System for identifying the pathological condition of edge cases and manufacturing implants

Abstract

A system and method for identifying edge case conditions for examination in patient-specific orthopedic implant procedures are disclosed. The system can analyze patient data, such as implant data, pre-operative data, post-operative data, or manufactured implant data, to identify edge case conditions in patient data that may affect implant placement in the patient. Based on the type of edge case condition, the system sends a notification to a human, such as a healthcare provider, to review the patient data before implanting a patient-specific orthopedic implant in the patient.

Description

Technical Field

[0001] Cross - Reference to Related Applications This application claims priority to U.S. Provisional Patent Application No. 63 / 437,966, filed on January 9, 2023, which is hereby incorporated by reference in its entirety.

[0002] The present disclosure generally relates to medical devices, and more particularly to systems and methods for identifying edge - case medical conditions for examination, designing patient - specific implants, and manufacturing patient - specific items.

Background Art

[0003] Orthopedic implants are used to correct a number of different diseases in a variety of situations, including spinal surgery, hand surgery, shoulder and elbow surgery, total joint reconstruction (arthroplasty), skull reconstruction, pediatric orthopedics, foot and ankle surgery, musculoskeletal oncology, surgical sports medicine, and orthopedic trauma. Spinal surgery itself can encompass a variety of procedures and targets, such as one or more of the cervical, thoracic, lumbar, or sacral vertebrae, and can be performed to treat spinal deformities or degenerations and / or related back pain, leg pain, or other body pain. Common spinal deformities that can be treated using orthopedic implants include irregular spinal curvatures such as scoliosis, lordosis, or kyphosis (hyper - curvature or hypo - curvature), and irregular spinal displacements (e.g., spondylolisthesis). Other spinal diseases that can be treated using orthopedic implants include osteoarthritis, lumbar degenerative disc disease or cervical degenerative disc disease, lumbar spinal stenosis, and cervical spinal stenosis. However, conventional orthopedic implants such as intervertebral discs and fixation rods do not actively cooperate to provide real - time adjustment and correction after surgery.

[0004] In addition, numerous types of data related to patient treatment and surgical interventions are available. To determine a patient's treatment protocol, physicians often rely on a subset of patient data available through the patient's medical records and historical outcome data. However, the amount of patient and historical data can be limited, and the available data may not be correlated with or relevant to the specific patient being treated. Furthermore, while the ability to collect and process digital data is improving, the collection mechanisms tend to be limited to a single physiological characteristic and / or a single disease / condition. For example, conventional techniques in the field of orthopedics may be limited to a limited set of devices, making other patient data or pre-treatment data unavailable. Moreover, such data may not be used by conventional implantable devices, and these devices may not be able to communicate with each other to adjust their operation based on the current medical condition / state. Thus, conventional techniques are limited in collecting data to achieve beneficial treatment outcomes, as well as in generating and optimizing patient-specific treatments (e.g., surgical interventions and / or implant designs). [Overview of the Initiative]

[0005] The accompanying drawings illustrate various embodiments of the system, the method, and various other embodiments of the present disclosure. Any person skilled in the art will understand that the boundaries of elements shown in the drawings (e.g., boxes, groups of boxes, or other shapes) represent one example of a boundary. In some examples, one element may be designed as multiple elements, or multiple elements may be designed as one element. An element shown as an internal component of one element in some examples may be implemented as an external component in another example, and vice versa. Furthermore, elements may not be drawn to a constant scale. A non-limiting and non-exclusive description is provided with reference to the following drawings. The components in each drawing are not necessarily to a constant scale, and instead, the emphasis is on illustrating the principle. [Brief explanation of the drawing]

[0006] [Figure 1] This is a network connection diagram showing a system for providing patient-specific medical care according to one or more embodiments of this technology. [Figure 2] This figure shows a computing device suitable for use in connection with the system shown in Figure 1, according to one or more embodiments of the present technology. [Figure 3] This is a system diagram showing an example of a computing environment in which the disclosed system operates in some embodiments. [Figure 4] This block shows the components that may be used in systems employing the disclosed technologies in some implementations. [Figure 5] This flowchart illustrates a process for identifying the pathological condition of an edge case for examination, according to one or more embodiments of the present technology. [Figure 6A] This flowchart illustrates a process for recommending a human review of a patient-specific implant procedure using one or more embodiments of this technology. [Figure 6B] This figure shows an exemplary review plan on a GUI that details an edge case pathology plan in relation to the method described herein, according to an embodiment. [Figure 7A] This figure shows an exemplary patient dataset that may be used and / or generated in connection with the methods described herein, according to the embodiments. [Figure 7B] This figure shows an exemplary patient dataset that may be used and / or generated in connection with the methods described herein, according to the embodiments. [Figure 7C] This figure shows an exemplary patient dataset that may be used and / or generated in connection with the methods described herein, according to the embodiments. [Figure 7D] This figure shows an exemplary patient dataset that may be used and / or generated in connection with the methods described herein, according to the embodiments. [Figure 8A]This figure shows an exemplary virtual model of a patient's spine that may be used and / or generated in connection with the methods described herein, according to the embodiments. [Figure 8B] This figure shows an exemplary virtual model of a patient's spine that may be used and / or generated in connection with the methods described herein, according to the embodiments. [Figure 9A-1] This figure shows an exemplary virtual model of the patient's spine in its preoperative and modified anatomical configurations. More specifically, Figures 9A-1 and 9A-2 show the patient's preoperative anatomical configuration, and Figures 9B-1 and 9B-2 show the modified anatomical configuration. [Figure 9A-2] This figure shows an exemplary virtual model of the patient's spine in its preoperative and modified anatomical configurations. More specifically, Figures 9A-1 and 9A-2 show the patient's preoperative anatomical configuration, and Figures 9B-1 and 9B-2 show the modified anatomical configuration. [Figure 9B-1] This figure shows an exemplary virtual model of the patient's spine in its preoperative and modified anatomical configurations. More specifically, Figures 9A-1 and 9A-2 show the patient's preoperative anatomical configuration, and Figures 9B-1 and 9B-2 show the modified anatomical configuration. [Figure 9B-2] This figure shows an exemplary virtual model of the patient's spine in its preoperative and modified anatomical configurations. More specifically, Figures 9A-1 and 9A-2 show the patient's preoperative anatomical configuration, and Figures 9B-1 and 9B-2 show the modified anatomical configuration. [Figure 10] This figure shows an exemplary surgical plan relating to a patient-specific surgical procedure that may be used and / or generated in connection with the methods described herein, according to the embodiments. [Figure 11-1] This figure shows an exemplary surgical plan report detailing the surgical plan shown in Figure 10, which may be used and / or generated in connection with the methods described herein for surgical review, according to the embodiments. [Figure 11-2]This figure shows an exemplary surgical plan report detailing the surgical plan shown in Figure 10, which may be used and / or generated in connection with the methods described herein for surgical review, according to the embodiments. [Figure 12A] This figure shows an exemplary patient-specific implant that may be used and / or produced in connection with the methods described herein, according to the embodiments. [Figure 12B] This figure shows an exemplary patient-specific implant that may be used and / or produced in connection with the methods described herein, according to the embodiments. [Figure 13] This figure shows the spinal segments of a patient after multiple patient-specific implants have been implanted, according to the embodiment. [Figure 14] This figure shows a remote device for controlling the operation of a patient's spine and intervertebral implants, according to an embodiment. [Figure 15] This figure shows an exemplary correction plan that may be used and / or generated in connection with the systems and methods described herein, according to the embodiments. [Figure 16A] This figure shows a patient's spine in different configurations according to the embodiment. [Figure 16B] This figure shows a patient's spine in different configurations according to the embodiment. [Figure 16C] This figure shows a patient's spine in different configurations according to the embodiment. [Figure 16D] This figure shows a patient's spine in different configurations according to the embodiment. [Modes for carrying out the invention]

