or ML model 30 to determine red, green, and blue (RGB) image details, and other color schemes from captured images. Additionally, AI engine 28 and / or ML model 30 determine shapes of implant site wound closures, whether images include shininess, and so on from captured images. Processing circuitry 20 can determine whether images of implant sites include potential abnormalities or whether implant sites otherwise meet predefined thresholds for a “normal” state based on various image processing techniques.

[0249] Figure 18 is a UI visualization of an example site check interface 1802 according to one or more techniques of the present disclosure. Figure 18 The example site check interface 1802 of illustrates an abnormal implant site result. When an abnormality is detected, processing circuitry 20 can provide a visual alert 1806. Processing circuitry can include descriptive information about the potential abnormality in visual alert 1806. In some examples, processing circuitry can include visual alert 1806, a medical intervention instruction for the user to perform some action (e.g., contact a clinic, etc.).

[0250] In some examples, processing circuitry 20 can provide augmented reality overlays 1804 (e.g., image overlays) on interface 1802. Processing circuitry 20 can do so in order to help users capture images of implant sites according to size references, at particular angles, and so on. Additionally, AI engine 28 and / or ML model 30 can provide adjustments to respective algorithms based on trends that post-operative images will also change as implant sites change (e.g., heal gradually, heal gradually less, etc.). In some examples, processing circuitry 20 can train AI engine 28 and / or ML model 30 to detect, in some examples, deviations from a “healthy” state. The healthy state can include characteristics of implant sites from previous site check sub-sessions that did not detect abnormalities.

[0251] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can identify an image overlay to augment a set of preview frames and direct a user to capture an image in a particular manner. In one example, processing circuitry 20 can determine patient data corresponding to a patient. Processing circuitry 20 in turn can determine image overlay data based at least in part on the patient data. As discussed herein, patient data includes various information about an IMD patient, including any one or more of: image data of an implant site (e.g., explant site), implant site characteristics, patient identification data (e.g., patient name, authentication data, login information, etc.), patient input data (e.g., entered via UI 22), or other information about IMD 6, including implant or explant date, IMD component details (e.g., lead, wiring routing, etc.). The image data can contain a current image or a previously taken image, such as an image taken shortly after implantation or prior to an implantation procedure. In such examples, the image overlay data can correspond to the IMD and / or other components of the IMD. As used herein, a preview frame generally refers to a frame of image data displayed to a user prior to and / or during image capture. The preview frame represents what the user observes on a display screen of, for example, computing device 2. In any case, the image overlay data defines an image overlay configured to augment the set of preview frames. In some cases, the image overlay can contain a wireframe similar to a subject’s body outline, such as patient 4 or a generic outline.

[0252] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can obtain an image of a patient’s body, where the image corresponds to the image overlay and represents one or more locations of the body where one or more components of the IMD (e.g., IMD 6, a lead of IMD 6) coincide. In one example, the image can represent a right or left pectoral region of the body, where IMD 6 includes an implant configured to be implanted in the pectoral region of the body. In another example, the image can represent a portion of a neck of patient 4, where the user’s task is to image the portion of the neck where a lead of IMD 6 coincides, such as a lead of IMD 6 that is routed to a brain region of patient 4. Processing circuitry 20 can determine an abnormality (e.g., infection, discharge, etc.) at the one or more locations of the body where one or more components of the IMD coincide from the image.

[0253] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can determine that the image overlay data includes a hollow outline of the body with an outline of the region of interest (e.g., the implant site) such that a user can determine that the image is properly aligned to capture the region of interest at a particular angle, size, orientation, etc. In one example, processing circuitry 20 can determine that the image overlay data includes a first portion that includes an outline of at least a portion of the body of patient 4, where the portion of the body corresponds to the implant site of the one or more IMD components, and a second portion of the image overlay data, where the second portion includes an internal overlay inside the outline and represents the region of interest (the implant site).

[0254] In illustrative and non-limiting examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can determine that the IMD coincides with a particular side of the pectoral region of the body based at least in part on the patient data. In one example, processing circuitry 20 can determine a first portion of the image overlay data to represent the particular side of the pectoral region of the body. In some examples, processing circuitry 20 can determine a second portion of the image overlay. In such examples, processing circuitry 20 can determine, based at least in part on the patient data, that the implant site 1704 includes a particular incision angle relative to an outline of at least a portion of the body of patient 4. That is, the image overlay data can be dynamic and based on the particular environment of the imaging of patient 4, such as based on the type of medical device 17, the location of medical device 17, etc. In any case, processing circuitry 20 can determine the second portion so as to represent the implant site including the particular incision angle.

[0255] Additionally, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can obtain one or more images of the body in accordance with the image overlay. In one example, processing circuitry 20 can obtain an image of the body by determining, from the set of preview frames, that the second portion of the image overlay coincides with the implant site of the one or more IMD components. In response to determining the overlay, processing circuitry 20 initiates automatic capture of an image of the body of patient 4. As used herein, “automatically” or “automatic” generally means without the need for user intervention or control.

[0256] In another example, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can determine image overlay data from an image overlay library. In one example, processing circuitry 20 can determine a static image overlay configured to provide a border for aligning an implant site within the static image overlay. In an illustrative example, the border can include a dotted box shape, similar to the overlay 1804 shown. Figure 18 In such examples, processing circuitry 20 can retrieve the static image overlay from the overlay library containing at least one image overlay. In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can determine a custom image overlay based at least in part on patient data, IMD information, and / or a template overlay (e.g., a static image overlay) as an image overlay customized for patient 4. In an illustrative example, processing circuitry 20 can obtain an image of a body by detecting movement of computing device 2 (e.g., through accelerometer data) and by keeping a particular orientation of the image overlay relative to a set of preview frames and one or more locations of the body of patient 4 compensated for the movement.

[0257] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can determine image overlay data by receiving a subset of preview frames via camera processor 38 and determining image overlay data based on the subset of preview frames. In such examples, AI engine 28 and / or ML model 30 can determine an initial estimate of what the scar / implant site looks like and what the patient’s body looks like (shoulder frame, body shape, etc.) based on patient data, IMD information, etc. Then, when camera 32 is pointed at the implant site, AI engine 28 and / or ML model 30 can fine-tune or morph the image overlay in real-time to better frame the wound site with reference to other characteristics of patient 4’s body. In this way, the user can be guided by the image overlay to precisely align the implant site within the overlay outline in a manner that the computing device 2 analyzes the image capture of the implant site according to one or more of the various techniques of the present disclosure and knows the angle of the wound site on the body, the size of the wound site, etc. Additionally, the image overlay data can be used to train the AI to determine image overlays for future imaging sessions. In another example, processing circuitry 20 can use a template overlay as an initial guess or can use a template that has already been modified and then can modify the template in real-time once actual image data is received by camera processor 38. In such examples, processing circuitry 20 can tailor the image overlay to fit or conform to the body shape / frame of patient 4 in addition to representing other characteristics that can be present in images of the implant site (e.g., clavicle for a chest implant, hairline for a DBS implant, etc.).