[0007] The present technology is directed to systems and methods for identifying edge case medical conditions for testing in patient-specific treatments. In the context of orthopedics, systems with improved computing capabilities (e.g., predictive analytics, machine learning, neural networks, artificial intelligence (AI)) can use large datasets to define improved or optimal implant designs for a particular patient. The present technology can be extended by identifying cases for detailed review. The present technology can, for example, identify cases for human review, compare results between automated and human reviews, confirm analysis results, and / or provide automated annotation / recommendations for review. This enables the present technology to be extended, for example, for effective intervention for patients with complex edge case medical conditions, unknown medical conditions, and the like.

[0008] This technology can characterize a patient's data (e.g., the entire data or a subset of the data) and compare it to data aggregated from a group of previous patients (e.g., parameters, metrics, pathology, treatments, outcomes). In some embodiments, the systems described herein use this aggregated data to formulate potential treatment solutions (e.g., surgical plans and / or implant designs for spinal and orthopedic procedures) and analyze the associated likelihood of success. These systems can further compare potential treatment solutions to determine the solution specific to the optimal patient expected to maximize the likelihood of successful outcomes. The systems described herein can use the aggregated data to formulate implant inspection protocols, implant quality criteria, and / or manufacturing plans or manufacturing parameters. This system can identify edge case medical conditions for additional review, including human review and / or automated review (e.g., review by machine learning of edge cases), and based on the additional review, can modify surgical plans, implant designs, and other techniques disclosed herein. This provides various levels of anatomical analysis and implant design for various medical conditions, increasing the likelihood of predicted outcomes.

[0009] For example, if a patient presents with a spinal deformity medical condition described by data including lumbar lordosis, Cobb angle, coronal parameters (e.g., coronal balance, overall coronal balance, coronal pelvic tilt, etc.), sagittal parameters (e.g., pelvic incidence, sacral slope, thoracic kyphosis, etc.), and / or pelvic parameters, an algorithm that uses these data points as inputs can be used to describe the optimal implant design to correct the subject's medical condition and improve the patient's outcome. Since additional data inputs are used to describe the medical condition (e.g., disc height, segment flexibility, bone quality, rotational displacement), the algorithm can further define the implant design optimal for that specific patient and the patient's medical condition using these additional inputs.

[0010] The system can generate implant data (e.g., design files, manufacturing instructions, etc.) for manufacturing implants. The system can manage access to implant data based on authentication levels. Authentication levels may include, but are not limited to, authenticating users (e.g., manufacturers, healthcare providers, etc.) based on geographical location, biometric data, blockchain access, tokens, or any other authentication method. Based on the determined authentication level of a user requesting access to implant data, the system may permit the user to access some or all of the implant data. The system may include one or more medical digital filing cabinets for storing implant data, patient information, electronic medical records, and / or additional patient-related information. In some embodiments, the system may share data (e.g., implant data, medical data, patient information, electronic medical records, and / or additional patient-related information) over a network without using digital filing cabinets. Digital filing cabinets can use blockchain and non-fungible token (NFT) technologies to control the collection and access of implant data.

[0011] In some embodiments, the technology can be expanded by identifying cases for detailed review. Conventional treatment planning software often designs treatments for patients with a qualifying set of conditions, thus limiting the number of patients that can be adequately treated using the software. Furthermore, conventional treatment planning software may design treatments that are unsuitable for edge-case conditions, such as patients who meet the qualifying set of conditions but have rare conditions that are neither detected nor considered by the software. This can lead to misdiagnosis and inadequate outcomes. In contrast, at least some embodiments of the technology can be expanded for the effective treatment of large populations with a wide range of conditions, including conditions not included in the machine learning training set, default parameters for designing treatments, etc. In some embodiments, the technology can automatically design treatments for patients who meet the criteria for automated surgical planning and can identify other “edge-case” patients for further analysis using human review. For example, the system can identify edge cases for human review, provide automated annotations / recommendations for human review, compare results between automated and human review, and / or review the results from the comparison and / or review. This will enable the technology to be expanded for effective interventions for patients with complex edge cases, unknown conditions, and other such conditions.

[0012] In some embodiments, the computer implementation method includes analyzing patient data (e.g., implant data, pre-operative data, post-operative data, manufactured implant data, etc.) to identify edge-case medical conditions in the patient data that may affect implant placement in the patient. For example, process 500 may scan pre-operative data (e.g., images such as MRI, CT, CAT, or X-rays, and / or implant data such as implant design files or manufactured implants) to identify edge-case medical conditions that may require human review. An edge case may refer to a medical condition in the patient data that has at least one extreme (e.g., minimum or maximum) parameter. The edge criteria for qualifying as an extreme parameter can be entered by the user, determined based on statistical analysis, or generated by a machine learning module. In some embodiments, the system may analyze one or more virtual models representing the patient's medical conditions to identify edge-case medical conditions. In some embodiments, the system may identify candidate edge cases based on prediction results simulated using the virtual models. The system can then analyze additional data (e.g., pre-operative images, implant designs, implant models, etc.) to determine whether the candidate edge cases qualify for edge-case analysis. Edge case analysis may include, but is not limited to, automated re-examination, human re-examination, surgical planning, and design changes.

[0013] The computer-aided approach can identify edge-case medical conditions based on one or more parameters, such as the patient's medical history (e.g., number / type of previous procedures), modifications of interbody fusion, anatomical abnormalities (e.g., hemivertebrae, grade 2+ spondylolisthesis), symptoms of osteotomy, unusual symptoms of implants, tumors in the patient (e.g., tumors near the implant site may be affected by implant placement), fractures, torn or partially torn tissue (e.g., ligaments, tendons, or muscles), implants adjacent to or near the patient's spine (e.g., fixation mass at the surgical level), transitional vertebrae (e.g., extra vertebrae), or any existing patient conditions that may affect, for example, implant placement in the patient, outcome, disease progression, or combinations thereof. The system can determine the parameters for identifying edge cases. For example, the system can identify parameters that may abnormally affect the patient's outcome at a threshold level, and then analyze the identified parameters to determine whether the patient has an edge-case medical condition.

[0014] The computer-aided system can determine thresholds for each type of parameter. For example, if a tumor is detected in patient data, the threshold could be the size of the tumor or the distance between the implant and the tumor. In the case of tissue damage (e.g., fracture, ligament injury, ligament rupture, intervertebral disc injury), the threshold can be tailored to the type of tissue damage. If a fracture is detected in patient data, the threshold could include, for example, a size threshold (e.g., fracture length), a fracture mode (e.g., type I, type II, etc.), the location of the fracture, and the load-bearing impact of the fracture. In the case of tissue rupture, the threshold could include, for example, the type of rupture (e.g., partial rupture, complete rupture, etc.) and the location of the rupture. Based on the type of edge case condition, the computer-aided system can send a notification for human review of the patient information (e.g., by a healthcare provider, implant designer, etc.). In some cases, in response to the identification of edge case condition parameters, the computer-aided system automatically sends a notification to a human for review of the patient information. For example, if a tumor is within a predetermined distance from the implant site, the physician is notified to verify that the implantation procedure will not cause the tumor to rupture or become inflamed during or after the implantation procedure. In some implementations, machine learning is used to identify edge-case conditions. Intraoperative and / or postoperative data may be used to further train the machine learning model. For example, machine learning may be retrained using edge-case condition data identified based on the results of a human review process.

[0015] The computer implementation method may further include receiving patient data for implant design, implant manufacturing, implant placement in the patient, etc. The received patient data may be analyzed to identify relevant training parameters. A reference dataset may be generated based on the relevant training parameters. A machine learning model may be trained at least in part on the reference dataset to examine the patient data to identify edge case conditions as well as patient data relevant to implant procedures specific to the patient. In some embodiments, the relevant training parameters may be classified spinal conditions, which are based on one or more default thresholds. Default thresholds may include, but are not limited to, threshold quality, lumbar lordosis specific to the threshold level, threshold Cobb angle, threshold pelvic intrinsic angle, and / or threshold intervertebral disc height. In some embodiments, the received patient data includes at least one comparison of the received patient data with patient data having identified and approved / unapproved edge case conditions for the patient. Embodiments of the present disclosure are described further in full below with reference to the accompanying drawings, in which the accompanying drawings illustrate examples of embodiments, with similar figures representing similar elements throughout multiple figures. However, the embodiments of the claims may be embodied in a variety of forms and should not be construed as being limited to the embodiments shown herein. The examples shown herein are, in particular, non-limiting examples and are merely examples among the possible examples.

[0016] The words “equipped,” “possessed,” “contained,” and “included,” as well as other forms of these words, are intended to be equivalent in meaning and are intended to be unrestricted in that one or more items following any one of these words are not intended to be an exhaustive list of such one or more items, nor are they intended to be limited to only the one or more items listed.

[0017] When used herein and in the appended claims, the singular forms "a," "an," and "the" include plural references unless otherwise explicitly indicated in the context.

[0018] While the disclosures herein primarily describe systems and methods for verifying the quality of patient-specific implants, the techniques may be similarly applicable to medical and device applications in other fields (e.g., other types of medical procedures). Furthermore, while many embodiments herein describe systems and methods relating to implanted devices, the techniques may be similarly applicable to other types of medical technologies and medical devices (e.g., non-implantable devices).

[0019] Figure 1 is a network connection diagram showing a computing system 100 for providing patient-specific medical care according to one or more embodiments of the present technology. As will be described in more detail herein, the system 100 is configured to collect, store, monitor, and / or update medical data. The system 100 may include one or more digital filing cabinets 180, which may include, but are not limited to, one or more electronic health records (EHRs), EMRs, patient information, digital wallets (e.g., tokens, credit cards, cryptocurrencies, payment information, etc.), and other medical data. The digital filing cabinets 180 can receive medical data and convert it into a digital format to improve the efficiency of finding and retrieving medical data. The digital filing cabinets 180 can receive medical data from patients, healthcare providers, health insurance agencies, financial institutions, imaging centers, and / or storage devices containing medical data. Based on the type of medical data, the digital filing cabinets 180 can organize the medical data by level of authentication. For example, a patient may have access to all medical data, while a healthcare provider may be limited to medical records and not have access to the patient's health insurance or payment information. The digital filing cabinet 180 contains patient medical data 108, which is organized by different authentication levels, such as data 108a, 108b, 108c, and 108d. Each group or set of medical data 108a, 108b, 108c, and 108d requires a different level of authentication for the user to access. Exemplary medical datasets and medical data are described in relation to Figures 8 to 11. After the user's authentication level is identified, the system 100 can send the medical data associated with the identified authentication level to the user. In some implementations, the system 100 sends the medical data to the patient's implant 150 or to a user device for the user to retrieve.

[0020] The number of medical data groups, permission settings, data to be stored, organization method, and / or other configurations can be set by the user, healthcare provider, etc. Data can be automatically collected and incorporated into the appropriate data group. In a cloud-based implementation, the digital filing cabinet 180 may be stored on a cloud server to provide remote access. In some implementations, the digital filing cabinet 180 may be stored locally to provide access to records at any time. Furthermore, the local storage of the digital filing cabinet 180, including a digital wallet containing blockchain information, may be stored locally. Each group of medical data 108a, 108b, 108c, and 108d can be associated by the user (or data management system) with, for example, a procedure, a physician, a healthcare provider, and / or a medical manufacturer. The user can add information, including annotations, personal notes, and other information, which may or may not be visible to other users, to the medical data 108a, 108b, 108c, and 108d, and can select the type, amount, and / or level of permission / access.