[0258] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of data server 94, or processing circuitry 40 of medical device 17, can determine image coverage data based on a determination of a particular camera configuration of camera 32. In such examples, processing circuitry 20 can determine whether a front-facing or rear-facing camera of a mobile computing device is being used to capture images of the body of patient 4, or whether camera 32 is a standalone camera unit that a user can stand near so that camera 32 can capture images of the body of patient 4. In illustrative examples, when the camera being used is a rear-facing camera of a mobile computing device rather than a front-facing camera of the mobile computing device, processing circuitry 20 can need a respective mirroring for image coverage. Additionally, camera configurations can include camera quantity (e.g., dual cameras, triple cameras) and lens type (wide angle, 360° lens, etc.), which will require different image coverage depending on the camera being used. In such examples, processing circuitry 20 can determine image coverage data based at least in part on the camera configuration. When the image coverage defined by the image coverage data overlaps with the implant site or other body part of interest of patient 4, processing circuitry 20 can automatically capture an image of the implant site. As used herein, “overlap” can not include a complete overlap, but can involve at least a substantial overlap so that a particular percentage of the body part overlaps with the image coverage. In one example, when 80-90% of the implant site is within the bounds of a particular portion of the image coverage, processing circuitry 20 can determine that there is an overlap so that the image coverage portion substantially overlaps with the body part.

[0259] In such examples, the image overlay data can be configured to direct image capture of the patient’s 4 body relative to a particular distance from the one or more computing devices to the body surface. In some examples, the processing circuitry 20 can determine the image overlay data by creating a wireframe (e.g., a custom wireframe, a template wireframe, etc.) via an overlay generator (e.g., the AI engine 28 and / or the ML model 30), where the overlay generator is trained to generate wireframes from patient data. A wireframe generally refers to an augmentation that includes a contour, such as a hollow outline, that is used to align objects in a scene captured within the wireframe. A user can perceive objects in the scene through the wireframe in the foreground and can also perceive the wireframe in the foreground, thereby increasing the image data frame as understood in the art. In some examples, obtaining an image of the patient’s 4 body can include receiving one or more frames of image data via the communication circuitry 26, where the one or more frames include an image of the patient’s 4 body. In another example, obtaining an image of the patient’s 4 body includes capturing video data of one or more locations of the patient’s body (e.g., with the aid of the image overlay). In such cases, the processing circuitry 20 can determine the abnormality includes identifying the abnormality from the video data and determining the post-implant report based on the identification of the abnormality.

[0260] As described herein, the processing circuitry 20 can use the image overlay (e.g., a box, a wireframe, etc.) to determine a first set of reference images only. The processing circuitry 20 can then align subsequent images to the reference images. In one example, the processing circuitry 20 can execute an implant site detection algorithm to detect an implant site in a subsequent image, isolate an image of the implant site from the rest of the image (e.g., define coordinates or an area of a box that includes the implant site), and adjust the image, including the detected implant site, to align with the alignment of the reference set of images. Additionally, the processing circuitry 20 can operate a detection algorithm (e.g., the AI engine 28 and / or the ML model 30) trained, for example, with the first set of reference images, to determine what a region of interest looks like based on what is captured within the image overlay. In some cases, it can be necessary (e.g., by a user or automatically) to adjust the image overlay of the reference set to capture the site of interest within the box, such as in a central region of the box, or to capture at least a threshold amount of the image in the region corresponding to the box of the image overlay.

[0261] Figure 19is a flowchart showing an example method of displaying images of a body according to one or more techniques of the present disclosure. In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of data server 94, or processing circuitry 40 of medical device 17, can create a personalized baseline for patient 4 and allow an AI engine 28 (e.g., inference engine) and / or ML model 30 to provide trend-based analysis. In one example, processing circuitry 20 can deploy an AI engine 28 (e.g., inference engine) and / or ML model 30 configured to evaluate images based on historical frames of image data to fit a trend line, rather than evaluate images based on a snapshot (e.g., a single still image) or even successive snapshots that evaluate against each other rather than against a baseline characteristic of patient 4. In any case, such evaluation can inform patient 4 and / or a HCP about how an implant site (e.g., explant site) is changing (e.g., developing) over time toward healing or an abnormality (e.g., rapidly worsening infection or slowly developing infection). In one example, processing circuitry system 20 can enable a pre-processing algorithm to automatically adjust scale and orientation of captured images so that a user reviewing image time-lapse obtains a consistent view while reviewing successive images over time. Additionally, processing circuitry 20 can enable a higher predictive confidence in image analysis due to the presence of contextual data available at the point of inference, such as an inference AI engine that reviews images of a particular site check sub-session in addition to other data items of a review interactive session to determine a comprehensive post-implant report.

[0262] In some examples, AI engine 28 and / or ML model 30 can be trained on a plurality of images that have been labeled as corresponding to whether an abnormality (e.g., whether an infection). In such examples, AI engine 28 and / or ML model 30 can be trained with data that has been labeled based on improvement of an implant site. In one example, AI engine 28 and / or ML model 30 can detect over time that an implant site is “improving” or an implant site is “not improving” as an implant status.

[0263] In some examples, processing circuitry system 20 can acquire images of an implant site from a plurality of interactive sessions over time. Processing circuitry system 20 can maintain a chronology of images to detect improvement or non-improvement of an implant site. In one example, processing circuitry 20 can collect images on a pre-determined schedule and align images over a period of time post-implant. In such examples, processing circuitry system 20 can train AI engine 28 and / or ML model 30 according to how much an implant site characteristic changes over time. In some examples, AI engine 28 and / or ML model 30 can determine a post-implant report based on changes between successive images that identify differences between images over time.

[0264] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of data server 94, or processing circuitry 40 of medical device 17, can identify a first set of images representative of a particular location of a body of patient 4 (1902). In one example, processing circuitry 20 can identify a first set of images representative of a particular location of a body in which at least one component of IMD 6 (e.g., IMD 6, lead, etc.) coincides. In some examples, the first set of images can include a single image, while in other examples, the first set of images can include multiple images.

[0265] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of data server 94, or processing circuitry 40 of medical device 17, can determine a projection of a change characteristic, such as a healing characteristic (1904). In one example, processing circuitry 20 can determine a trend line or model representative of a projection of a change characteristic over time. That is, the projection is generally representative of a change characteristic of a particular location of a body of patient 4 over time. The particular location of the body can include an implant site of IMD 6 or can include other body locations of patient 4 that coincide with other components of IMD 6. In some examples, processing circuitry 20 can determine a projection of a change characteristic based on images of other IMD recipients, such as recipients aligned with a particular cohort of patient 4.