[0021] In some embodiments, a group of medical data 108a may be associated with an implant 150 in a patient (not shown) and may include the surgical plan for the implant 150, manufacturing data for the implant 150, notifications (e.g., recall notices), predicted post-treatment analysis, physician information, and other information associated with the implant 150 or procedure (e.g., pre-operative, intra-operative, and / or post-operative information). A user can set up one or more rules to allow authorized users to access medical data 108a or medical data 108 (e.g., all or part of it). For example, a user may allow a physiotherapist to view post-operative data 108a, which can then access the data collected post-operatively to modify the user's treatment plan. A user may allow a primary care physician to access medical data 108 to provide general care, and a surgeon to access medical data 108a to evaluate surgical outcomes and recommend additional treatments, such as future surgical interventions.

[0022] Medical data 108b may include, for example, a patient's general electronic medical records (EMR), which include health records not associated with implants. Users may allow their primary care physician to access medical data 108b to provide general care. Users may also allow family members and third parties to access medical data 108b. Therefore, the access settings for medical data 108a and 108b may be the same or different.

[0023] Medical data 108c may include, for example, data from user device input. This data could be from, for example, wearables (e.g., smartwatches, pedometers, etc.), smartphones, biometric sensors (e.g., analyte sensors, glucose sensors, etc.), cardiac monitors, exercise monitoring devices, etc. The user may allow a family member to access the data 108c, for example, to help adhere to diet goals, exercise goals, or the goals of other users.

[0024] The data may be automatically provided to the digital filing cabinet 180. In some embodiments, for example, the implant retrieval function 160 may provide a command to access the digital filing cabinet 180. Imaging devices (e.g., MRI machines, X-ray machines, scanners, etc.) may read information from the retrieval function 160. This information may be transmitted to the digital filing cabinet 180 via the communication network 104. The transmitted information may include, but is not limited to, authorization information (e.g., login information for the digital filing cabinet), patient identification information, implant identifiers, patient images, and / or other information used to authorize, find, and / or classify the data.

[0025] The digital filing cabinet 180 can store data transmitted from the manufacturing system 124 via the communication network 104, analyze the received data, and associate data such as manufacturing data with the received implant data. Correlation settings can be changed or set by the user. Furthermore, surgical plans can be transmitted to the digital filing cabinet 180 from the implant design platform or system 106 (system 106) via the communication network 104. The surgical plans can be associated with manufacturing data, implant data, and other information associated with the implant 150. The digital filing cabinet 180 can transmit patient medical data to system 106 via the communication network 104. This enables newly available data to be automatically or periodically transmitted to the analysis system 106. The analysis system 106 can analyze the newly received data using, for example, one or more models, and provide the analysis results to client computing devices 102, the digital filing cabinet 180, the manufacturing system 124, physicians, etc. The client computing device 102 can receive analysis results and notifications from, for example, the digital filing cabinet 180, the analysis system 106, and / or other data sources.

[0026] In some embodiments, System 100 is configured to manage patient medical data on user devices, cloud-based devices, and / or healthcare provider devices. Medical data may include patient medical records, health insurance information, health indicators from wearable devices, surgical information, surgical plans, technology recommendations (e.g., device and / or instrument recommendations), and / or medical device information (e.g., implanted medical devices (e.g., also referred to herein as “implants” or “implanted devices”) or implant supply instruments). Digital wallets may be used to manage blockchain medical data (e.g., blockchain EHRs, EMRs, etc.), insurance activities (e.g., payments, claims submissions, etc.), etc.

[0027] In some embodiments, System 100 manages the authentication required to access medical records. Authentication may include blockchain, tokens, keys, biometrics, geographic location, passwords, or any authentication information. Medical data that is unique to a patient is referred to herein as “patient-specific” or “personalized” medical data. The digital filing cabinet 180 may store one or more keys (e.g., private key, public key, etc.), authentication information, and / or other information for accessing data, including patient-associated electronic medical records from a distributed blockchain ledger of electronic medical records. U.S. Patent Application No. 17 / 463,054 discloses a system and method for tracking patient medical records using keys, for example, and is incorporated in its entirety by reference. System 100 may include systems and functions for linking medical devices with patient data, as disclosed in U.S. Patent No. 16 / 990,810, which is incorporated in its entirety by reference. The digital filing cabinet may be used to receive user feedback, as described in U.S. Patent Application No. 16 / 699,447, which is incorporated in its entirety by reference. The system disclosed herein may include a digital filing cabinet for designing medical devices using the method disclosed in U.S. Patent Application No. 16 / 699,447.

[0028] System 100 includes a client computing device 102, which can be a user device such as a smartphone, mobile device, laptop, desktop, personal computer, tablet, phablet, wearable device (e.g., smartwatch), or other device known in the art. As further described herein, the client computing device 102 may include one or more processors and memory storing instructions that can be executed by one or more processors to perform the methods described herein. The client computing device 102 may be associated with a healthcare provider or a patient. Figure 1 shows a single client computing device 102, but in an alternative embodiment, the client computing device 102 may instead be implemented as a client computing system encompassing multiple computing devices, so that the operations described herein with respect to the client computing device 102 can instead be performed by a computing system and / or multiple computing devices.

[0029] The client computing device 102 is configured to receive patient medical data 108 associated with a patient. The patient medical data 108 may include data representing the patient's condition, biostructure, medical status, medical history, preferences, and / or any other information or parameters related to the patient. For example, patient medical data 108 may include medical history, surgical intervention data, treatment outcome data, progress data (e.g., physician's notes), patient feedback (e.g., quality of life questionnaires, feedback obtained using surveys), clinical data, provider information (e.g., physician, hospital, surgical team), patient information (e.g., demographics, sex, age, height, weight, type of illness, occupation, activity level, tissue information, health assessment, comorbidities, health-related quality of life (HRQL)), vital signs, diagnostic results, drug information, allergies, imaging data (e.g., camera images, magnetic resonance imaging (MRI) images, ultrasound images, computerized tomography (CAT) scan images, positron emission tomography (PET) images, X-ray images), and diagnostic equipment information (e.g., manufacturer, model number, specifications, user-selected settings / configurations, etc.). In some embodiments, patient medical data 108 includes data representing one or more of the following: patient identification number (ID), age, sex, body mass index (BMI), lumbar lordosis, Cobb angle, pelvic intrinsic angle, intervertebral disc height, segmental flexibility, bone quality, rotational displacement, and / or spinal treatment level. In some embodiments, the client computing device 102 can locally store the digital filing cabinet 180, medical data 108, and / or other information. The client computing device 102 can store account information to enable the user to automatically access the remote digital filing cabinet or account with or without login credentials.In some embodiments, the client computing device 102 can periodically or continuously receive newly available data (e.g., biometrics from a wearable, user input, etc.) and can transmit all or part of the newly available data to a remote storage system, such as a digital filing cabinet 180 or a server 106.

[0030] The client computing device 102 is operationally connected to the server 106 via the communication network 104, thus enabling data transfer between the client computing device 102 and the server 106. The communication network 104 may be a wired network and / or a wireless network. If the communication network 104 is wireless, it may be implemented using communication technologies such as Visible Light Communication (VLC), Worldwide Interoperability for Microwave Access (WiMAX), Long Term Evolution (LTE), Wireless Local Area Network (WLAN), Infrared Communication (IR), Public Switched Telephone Network (PSTN), radio waves, and / or other communication technologies known in the art.

[0031] Server 106, sometimes referred to as a “healthcare data network” or “healthcare data analytics network,” may include one or more computing devices and / or systems. As further described herein, Server 106 may include one or more processors and memory storing instructions executable by one or more processors to perform the methods described herein. In some embodiments, Server 106 is implemented as a distributed “cloud” computing system or function across any suitable combination of hardware and / or virtual computing resources.

[0032] The cloud analytics integration platform 126 is connected to the communication network 104. The analytics integration platform 126 can analyze medical data and integrate data collected from patients and healthcare providers into a digital filing cabinet. The analytics integration platform 126 can integrate surgical and patient plans, identify important health indicators for patients, display patient information and goals, perform post-operative analysis, and generate notifications for healthcare providers or patients (e.g., monitoring based on data from wearables, requesting updated information, scheduling appointments, or notifying of emergencies).

[0033] The medical implant 150 may be an intervertebral device comprising a body 152 configured to interface with one or more identified anatomical structures (e.g., one or more vertebral bodies or vertebral endplates) at and / or in close proximity to the target implant site (e.g., between one or more vertebral bodies or vertebral endplates). The implant body 152 may include one or more structural features designed to interlock with one or more identified anatomical structures. For example, in the shown embodiment, the implant 150 may include an upper surface 165 and a lower surface (not shown) configured to be fixed to a vertebral body of the spine. In some embodiments, the upper surface 165 and the lower surface may have contours that match the contours of the vertebral endplates so that they "pair" with the corresponding vertebral endplates with which they interlock. Dimensions, contours, topology, composition, and / or other implant data may be part of the EMR. In some embodiments, such as the shown embodiment, the upper surface 165 and / or the lower surface may have texture (e.g., by roughening, carving, ridges, etc.). Texture data can be part of the manufacturing data stored in the EMR. In the case of lordosis correction, the upper and lower surfaces may be angled relative to each other, and the EMR can include the angles and sizes of these surfaces.

[0034] Users (e.g., physicians, healthcare providers, patients, etc.) can access the EMR using the search function 160. For example, in an embodiment where the search function 160 is a barcode corresponding to a unique identifier, the user can scan the search function 160 using, for example, one or more cameras on a computing device, and / or otherwise input the unique identifier into the computing device. After the unique identifier has been input into the computing system, the computing system may send the unique identifier to a remote server (e.g., via a communication network) along with a request to provide a medical dataset specific to the corresponding patient. In response to this request, as described above, the server may find a specific dataset associated with the unique identifier and send this dataset to the computing device for display to the user. The implant 150 may include other functions to assist in accessing the ledger and displaying the EMR.

[0035] The search function 160 may be used to carry patient data, such as a private key for unlocking patient medical records stored in the blockchain ledger. The medical implant 150 may be blockchain-enabled to establish communication contact using proximity mode. The private key stored in the search function 160 may be used to access patient-specific medical data. In some implementations, the medical implant 150 also includes a private blockchain ledger to track EMRs associated with the patient. As the patient undergoes various treatments, new EMRs and updates to existing EMRs are generated and stored as “transactions” in the blockchain ledger. To access the EMRs associated with the patient, a private key from the medical implant 150 must be used to “unlock” the EMRs stored in the blockchain ledger. The patient may provide this private key to healthcare providers and other parties via a secure platform, mobile application, digital key, etc. In some embodiments, the EMRs are encrypted using an encryption key that is decrypted by the healthcare provider. Additionally or alternatively, key change protocols, certificate management protocols (e.g., registration certificate protocols, transaction certificate protocols, etc.), and other protocols may be used for variable access and permissions. Patients can manage their EMR data to share only selected data. For example, a patient can share another section of their EMR while keeping one section private. The system also allows for user-controlled settings, such as settings for minors, family members, relatives, and / or settings controlled by other users.

[0036] EMR may include patient data associated with implant design and the design process. If the implant is an artificial intervertebral disc, for example, the stored data may include kinematic data (e.g., pre-operative patient data, target kinematic data, etc.), manufacturing data, design parameters, target service life data, physician recommendations / notes, etc. The intervertebral disc may include the articular joint implant body, which includes plates contoured to match the vertebral endplate, and custom articular joint members between the plates to provide patient-specific movement. If the implant is an intervertebral cage, the stored data may include the material specifications of the implant body, the dimensions of the implant body, manufacturing data, design parameters, target service life data, physician recommendations / notes, etc. Applications and patents incorporated by reference disclose data that may be associated with the search function 160 (e.g., surgical plans, implant specifications, datasets, etc.).

[0037] In some implementations, patients can set variable permissions regarding access to transactions and details stored on the blockchain ledger. For example, a particular healthcare provider may only be granted access to certain transactions related to a specific type of medical procedure. In other implementations, permissions may be set on a patient basis, including child-specific settings for children using implants.

[0038] The medical implant 150 can also track and monitor various health-related data of a patient. For example, the medical implant 150 may include one or more sensors configured to measure pressure, load, or force applied by anatomical elements to monitor, for example, activity, load, etc. The medical implant 150 can continuously or periodically collect data indicating activity levels, activities performed, disease progression, etc. For example, the load on the implant 150 may be tracked over a period of time. The applications and patents incorporated by reference disclose techniques for monitoring and collecting data and transmitting data. In some embodiments, the medical implant 150 can identify events such as excessive load, spinal imbalance, etc. In some embodiments, the patient is automatically updated and monitored on a blockchain based on activities (e.g., surgical procedures, changes in condition, etc.), physical disabilities (e.g., new physical disabilities, progression of physical disabilities, etc.), and / or medical events. Medical events may include imaging, diagnosis, treatment, and / or outcomes, as well as event data that can be encoded within the blockchain. The collected data may be used as historical patient data to treat another patient. Applications and patents incorporated by reference also disclose the use of historical data, imaging data, surgical plans, simulations, modeled results, treatment protocols, and result values, which can be encoded within the blockchain. The digital filing cabinet can also track and monitor various health-related data of patients and may include one or more blockchain digital wallets for managing blockchain data. The number, configuration, and / or contents of the digital wallets may be selected by the user, physician, etc. The digital wallets may be used to access the blockchain and automatically update the blockchain for any number of implants.