[0266] In an illustrative example, processing circuitry 20 can determine a projection of a change characteristic from a first set of images. In such examples, processing circuitry 20 can determine a common alignment reference for aligning a first set of images in a time-lapse representation of the first set of images. In some examples, processing circuitry 20 can align multiple images from the first set of images according to the common alignment reference to determine the time-lapse representation. In some examples, processing circuitry 20 can determine a projection of a change characteristic from the multiple aligned images. In some examples, each image in the first set of images can be validated to determine whether the image should be included in the multiple images. In one example, processing circuitry 20 can validate a quality (e.g., blurriness) of an image, an orientation (e.g., portrait only), and can discard images that fail the validation test.

[0267] As described herein, in various examples, the particular location of the body can include an implant site of at least one component of the IMD 6. In such examples, the common alignment reference includes an alignment truth utilizing one or more of: a relative angle of the implant site, a relative size of the implant site, used to determine and implement the alignment truth, for example, to align images in a time-lapse configuration. In some examples, the common alignment reference can further include: an image illumination characteristic, a skin pigmentation characteristic, a relative orientation of the implant site, or a relative position of the implant site in respective frames of image data representing the first set of images.

[0268] In an illustrative example, the processing circuitry 20 can align the plurality of images from the first set of images. In one example, the processing circuitry 20 can provide an image overlay configured to augment a set of preview frames during image capture. The processing circuitry 20 can determine that an image of the particular location of the body from the set of preview frames overlaps the image overlay. The processing circuitry 20 can obtain a first image of the first set of images based at least in part on the overlap, where the first image represents the particular location of the body according to the common alignment reference. The processing circuitry 20 can align the image of the implant site with the common alignment reference through the plurality of images. In one example, the processing circuitry 20 can automatically capture at least one first image of the first set of images upon detecting that the implant site is aligned with the image overlay. In this way, the processing circuitry 20 can obtain a first image that satisfies an alignment criterion such that the at least one first image of the first set of images represents a particular view of the implant site defined by the image overlay. Subsequent images from the first set of images can then be aligned based on the alignment of the at least one first set of images in order to determine a time series of multiple images captured with common characteristics (e.g., scaling, angle, size, color, illumination, brightness, etc.) over time. In some cases, the first set of images can include only the first image captured according to the application of the image overlay to augment the set of preview frames.

[0269] In some examples, the processing circuitry 20 can determine a set of pre- implant images representing the particular location of the body prior to implantation. In one example, the set of pre-implant images can include images captured prior to an implantation procedure such that the pre-implant images represent a baseline of the patient 4 to which the patient 4 will return after healing, sometimes with unavoidable increases in surgical artifacts (e.g., scarring, etc.). In such examples, a physician or an assistant to the physician can capture the set of pre-implant images while placing a white balance card (e.g., representing a pure white color) next to the body of the patient 4. In an illustrative and non-limiting example, a nurse or image acquisition representative can capture one or more skin color baseline images of the implant site (e.g., the intended implant site) of the patient 4 through the camera 32 prior to implanting the IMD 6 while the image acquisition simultaneously or concurrently captures images of a color reference, such as a white balance reference card held within a frame.

[0270] The processing circuitry 20 can determine the set of pre-implant images with color references to allow the processing circuitry 20 to identify the color reality of the pigmentation and / or skin type of the patient 4. The processing circuitry 20 can store the set of pre-implant images to a storage device (e.g., cloud storage or the storage device 24), where the AI engine 28 and / or the ML model 30 can determine (e.g., via a cloud solution) the color reality of the patient 4. As described herein, the color reality can be used to determine baseline characteristics of the patient 4 and to project healing characteristics of the patient 4 over time based on post-implant captured images. In this way, given the diversity of skin tones and body types around the world, the processing circuitry 20 can obtain images of the implant site pre-implant and immediately post-implant in order to accurately identify abnormalities (e.g., excessive or abnormal bruising) at the implant site across various cohorts.

[0271] In some examples, the set of pre-implant images includes a single image or multiple images representing a particular location of the body of the patient 4 captured according to a particular lighting condition or from a particular vantage point relative to a set of potential vantage points representing various different views of the implant site. In an illustrative example, the processing circuitry 20 can determine baseline characteristics of the body of the patient 4 pre-implant from the set of pre-implant images. In some examples, the baseline characteristics can include pigmentation, skin type, etc.

[0272] In some examples, the AI engine 28 and / or the ML model 30 can automatically determine a micro-cohort of the patient 4 based on the set of pre-implant images, such as a micro-cohort based on skin pigmentation. In some examples, a user can manually self-identify a cohort that the user self-identifies via the UI 22, such as based on skin type or pigmentation. In another example, the processing circuitry 20 can utilize manually entered information and baseline characteristic data to automatically identify a micro-cohort of the patient 4 that uses the manually entered information as an assumed input to the AI engine 28 and / or the ML model 30 that is configured to automatically or at least semi-automatically identify micro-cohorts for further image processing and / or training purposes as described herein. In some examples, the AI engine 28 and / or the ML model 30 can automatically infer a cohort of the patient 4 that will yield the best results for analyzing images of the patient 4 according to pre-processing at the time the first image of the patient 4 is taken.

[0273] In an illustrative example, processing circuitry 20 can transmit the pre-implant images to a device (e.g., another computing device in computing device 2) configured to operate a skin color classification algorithm for pre-processing via communication circuitry 26. In another example, processing circuitry 20 can operate the skin color classification algorithm, in which case the pre-implant images can be stored only to storage device 24. The algorithm can be trained on thousands of skin photos and use a color scale, such as Von Luschan’s chromatic scale, Fitzpatrick scale, or a combination thereof. Processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can utilize the color scale to classify the skin type of patient 4 into various skin type classifications, such that ML model 28 and / or AI engine 30 can accurately identify abnormalities of the skin after implantation of one of medical devices 17 (e.g., IMD 6). The skin type classification can include classifying or organizing the skin type of patient 4 into subsets of skin types, such as various subsets corresponding to a particular color scale, such as Von Luschan’s scale and / or Fitzpatrick scale. In any case, processing circuitry 20 can automatically group patient 4, e.g., based on the skin type classification. That is, processing circuitry 20 can automatically determine a group of patient 4 based on the one or more pre-implant images. In another example, processing circuitry 20 can automatically determine a group of patient 4 based on the one or more post-implant images (e.g., the first set of images and / or the second set of images).