[0039] In some implementations, two or more implants may be used. For example, a patient may have both a spinal implant containing an encoded chip that includes a private blockchain ledger containing a secret key and / or the patient's EMR, and a subcutaneous digital implant. The subcutaneous digital implant acts as an intermediate device and communicates with both the spinal implant containing the secret key and / or private blockchain ledger, and an external computing device such as a patient treatment computing system. The subcutaneous digital implant may contain its own data, such as patient identification information and biometric data. In some embodiments, the subcutaneous digital implant may contain a private blockchain ledger containing the secret key and / or the patient's EMR.

[0040] A patient-specific implant may be any of the implants described herein or any of the patent references incorporated herein by reference. For example, a patient-specific implant may include one or more of the following: screws (e.g., bone screws, spinal screws, pedicle screws, small articular surface screws), intervertebral implant devices (e.g., intervertebral implants), cages, plates, rods, intervertebral discs, fixation devices, spacers, rods, expandable devices, stents, brackets, cords, skeletons, fixation devices, anchors, nuts, bolts, rivets, connectors, ropes, fasteners, joint replacements (e.g., artificial intervertebral discs), gluteal implants, etc.

[0041] A patient-specific implant design may include data representing one or more of the implant's physical properties (e.g., size, shape, volume, material, mass, weight), mechanical properties (e.g., stiffness, strength, modulus of elasticity, hardness), and / or biological properties (e.g., osteointegration, cell adhesion, antimicrobial properties, antiviral properties). For example, the design of an orthopedic implant may include the implant's shape, size, material, and / or effective stiffness (e.g., lattice density, number of struts, strut placement, etc.).

[0042] Additional implant types, configurations, and structural features suitable for interlocking with identified anatomical features are described, for example, in U.S. Patent Application No. 16 / 207,116 filed December 1, 2018, and U.S. Patent Application No. 16 / 987,113 filed August 6, 2020, the disclosures of which are incorporated herein by reference in their entirety. For example, medical implants may be pedicles, patient-specific implants, intervertebral implant systems, artificial intervertebral discs, expandable intervertebral implants, sacroiliac joint implants, plates, arthroplasty devices for orthopedic joints, non-structural implants, or other devices disclosed in patents and applications incorporated herein by reference.

[0043] Medical implant 150 may be used to track and monitor medical data associated with a patient. U.S. Patent Application No. 63 / 218,190 discloses an implant capable of collecting data, assigning weights / values, and communicating with other devices. Monitoring may be used in conjunction with normative systems such as those disclosed in U.S. Patent No. 10,902,944 and U.S. Patent Application No. 17 / 342,439, which are incorporated in their entirety by reference. For example, patient data may be incorporated into one or more training sets for machine learning systems or other systems disclosed in the patents and applications incorporated by reference. Medical implant 150 may also be a multipurpose implant that provides a structure to address medical problems within a patient's body while also carrying information about the patient. For example, medical implant 150 may be a pacemaker, a plate or pin to correctly position a previously fractured bone or a set of bones, etc. The digital filing cabinet 180 can also be incorporated into the systems disclosed in U.S. Patent No. 10,902,944 and U.S. Patent Application No. 17 / 342,439 for tracking and monitoring patient-managed medical data.

[0044] The client computing device 102 and the server 106 can individually or collectively perform the various methods described herein for storing and retrieving medical data. For example, some or all of the steps of the methods described herein can be performed by the client computing device 102 alone, the server 106 alone, or a combination of the client computing device 102 and the server 106. In some embodiments, the client computing device 102 includes one or more digital filing cabinets 180. It should be understood that although certain operations are described herein in relation to the server 106, these operations can also be performed by the client computing device 102, and vice versa.

[0045] Server 106 includes at least one database 110 configured to store reference data that helps provide, manage, or analyze patient-specific medical data from the implantation methods described herein. The reference data may include historical and / or clinical data from the same or other patients, data collected from previous surgeries and / or treatments of the patient by the same or other healthcare providers, data related to medical device design, data collected from laboratory or research groups, data from practice databases, data from academic institutions, data from implant manufacturers or other medical device manufacturers, data from imaging studies, data from simulations, clinical trials, demographic data, treatment data, outcome data, mortality rates, etc.

[0046] In some embodiments, the database 110 includes multiple reference patient datasets, each patient reference dataset associated with a corresponding reference patient. For example, a reference patient may be a patient who has previously received treatment or is currently receiving treatment. Each reference patient dataset may include data representing any other information or parameters about the reference patient, such as the corresponding reference patient's condition, biostructure, medical condition, medical history, disease progression, preferences, and / or any of the data described herein with respect to medical data 108. In some embodiments, the reference patient dataset includes pre-operative data, intra-operative data, and / or post-operative data. For example, the reference patient dataset may include data representing one or more of the following: patient ID, age, sex, BMI, lumbar lordosis, Cobb angle, pelvic intrinsic angle, intervertebral disc height, segmental flexibility, bone quality, rotational displacement, and / or spinal treatment level. As another example, the reference patient dataset may include treatment data relating to at least one treatment procedure performed on the reference patient, such as a description of the surgical procedure or intervention (e.g., surgical technique, osteotomy, surgical manipulation, corrective operation, implant or other device placement). In some embodiments, treatment data may include medical device design data for at least one medical device used to treat a reference patient, such as physical properties (e.g., size, shape, volume, material, mass, weight), mechanical properties (e.g., stiffness, strength, modulus of elasticity, hardness), and / or biological properties (e.g., osteointegration, cell adhesion, antimicrobial properties, antiviral properties). In yet another example, the reference patient dataset may include outcome data representing the treatment outcomes of the reference patient, such as modified anatomical indicators, presence of fixation, HRQL, activity level, return to work, complications, recovery time, efficacy, mortality, and / or reoperation.

[0047] In some embodiments, server 106 receives at least a portion of reference patient datasets from multiple healthcare provider computing systems (e.g., systems 112a-112c, collectively 112), digital filing cabinets, or a combination thereof. Server 106 may be connected to the healthcare provider computing systems 112 via one or more communication networks (not shown). Each healthcare provider computing system 112 may be associated with a corresponding healthcare provider (e.g., a physician, surgeon, clinic, hospital, healthcare network, etc.). Each healthcare provider computing system 112 may include at least one reference patient dataset (e.g., reference patient datasets 114a-114c, collectively 114) associated with reference patients treated by the corresponding healthcare provider. The reference patient datasets 114 may include, for example, electronic medical records, electronic health records, biomedical datasets, etc. The reference patient datasets 114 can be received from the healthcare provider computing systems 112 by server 106 and may be reformatted to different formats for storage in the database 110. Optionally, the reference patient dataset 114 may be processed (e.g., by removing unnecessary parameters) to ensure that the expressed patient parameters are more likely to be useful in the treatment planning methods described herein.