[0274] In such examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of data server 94, or processing circuitry 40 of medical device 17, can apply a particularly defined implant site analysis algorithm according to each group to yield the best results for a particular group, such as an automatically determined group, by accounting for differences in skin type. Additionally, during application load, processing circuitry 20 can provide patient 4 with an option to self-identify their ethnicity, age, weight, physiological sex, and body type through UI 22. If provided, processing circuitry 20 can use these inputs to fine-tune the results of AI engine 28 and / or ML model 30 as described herein.

[0275] Processing circuitry 20 can determine a projected change in characteristics based at least in part on the baseline characteristics. In such examples, the change in characteristics from the projection is configured to approach the baseline characteristics of the patient’s body over time. That is, processing circuitry 20 can determine a trajectory to return to the baseline of patient 4 from the post-implant images, such as a state of return to a fully healed scar at the implant site of IMD 6.

[0276] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of data server 94, or processing circuitry 40 of medical device 17, can identify a second set of images (1906). In one example, processing circuitry 20 can identify a second set of images that represent a particular location of the body at a progression interval in time following the first set of images. Similar to the first set of images, the second set of images can contain a single image, while in other examples, the second set of images can contain multiple images. Likewise, processing circuitry 20 can perform similar verifications on the images of the second set of images, and can discard images that fail the verification test. In one example, processing circuitry 20 can determine whether the images are captured according to a particular orientation (e.g., portrait, angle, etc.), and can prompt the user to capture the images according to the particular orientation through UI 22 when an image is identified that is captured in an incorrect orientation. In another example, processing circuitry 20 can identify objects through object recognition, and notify the user through UI 22 when an image represents an incorrect object, such as an object that does not appear to represent the relevant implant site.

[0277] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of data server 94, or processing circuitry 40 of medical device 17, can determine a second set of change characteristics (1908). In one example, processing circuitry 20 can determine the second set of change characteristics from the second set of images.

[0278] In an illustrative example, processing circuit 20 can determine that the first set of images includes at least one first post-implant image and at least one second post-implant image, where the images are captured at consecutive time intervals. The consecutive time intervals can be spaced apart in time by a first duration of time. In one example, the duration of time can include a relatively short period of time, such as an hour or a day. In such an example, the second set of images can include at least one third post-implant image captured at a registration time following the time at which the first set of images was captured.

[0279] In such examples, the second duration can be greater than the first duration, such as a week after the last capture of the first set of images. That is, the first set of images can be captured at frequent intervals immediately after implantation or at least after the removal of a bandage from the implant site, such as every hour or every other hour for the first day. The first set of images is captured at frequent intervals to determine how the characteristics of the images change over time (e.g., healing indicators). In such examples, the second set of images includes consecutive fourth images captured at a second enrollment time a third duration after the first enrollment time, such as two or three weeks after the first image of the baseline set is captured. In such examples, the second set of change characteristics includes a healing trend of the relative change characteristics between the consecutive images of the second set of images. That is, the second set of change characteristics can be based on a healing progression between at least two images of the second set of images, which can be compared to the first set of change characteristics determined from the first set of images and / or the pre-implant set of images. The processing circuitry 20 can further determine the second set of change characteristics based on the pre-implant set of images and / or the first set of images in order to track a trajectory from the first set of images to the second set of images, a trajectory from the second set of images to a pre-implant image-based baseline of the patient 4, and / or a combination of such trajectories.

[0280] In some examples, the processing circuitry, such as the processing circuitry 20 of the computing device 2, the processing circuitry 64 of the edge device 12, the processing circuitry 98 of the data server 94, or the processing circuitry 40 of the medical device 17, can compare the second set of change characteristics to the projection of change characteristics (1910). In one example, the processing circuitry 20 can compare the second set of change characteristics to the projection. In an illustrative example, to compare the second set of change characteristics to the projection, the processing circuitry 20 can determine an amount of difference from the second set of change characteristics and the projection. In some examples, the processing circuit 20 can determine the amount of difference from the second set of change characteristics and the projection by comparing the second set of change characteristics to a particular portion of the projection corresponding to a projection time corresponding to the second set of images (e.g., a time when the second set of images is identified, such as when the second set of images is acquired, captured, received, time-stamped, communicated from another device, and / or obtained), where the projection time is based on projecting at least in part from an interval or time point corresponding to the historical set of images (e.g., the first set of images, the pre-implant images, the first set of images and the pre-implant images, etc.).

[0281] In a non-limiting example, processing circuitry 20 can determine a projection that indicates, based on the set of historical images, that a particular change should occur at a particular time in the future (e.g., a projection time) relative to the implant site of patient 4. In some cases, processing circuitry 20 can prompt patient 4 to capture a second set of images (e.g., one image or multiple images) at the projection time and can compare the characteristics of the second set of images to the projection to determine the extent of healing of patient 4 tracked on the projection. In cases where the difference between the projection and the second set of characteristics exceeds a particular threshold, processing circuitry 20 can determine that further analysis needs to be performed or that additional images of the implant site should be captured to supplement or enhance the analysis. In another example, processing circuitry 20 can identify the second set of images and determine, from the projection, a particular portion of the projection that coincides (e.g., coincides in terms of time interval) with the second set of images, rather than prompting patient 4 for images. In any case, processing circuitry can compare the second set of images to the projection to determine the presence of an anomaly, as in cases where the comparison indicates a particular amount of deviation that exceeds a predetermined threshold.

[0282] In some cases, processing circuitry 20 can determine, from the set of historical images, a projection of how much the implant site of patient 4 should change over time at a particular time in the future (e.g., an expected time). In such cases, processing circuitry 20 can compare the second set of change characteristics to the projection to identify a potential anomaly, a likelihood of a potential anomaly, or whether further analysis is needed, such as analysis by a HCP, technician, or the like. In another example, processing circuitry 20 can determine, from the second set of images and in some cases from the first set of images, how much the implant site of patient 4 has changed or appears to have changed over time at a time corresponding to the second set of images. Processing circuitry 20 can determine the second set of change characteristics from the determination of how much or appears to have changed the implant site of patient 4 over time, and processing circuitry 20 can compare the second set of change characteristics to an expected projection, where the expected projection indicates how much processing circuitry 20 expects the implant site of patient 4 to change over time at a particular time in the projection that aligns or at least approximates to the time corresponding to the one or more images from the second set of images (e.g., the time at which the second set of images was obtained).

[0283] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of data server 94, or processing circuitry 40 of medical device 17, can output, through UI 22, an estimate of a time at which a particular location of the body is expected to change according to the projection one or more milestones. In one example, the milestone estimate can include an indication of a time at which redness should subside, a time at which soreness should subside below a predefined threshold relative to a pain threshold of patient 4, a time at which a likelihood of infection falls below a predefined threshold, a time at which healing of the implant site exceeds a predefined healing threshold, and the like. Additionally, processing circuitry 20 can determine the milestone estimate based on a comparison of the second set of change characteristics to the projection. In such examples, the comparison indicates an amount of deviation that processing circuitry 20 can utilize to determine an estimate informed by current images of patient 4 and past images of patient 4. Additionally, processing circuitry 20 can determine the estimate based on (e.g., informed by) images and / or healing trends of other IMD patients (e.g., patients of the same micro cohort as patient 4). In any case, the amount of deviation can indicate a degree to which healing of patient 4 deviates from the projection (e.g., a projection based on historical images of patient 4 and / or other IMD patients), as informed by the determination of the second set of change characteristics. As such, processing circuitry 20 can provide an informed estimate of a time at which patient 4 can be expected to reach various change milestones of the healing process of patient 4.

[0284] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of data server 94, or processing circuitry 40 of medical device 17, can identify a potential abnormality from the comparison (1912). In one example, processing circuitry 20 can identify a potential abnormality at the particular location of the body based at least in part on the comparison. In some examples, processing circuit, such as processing circuit 20 of computing device 2, processing circuit 64 of edge device 12, processing circuit 98 of data server 94, or processing circuit 40 of medical device 17, can identify the potential abnormality by determining an abnormality trend that indicates an expected progression toward the potential abnormality or toward a worsening abnormality relative to the potential abnormality. In any case, in response to determining the potential abnormality, processing circuit 20 can transmit the first set of images and the second set of images to one or more HCPs’ devices over network 10 (e.g., via communication circuit 26). In another example, processing circuit 20 can utilize the abnormality determination to determine a post-implant report, which in some examples includes a basis for physiological parameter data, device interrogation data, and the like.

[0285] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of data server 94, or processing circuitry 40 of medical device 17, can identify (e.g., through an inference engine) one or more abnormality control procedures corresponding to at least one component of the IMD (e.g., gauze, TYRX TM In one example, processing circuitry 20 can determine such information from patient input during a patient input sub-session. That is, a user (e.g., patient 4) can input medication information, gauze information, and the like, and processing circuitry 20 can reference the information through an AI inference engine to determine projection data based at least in part on one or more abnormality control procedures. In such examples, the projection of the healing timeline and the healing trend can differ as in the case of using TYRX TM control procedures or using a particular medication.

[0286] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of data server 94, or processing circuitry 40 of medical device 17, can output the results of the comparison to a storage device (e.g., storage device 24, storage device 65, storage device 96, and / or storage device 50). In one example, as part of outputting the second set of images to one or more HCPs, processing circuitry 20 can transmit abnormality data items configured to visually represent potential abnormalities identified in the second set of images. In an illustrative example, where processing circuitry 20 flags the results of the images as including a “potential infection,” processing circuitry 20 can generate an alert and provide an alert indication to one or more HCPs (e.g., through UI 22). The alert indication can include a summary of the results, a post-implant report, one or more images, and in some cases, highlight images to indicate characteristics of the potential abnormality. Additionally, the alert indication can include a time-lapse of images created according to one or more of the various techniques of the present disclosure. In any case, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of data server 94, or processing circuitry 40 of medical device 17, can maintain a chronology of the first set of images (e.g., through the respective storage devices).

[0287] In some examples, the projection techniques of the present disclosure allow processing circuitry 20 to individualize the algorithm and make the feedback most useful to the patient 4 and clinician. This is because images of the implant site can be captured at a high frequency immediately after implantation, thus allowing processing circuitry 20 to produce a time series (e.g., time-lapse) dataset. Processing circuitry 20 can then utilize the time series to create a personalized baseline for the patient and allow the model to produce trend analysis, as described herein. In another example, processing circuitry 20 can utilize a pre-processing algorithm to automatically adjust the scale and orientation of all photos so that when the HCP seems to review historical images over time, the physician viewing the time series gets a consistent view. In such examples, processing circuitry 20 can provide a predetermined reminder to prompt the user to capture particular images (e.g., with particular zoom levels, lighting, angles, etc.) at a first frequency to determine a first set of reference images (e.g., a first set of post-implant images), and can provide a predetermined reminder to prompt the user to capture particular images at a second frequency, where the second frequency can include a variable frequency, but generally can include a frequency that is less than the first frequency, to determine a second set of ongoing enrollment images (e.g., a second set of post-implant images). Based on the at least two sets of images, processing circuitry 20 can accurately determine the presence of an anomaly and / or determine a post-implant report that indicates an anomaly informed by other sets of data items (e.g., physiological parameters, etc.). It should be noted that the various techniques of the present disclosure (e.g., projection techniques, sub-session techniques, overlay techniques, etc.) are applicable to both “cloud” implementations (e.g., Medtronic CareLink® network) and “edge” implementations (e.g., with respect to mobile applications, tablet applications, IoT applications, etc.), as described herein.

[0288] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can synchronize data and results of image analysis later for ease of access and / or algorithm adjustment. In one example, processing circuitry 20 can coordinate with other processors of the processing system to synchronize data and results of image analysis between edge networks (e.g., computing device 2, edge device 12, medical device 17, etc.) and cloud networks (e.g., server 94).

[0289] Figure 20 ​is a flowchart of an exemplary method of a set of UI interfaces that illustrate a navigational virtual enrollment process (e.g., an interactive session) in accordance with one or more techniques of this disclosure. Processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can determine a data item comprising image data representative of a body of patient 4 in accordance with a site examination sub-session (2002). In another example, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can determine a data item comprising physiological parameters in accordance with a physiological parameter sub-session (2004). In an optional example, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can determine a data item comprising patient input in accordance with a patient status sub-session (2006). As described herein, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can determine a data item comprising device interrogation data in accordance with a device examination sub-session (2008). It should be understood that the sub-sessions can be performed in various different orders, and with additional sub-sessions or without one or more of the sub-sessions described herein. In any case, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can determine an abnormality in accordance with a combination and / or synthesis of the image data and one or more of the other data items described herein (2010).