[0048] As will be described in more detail herein, the server 106 may be configured using one or more algorithms that generate patient-specific treatment plan data (e.g., treatment procedures, medical devices) based on reference data. In some embodiments, patient-specific data is generated based on correlations between patient dataset 108 and reference data. Optionally, the server 106 can predict outcomes, including recovery time, efficacy based on clinical endpoints, likelihood of success, predicted mortality, predicted related reoperations, etc. In some embodiments, the server 106 can continuously or periodically analyze patient data (including patient data acquired during the patient's stay) to determine near real-time or real-time risk scores, mortality predictions, etc.

[0049] In some embodiments, the server 106 includes one or more modules for performing one or more steps of the patient-specific treatment planning method described herein. For example, in the shown embodiment, the server 106 includes a data analysis module 116 and a treatment planning module 118. In alternative embodiments, one or more of these modules may be combined with each other or omitted. Thus, while certain operations are described herein in relation to a particular one or more modules, this is not intended to be limiting, and in alternative embodiments, such operations may be performed by different one or more modules.

[0050] The data analysis module 116 is configured using one or more algorithms to identify a subset of reference data from database 110 that is likely to be useful in developing a patient-specific treatment plan. Database 110 can retrieve or receive data from a client computing device 102, a digital filing cabinet 180, or other data sources. For example, the data analysis module 116 can compare patient-specific data (e.g., a patient dataset 108 received from a client computing device 102) with reference data from database 110 (e.g., a reference patient dataset) to identify similar data (e.g., one or more similar patient datasets within the reference patient dataset). The reference data may be updated in real time or near real time using other patient data accessible via network 104. This comparison may be based on one or more parameters, such as age, sex, BMI, lumbar lordosis, pelvic intrinsic angle, and / or treatment level. These parameters may be used to calculate a similarity score for each reference patient. The similarity score may represent a statistical correlation between the patient dataset 108 and the reference patient dataset. Therefore, similar patients can be identified based on whether their similarity score is above, below, or at a specified threshold. For example, as will be explained in more detail below, this comparison can be performed by assigning values ​​to each parameter and determining the sum of the differences between the patient in question and each reference patient. Reference patients whose sum of differences is below the threshold may be considered similar patients.

[0051] The data analysis module 116 may further be configured using one or more algorithms for selecting a subset of reference patient datasets based on similarity to the patient dataset 108 and / or the treatment outcomes of corresponding reference patients. For example, the data analysis module 116 may identify one or more similar patient datasets within the reference patient dataset and then select a subset of similar patient datasets based on whether the similar patient datasets contain data indicating beneficial or desirable treatment outcomes. The outcome data may include data representing one or more outcome parameters, such as modified anatomical indicators, presence of fixation, HRQL, activity level, complications, recovery time, efficacy, mortality, or reoperation. In some embodiments, as described in more detail below, the data analysis module 116 calculates an outcome score by assigning a value to each outcome parameter. If the outcome score is higher than, lower than, or at a specified threshold, the patient may be considered to have a beneficial outcome.

[0052] In some embodiments, the data analysis module 116 selects a subset of reference patient datasets based at least in part on user input (e.g., from clinicians, surgeons, physicians, or healthcare providers). For example, user input may be used to identify similar patient datasets. In some embodiments, weighting of similarity and / or outcome parameters may be selected by the healthcare provider or physician to adjust similarity and / or outcome scores based on clinician input. In further embodiments, the healthcare provider or physician may each select (or define new similarity and / or outcome parameters) a set of similarity and / or outcome parameters to be used to generate similarity and / or outcome scores.

[0053] In some embodiments, the data analysis module 116 includes one or more algorithms used to select a set or subset of reference patient datasets based on criteria other than patient parameters. For example, one or more algorithms may be used to select a subset based on healthcare provider parameters (e.g., based on healthcare provider ranking / scores, such as hospital / physician expertise, number of procedures performed, hospital ranking, etc.) and / or healthcare resource parameters (e.g., diagnostic equipment, facilities, surgical equipment such as surgical robots), or non-patient related information that can be used to predict outcomes and risk profiles for current healthcare provider procedures. For example, reference patient datasets including images captured from similar diagnostic equipment may be aggregated to mitigate or limit irregularities caused by variability between diagnostic equipment. Furthermore, data from similar healthcare providers (e.g., healthcare providers that traditionally have similar outcomes, physician expertise, surgical teams, etc.) may be used to develop patient-specific treatment plans for a particular healthcare provider. In some embodiments, reference healthcare provider datasets, hospital datasets, physician datasets, surgical team datasets, post-treatment datasets, and other datasets may be available. For example, a patient-specific treatment plan for performing surgery in a battlefield can be based on reference patient data from similar battlefield surgeries and / or datasets associated with battlefield surgeries. In another example, a patient-specific treatment plan may be generated based on available robotic surgery systems. The reference patient dataset may be selected based on patients who have undergone surgery using equivalent robotic surgery systems according to similar conditions (e.g., size and capabilities of the surgical team, hospital resources, etc.).

[0054] The treatment planning module 118 is configured with one or more algorithms for generating at least one treatment plan or recovery protocol (e.g., pre-operative plan, surgical plan, post-operative plan, etc.) based on the output from the data analysis module 116. In some embodiments, the treatment planning module 118 is configured to develop and / or implement at least one predictive model, also known as a “prescriptive model,” to generate a patient-specific treatment plan. The predictive model may be developed using clinical knowledge, statistics, machine learning, AI, neural networks, etc. In some embodiments, the output from the data analysis module 116 is analyzed (e.g., using statistics, machine learning, neural networks, AI) to identify correlations between datasets, patient parameters, healthcare provider parameters, medical resource parameters, treatment procedures, medical device designs, and / or treatment outcomes. These correlations may be used to develop at least one predictive model that predicts the likelihood that a treatment plan will produce beneficial outcomes for a particular patient. The predictive model may be validated, for example, by inputting data into the model and comparing the model’s output with expected outputs and actual outputs after treatment.