[0290] Figure 21 is a UI visualization of an exemplary full enrollment interface 2102 in accordance with one or more techniques of this disclosure. Full enrollment interface 2102 illustrates a UI that processing circuitry 20 can cause to be displayed upon completion of one or more of the sub-sessions of the virtual enrollment session (e.g., the device examination sub-session, the site examination process, etc.). In some cases, processing circuitry 20 can output full enrollment interface 2102 upon submission of information to be forwarded to an HCP / clinician, etc. Figure 21 illustrates a UI that computing device 2 can generate and present to indicate completion of the interactive session corresponding to all of the four tiles shown. As described herein, in some examples, full enrollment interface 2102 can involve a fewer or greater number of tiles, with computing device 2 utilizing additional or fewer interfaces to obtain data regarding patient 4 and / or medical device 17. Full enrollment interface 2102 can also include a graphical icon 2104 indicating completion of, for example, the site examination sub-session. Figure 7 illustrates a UI that computing device 2 can generate and present to indicate completion of the interactive session corresponding to all of the four tiles shown. As described herein, in some examples, full enrollment interface 2102 can involve a fewer or greater number of tiles, with computing device 2 utilizing additional or fewer interfaces to obtain data regarding patient 4 and / or medical device 17. Full enrollment interface 2102 can also include a graphical icon 2104 indicating completion of, for example, the site examination sub-session.

[0291] In some examples, the images can be adjudicated by a remote dermatology service. In such cases, the remote dermatology service can be configured to generate a report (e.g., a summary report, a post-implant report, etc.) based on the images. Additionally, the computing device 2 can receive the report from the remote dermatology service and provide the report for review through the UI 22. In such examples, the computing device 2 can interface with the remote dermatology service through the network 10 and / or through a mediator, such as through one of the edge devices 12. As described herein, the edge device 2 can include one of the computing devices 2, where the computing device 2 is configured to interface with and / or operate the remote dermatology service to generate an outputted summary report based on various images.

[0292] In some cases, processing circuitry, such as the processing circuitry 20 of the computing device 2, the processing circuitry 64 of the edge device 12, the processing circuitry 98 of the server 94, or the processing circuitry 40 of the medical device 17, can generate a summary report that matches a proficiency level of a user. In one example, the processing circuitry 20 can identify a particular proficiency level of a user and generate a summary report for the user according to the proficiency level. In such cases, the proficiency level can represent a level of proficiency of the user in reviewing reports such as remote dermatology reports. The processing circuitry 20 can automatically determine such a proficiency level from user data and / or can receive user input indicating such a proficiency level, such as by requesting such information from the user via a questionnaire or other input mechanism.

[0293] In an illustrative example, the processing circuitry 20 can receive a single report data file. The processing circuitry 20 can then transform the data of the data file in order to generate a user-oriented summary report (e.g., a post-implant report). The processing circuitry 20 can transform the data in different ways, for example, where the processing circuitry 20 determines that the report is being presented to a “novice” user (e.g., an elderly patient who takes pictures autonomously) rather than to another user such as an “expert” user (e.g., a particular HCP, a nursing home staff member, etc.). The summary report can include an HCP adjudication of the results of the automatic image analysis of the images, physiological parameters, patient status updates, device interrogation data (e.g., device diagnostics), and / or interactive report sessions.

[0294] In some examples, the full enrollment interface 2102 can provide an option for scheduling an in-person clinic follow-up. This option can be presented by the processing circuitry 20 through the UI 22 when a potential abnormality is detected or in cases where the processing circuitry 20 is unable to rule out an abnormality from an evaluation of the camera images.

[0295] Additionally, the full enrollment interface 2102 can include a "generate report" icon 2106 and / or a "send report to clinic" icon 2108. The report can include, among other information, images captured by the processing circuitry 20. Additionally, the report can include a synthesis of data including failure criteria based on failure severity (e.g., criteria for finding abnormalities), number of failed tests, number of failed attempts, number of failed tiles, etc. Additionally, the processing circuitry 20 can generate the report based on physician preferences (e.g., report type, details of data weighting, etc.). Additionally, the processing circuitry system 20 can transmit the report based on physician preferences for notifications and / or notification frequency. The processing circuitry system 20 can further synthesize the data based on characteristics of the patient 4 and / or type of medical device 17 (e.g., IMD 6).

[0296] In an illustrative example, the processing circuitry system 20 can employ various abnormality scoring algorithms based on the severity of potential abnormalities detected in one or more images. The abnormality scoring algorithms can determine an abnormality score that maps to actionable responses (e.g., transmitting a report to an HCP, automatically scheduling a clinician visit, etc.). The processing circuitry 20 can determine a sensitivity of the abnormality scoring algorithm, for example, based on the severity of potential abnormalities detected in images of a body site of the patient 4 (e.g., an implant site), based on the IMD type, etc. The sensitivity can define a threshold of conservative threshold variation that can result in a particular output (e.g., a failure determination, transmitting a report to an HCP, etc.) from a response to a determination of any image indicating an abnormality or an analysis of one or more images against projections indicating an abnormality. In another example, the sensitivity can define a complex threshold that utilizes a weighted composite score that takes into account, among other things, failure severity across sub-sessions of any potential medical condition of the patient 4 and underlying algorithms that define each sub-session (whether failed or not) to determine whether a failure requiring a particular response was determined, for example, based on a mapping of abnormality scores to particular response outputs. Additionally, the sensitivity can further define a threshold that takes into account physician programmable limits of the medical device 17.

[0297] In some examples, the processing circuitry system 20 can provide feedback to the patient 4 through graphical simplified icons. In another example, the processing circuitry system 20 can obtain an indication from the HCP of a preference for complex results over simplified icons. In such cases, the processing circuitry 20 can provide feedback to the patient 4 in the form of complex results.

[0298] Figure 22UI visualization of an example full enrollment interface 2202 in accordance with one or more techniques of the present disclosure. In one example, the full enrollment interface 2202 indicates an HCP office or a receipt confirmation 2204 of the submission to the HCP office by an intermediary system. That is, the processing circuitry 20 can receive a confirmation from the intermediary system (e.g., the edge device 12), and in turn can provide the receipt confirmation 2204. The receipt confirmation 2204 can represent that the report was received at the office of the HCP of the patient 4. In another example, the receipt confirmation 2204 can represent a confirmation that the post-implant report has been successfully saved to a database (e.g., one of the servers 94, a plurality of servers 94 including a cloud storage database, a server 94 and edge device 12 including a cloud storage database, etc.). In such examples, the authorized HCP can access the report from the database via the computing device 2 of the HCP. In the illustrated example, the receipt confirmation can further include uploading a summary report to an EMR database. As described herein, the techniques of the present disclosure are applicable to both “cloud” implementations (e.g., Medtronic CareLink® Network) and / or “edge” implementations (e.g., with respect to mobile applications). Thus, the post-implant report can be uploaded and stored in any number of different locations, and can be accessed from these locations in any number of different ways.

[0299] Figure 23 is a flowchart showing an example method of determining instructions for a medical intervention with respect to an IMD patient in accordance with one or more techniques of the present disclosure. In some examples, processing circuitry, such as the processing circuitry 20 of the computing device 2, the processing circuitry 64 of the edge device 12, the processing circuitry 98 of the servers 94, or the processing circuitry 40 of the medical device 17, can determine a health condition of at least one medical device in the medical device 17 (e.g., the IMD 6) and / or the patient 4 from the post-implant report (2302). The health condition can include an indication that the processing circuitry 20 has determined an abnormality at the implant site of the IMD 6 and / or an abnormality of one or more medical devices 17 (e.g., the IMD 6) performance, an abnormality of a physiological parameter, and / or an abnormality of a patient status input. In some cases, the processing circuitry 20 can determine the health condition based on a synthesis (e.g., combination) of data items obtained via multiple sub-sessions of the interactive enrollment session (e.g., a site check sub-session and a physiological parameter sub-session). In some examples, the processing circuitry system 20 determines the health condition based on images received through the UI 22. Although described as generally being performed by the computing device 2, the example method of Figure 23 may be performed by, for example, any one or more of the edge device 12, the medical device 17, or the servers 94, such as by the processing circuitry of any one or more of these devices.

[0300] ​Processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can determine instructions for medical intervention based on the health condition of patient 4 (2304). For example, in instances where processing circuitry system 20 determines that an abnormality exists at the implant site of IMD 6, processing circuitry system 20 can determine instructions for medical intervention based on the abnormality. In another example, when processing circuitry 20 determines that an abnormality at the implant site of IMD 6 and an abnormality of a physiological parameter of patient 4 exist, processing circuitry 20 can determine instructions for medical intervention based on a post-implant report that processing circuitry 20 can generate based at least in part on the multiple abnormalities. In some examples, processing circuitry system 20 can determine different instructions for different levels of severity or classifications of abnormalities. For example, processing circuitry system 20 can determine a first set of instructions for one abnormality that processing circuitry system 20 determines can not have another abnormality as severe. In some examples, processing circuitry system 20 can not determine intervention instructions in instances where processing circuitry system 20 determines that the level of abnormality does not satisfy a predefined threshold. In some examples, processing circuitry 20 can provide an alert, such as a text or graphical based notification, a visual notification, or the like. In some examples, processing circuitry 20 can cause a sound alarm to sound or cause a haptic alarm, alerting patient 4 of the determined abnormality. In other examples, computing device 2 can provide a visible light indication, such as a red light for a high level of severity or a yellow light for a moderate level of severity. The alert can indicate a potential, possible, or predicted abnormal event (e.g., a potential infection).

[0301] In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can transmit instructions (2306) for a medical intervention to be displayed via a user interface, such as UI 22. In some examples, processing circuitry 20 can transmit the instructions to a device of an HCP (e.g., a caregiver), such as a pager of the HCP. In examples where processing circuitry 64 generates the instructions, processing circuitry 20 can transmit the instructions for the medical intervention to a user interface, such as UI 22. The instructions can include post-implantation reports and / or individual abnormality indications (e.g., ECG abnormalities, etc.). In some examples, edge device 12, medical device 17 (e.g., IMD 6), server 94, and / or computing device 2 can use the collected data to predict adverse health events (e.g., worsening infection) using a comprehensive diagnostic approach. That is, computing device 2 can use abnormality determinations (including likelihood or severity determinations) as evidence nodes for a probabilistic model deployed by, for example, AI engine 28 and / or ML model 30, in order to determine a probability score indicating a likelihood that an implant site of patient 4 is infected, a likelihood that the implant site can be infected within a pre-determined amount of time, a likelihood that one of medical devices 17 (e.g., IMD 6) can experience a functional abnormality (e.g., malfunction), etc. In some examples, processing circuitry, such as processing circuitry 20 of computing device 2, processing circuitry 64 of edge device 12, processing circuitry 98 of server 94, or processing circuitry 40 of medical device 17, can also include failure criteria (e.g., criteria for finding abnormalities) based on failure severity, a number of failed tests, a number of failed attempts, a number of failed tiles or sub-sessions, patient characteristics, a type of medical device 17 (e.g., IMD 6), physician preferences, etc., as evidence nodes for the probabilistic determination (e.g., abnormality prediction).

[0302] While described from the perspective of a user computing device 2 executing techniques of the present disclosure, it should be noted that the system of the present disclosure supports two-way communication between a user (e.g., patient 4) and an HCP. The two-way communication can operate using similar UIs at both ends of the communication line. In such examples, the HCP can access images, physiological parameter data items, medical device information, patient information, etc., uploaded by the user via the user’s computing device 2 via the HCP’s computing device 2. Additionally, the HCP can determine the presence or absence of abnormalities from the data upload via the HCP’s computing device 2 and transmit a summary report back to the user’s (e.g., patient 4’s) computing device 2, for example, including one or more indications of abnormalities.

[0303] Illustrative examples of the present disclosure include:

[0304] Example 1: A method of monitoring a patient having an IMD, the method comprising: providing, via a computing device, an interactive session, the interactive session configured to allow a user to navigate a plurality of sub-sessions, the plurality of sub-sessions including at least a first sub-session and a second sub-session different from the first sub-session, wherein the first sub-session includes capturing image data via one or more cameras; determining, by the computing device, a first set of data items in accordance with the first sub-session of the interactive session, the first set of data items including the image data; determining, via the computing device, a second set of data items in accordance with the second sub-session of the interactive session, the second set of data items different from the first set of data items and including one or more of: data obtained from the IMD, at least one physiological parameter of the patient, or user input data; determining, based at least in part on the first set of data items and the second set of data items, an abnormality corresponding to at least one of the patient or the IMD; and outputting, via the computing device, a post-implant report of the interactive session, wherein the post-implant report includes an indication of the abnormality and an indication of an amount of time that has occurred since a date of implanting the IMD.

[0305] Example 2: The method of Example 1, wherein the first sub-session includes determining a change projection of a body part over time from a set of images; and determining an abnormality based at least in part on a comparison of the image data to the change projection.

[0306] Example 3: The method of any of Examples 1 or 2, wherein providing the interactive session includes: training, via the computing device, a session generator with one or more of cohort parameters or IMD information; deploying, via the computing device, the session generator to generate the interactive session; and providing, via the session generator, the interactive session personalized at least in part for the user as a result of the training.

[0307] Example 4: The method of any of Examples 1-3, wherein the IMD information includes interrogation data, and wherein the method further comprises: performing, via the computing device, a pairing process configured to pair the computing device with the IMD; and receiving, via the pairing process, the interrogation data regarding the IMD; and storing, via the computing device, the interrogation data as historical interrogation data for reference.

[0308] Example 5: The method of any of examples 1-4, wherein providing the interactive session comprises: identifying a follow-up schedule for the patient, the follow-up schedule defined by one or more time periods in which the computing device is configured to prompt the user for the interactive session; and providing the interactive session in accordance with the follow-up schedule.

[0309] Example 6: The method of example 5, wherein identifying the follow-up schedule comprises: receiving a push notification via a communication circuit of the computing device; and determining a first time period of the follow-up schedule based on the push notification.

[0310] Example 7: The method of any of examples 5 or 6, the one or more time periods including at least one time period corresponding to a predetermined amount of time from a date of implantation of the IMD.

[0311] Example 8: The method of any of examples 1-7, wherein the image data comprises one or more frames representing an image of a body of the patient.

[0312] Example 9: The method of example 8, wherein the portion of the body of the patient comprises an implantation site of the IMD, and wherein the one or more frames represent an image of the implantation site.

[0313] Example 10: The method of any of examples 1-9, wherein providing the interactive session comprises: providing a top-l...

Claims

1. A method of imaging a body of a patient having an implantable medical device (IMD), the method comprising: identifying, via processing circuitry of a computing device, a first set of images representing a location of the body in which at least one component of the IMD coincides; determining, from the first set of images, a projection of a change characteristic of the location of the body over time; identifying, via the processing circuitry, a second set of images representing the location of the body at a progression interval in time subsequent to the first set of images; determining, from the second set of images, a second set of change characteristics; comparing, via the processing circuitry, the second set of change characteristics to the projection; and identifying, based at least in part on the comparison, a potential abnormality at the location of the body.

2. The method of claim 1, wherein determining the projection comprises: determining a common alignment reference for aligning the first set of images in a time-lapse representation of the first set of images; aligning, from the first set of images, a plurality of images according to the common alignment reference to determine the time-lapse representation; and determining, from the plurality of aligned images, the projection of the change characteristic.

3. The method of claim 2, wherein the location of the body comprises an implantation site of the at least one component of the IMD, and wherein the common alignment reference comprises one or more of: a relative angle of the implantation site, a relative size of the implantation site, an image illumination characteristic, a skin pigmentation characteristic, a relative orientation of the implantation site, or a relative position of the implantation site in respective frames of image data representing the first set of images.

4. The method of claim 2, wherein aligning the plurality of images comprises: providing, via the processing circuitry, an image overlay configured to enhance a set of preview frames of image data; determining, via the processing circuitry, that an image of the location of the body from the set of preview frames overlaps with the image overlay; obtaining, based at least in part on the overlap, a first image of the first set of images, wherein the first image represents the location of the body according to the common alignment reference; aligning, via the plurality of images, an image of an implantation site of the at least one component of the IMD to the common alignment reference.

5. The method of claim 1, wherein determining the projection comprises: determining, via the processing circuitry of the computing device, a set of pre-implant images representing the location of the body prior to implantation; determining, from the set of pre-implant images, a baseline characteristic of the body of the patient prior to the implantation; and determining, based at least in part on the baseline characteristic, the projection of the change characteristic.

6. The method of claim 5, wherein the change characteristic from the projection is configured to approximate the baseline characteristic of the body of the patient over time. ​ ​ ​ 7. The method of claim 5, wherein the set of pre-implantation images comprises a single image representative of the position of the body captured from a vantage point that is a function of an illumination condition or from a set of potential vantage points relative to various different views of an implantation site representative of the at least one component of the IMD.

8. The method of claim 1, wherein the first set of images comprises at least a first image and a second image captured at consecutive time intervals, wherein the consecutive time intervals are spaced apart in time by a first duration, and wherein the second set of images comprises a third image captured at a registration time that is a second duration after a first interval of the consecutive time intervals, wherein the second duration is greater than the first duration.

9. The method of claim 8, wherein the registration time comprises a first registration time, and wherein the second set of images comprises a consecutive fourth image captured at a second registration time that is a third duration after the first registration time, wherein the third duration is greater than the first duration, and wherein the second set of change characteristics comprises a healing trend in relative change characteristics between consecutive images of the second set of images.

10. The method of claim 1, further comprising: outputting, via a user interface of the computing device, an estimate of a time at which one or more milestone changes in the position of the body are expected according to the projection.

11. The method of claim 1, wherein determining the projection comprises: identifying, via an inference engine, one or more abnormal control programs corresponding to the at least one component of the IMD or the patient; and determining the projection based at least in part on the one or more abnormal control programs.

12. The method of any one or more of claims 1, wherein identifying the potential abnormality comprises: determining an abnormal trend indicative of an expected progression toward the potential abnormality or toward a worsening abnormality relative to the potential abnormality.

13. The method of claim 1, further comprising: in response to determining the potential abnormality, transmitting, by a computing network, the first set of images and the second set of images to a device of one or more healthcare professionals (HCPs) via a communication circuit of the computing device.

14. The method of claim 13, further comprising: outputting, via the processing circuit of the computing device, results of the comparison to a storage device of the computing device; and as part of transmitting the second set of images, transmitting an abnormality data item configured to visually represent the potential abnormality identified in the second set of images.

15. The method of claim 13, further comprising: maintaining, via a storage device of the computing device, a chronology of the first set of images.

16. A system for imaging a body of a patient having an implantable medical device (IMD), the system comprising: a memory configured to store one or more frames of image data, the one or more frames representative of at least a first set of images; and one or more processors in communication with the memory, the one or more processors configured to: identify a first set of images representing a location of the body in which at least one component of the IMD coincides; determine, from the first set of images, a projection of a change characteristic of the location of the body over time; identify a second set of images representing the location of the body at an advancement interval relative to the first set of images; determine a second set of change characteristics from the second set of images; compare the second set of change characteristics to the projection; and identify a potential abnormality at the location of the body based at least in part on the comparison.

17. The system of claim 16, wherein to determine the projection, the one or more processors are configured to: determine a common alignment reference for aligning the first set of images; align a plurality of images from the first set of images according to the common alignment reference; and determine the projection of the change characteristic from the plurality of aligned images.

18. The system of claim 16, wherein to identify the potential abnormality, the one or more processors are configured to: determine an abnormality trend indicating an expected progression toward the potential abnormality or toward a worsening abnormality.

19. The system of claim 16, wherein to determine the projection, the one or more processors are configured to: identify a set of pre-implant images representing the location of the body; determine, based at least in part on the set of pre-implant images, a baseline characteristic of the body of the patient; and determine the projection of the change characteristic based at least in part on the baseline characteristic.

20. A non-transitory computer-readable storage medium having instructions stored thereon that, when executed, cause one or more processors to at least: identify a first set of images representing a location of a body of a patient in which at least one component of an implantable medical device (IMD) coincides; determine, from the first set of images, a projection of a change characteristic of the location of the body over time; identify a second set of images representing the location of the body; determine a second set of change characteristics from the second set of images; compare the second set of change characteristics to the projection; and identify a potential abnormality at the location of the body based at least in part on the comparison. ​

Images