[0055] In some embodiments, the treatment planning module 118 is configured to generate a treatment plan based on previous treatment data from a reference patient. For example, the treatment planning module 118 can receive a selected subset of reference patient datasets and / or similar patient datasets from the data analysis module 116 and determine or identify treatment data from the selected subset. Treatment data may include, for example, treatment procedure data (e.g., surgical procedure or intervention data) and / or medical device design data (e.g., implant design data) associated with beneficial or desirable treatment outcomes for the corresponding patient. The treatment planning module 118 can analyze the treatment procedure data and / or medical device design data to determine the optimal treatment protocol for treating the patient. For example, treatment procedures and / or medical device designs may be assigned values ​​and aggregated to generate a treatment score. A patient-specific treatment plan may be determined by selecting a treatment plan based on this score (e.g., a higher or best score, a lower or worst score, a score higher or lower than a specified threshold, or a score at or above a specified threshold). A personalized, patient-specific treatment plan can be based at least partially on patient-specific techniques or patient-specific selected techniques.

[0056] Alternatively, or in combination, the treatment planning module 118 can generate treatment plans based on correlations between datasets. For example, the treatment planning module 118 can associate treatment procedure data and / or medical device design data from similar patients with beneficial outcomes (such as those identified by the data analysis module 116). Correlation analysis may include converting correlation coefficient values ​​into values ​​or scores. These values / scores may be aggregated, filtered, or otherwise analyzed to determine one or more statistical significance. These correlations may be used to determine which treatment procedures and / or medical device designs are optimal for treating a patient or are likely to produce beneficial outcomes.

[0057] Alternatively, or in combination, the treatment planning module 118 may generate treatment plans using one or more AI technologies. AI technologies may be used to develop computing systems that can simulate aspects of human intelligence, such as learning, reasoning, planning, problem-solving, and decision-making. AI technologies may include, but are not limited to, case-based reasoning, rule-based systems, artificial neural networks, decision trees, support vector machines, regression analysis, Bayesian networks (e.g., naive Bayesian classifiers), genetic algorithms, cellular automata, fuzzy logic systems, multi-agent systems, swarm intelligence, data mining, machine learning (e.g., supervised learning, unsupervised learning, reinforcement learning), and hybrid systems.

[0058] In some embodiments, the treatment planning module 118 generates treatment plans using one or more trained machine learning models. Various types of machine learning models, algorithms, and techniques are suitable for use with this technique. In some embodiments, the machine learning model is first trained on a training dataset, which is a set of examples used to fit the model's parameters (e.g., the weights of the connections between "neurons" in an artificial neural network). For example, the training dataset may include any of the reference data stored in the database 110, such as multiple reference patient datasets or selected subsets thereof (e.g., multiple similar patient datasets).

[0059] In some embodiments, a machine learning model (e.g., a neural network or a Naive Bayes classifier) ​​may be trained on a training dataset using a supervised learning method (e.g., gradient descent or stochastic gradient descent). The training dataset may include pairs of generated "input vectors" with associated corresponding "response vectors" (generally referred to as targets). The current model is run on the training dataset to produce results, which are then compared to targets for each input vector in the training dataset. Based on the results of the comparison and the specific learning algorithm used, the model's parameters are adjusted. Model fitting may include both variable selection and parameter estimation. The fitted model may be used to predict responses to observations in a second dataset called a validation dataset. The validation dataset can provide an unbiased assessment of the model's fit to the training dataset while adjusting the model parameters. The validation dataset may be used for regularization by stopping training early, for example, when the error on the validation dataset increases, which may be an indication of overfitting to the training dataset. The error on the validation dataset can fluctuate during training, so that in some embodiments, an ad-hoc rule may be used to determine when overfitting truly began. Finally, the test dataset may be used to provide an unbiased assessment of the final model fit to the training dataset.

[0060] To generate treatment plans, a patient dataset 108 may be input to a trained machine learning model. Additional data, such as a selected subset of a reference patient dataset and / or similar patient datasets, and / or treatment data from the selected subset, may also be input to the trained machine learning model. The trained machine learning model can then calculate whether various candidate treatment procedures and / or medical device designs are likely to produce beneficial outcomes for the patient. Based on these calculations, the trained machine learning model can select at least one treatment plan for the patient. In embodiments where multiple trained machine learning models are used, the models can be run sequentially or concurrently to compare results and may be periodically updated using the training dataset. The treatment planning module 118 may use one or more of the machine learning models based on the predicted accuracy scores of the models.

[0061] A patient-specific treatment plan generated by the treatment planning module 118 may include at least one patient-specific treatment procedure (e.g., a surgical procedure or intervention) and / or at least one patient-specific medical device (e.g., an implant or implant supply device). The patient-specific treatment plan may include the entire surgical procedure or a portion thereof. Furthermore, one or more patient-specific medical devices may be selected or designed specifically for the corresponding surgical procedure, thus enabling the combination and use of various patient-specific components to treat the patient.

[0062] In some embodiments, patient-specific treatment procedures include orthopedic procedures such as spinal surgery, hip surgery, knee surgery, jaw surgery, hand surgery, shoulder surgery, elbow surgery, total joint reconstruction (arthroplasty), skull reconstruction, foot surgery, or ankle surgery. Spinal surgery may include spinal fusion procedures such as posterior lumbar interbody fusion (PLIF), anterior lumbar interbody fusion (ALIF), transverse or transforaminal lumbar interbody fusion (TLIF), lateral lumbar interbody fusion (LLIF), direct lateral lumbar interbody fusion (DLIF), or extreme lateral lumbar interbody fusion (XLIF). In some embodiments, patient-specific treatment procedures include instructions for performing and / or commands for performing one or more aspects of a patient-specific surgical procedure. For example, a patient-specific surgical procedure may include one or more of surgical techniques, corrective operations, osteotomies, or implant placements.

[0063] In some embodiments, patient-specific medical device design includes the design of orthopedic implants and / or instruments for supplying orthopedic implants. Examples of such implants include, but are not limited to, screws (e.g., bone screws, spinal screws, pedicle screws, articular surface screws), intervertebral implant devices (e.g., intervertebral implants), cages, plates, rods, intervertebral discs, fixation devices, spacers, rods, expandable devices, stents, brackets, cords, skeletons, fixation devices, anchors, nuts, bolts, rivets, connectors, ropes, fasteners, joint replacements, gluteal implants, etc. Examples of instruments include, but are not limited to, screw guides, cannulas, ports, catheters, insertion tools, etc.

[0064] A patient-specific medical device design may include data representing one or more of the physical properties (e.g., size, shape, volume, material, mass, weight), mechanical properties (e.g., stiffness, strength, modulus of elasticity, hardness), and / or biological properties (e.g., osteointegration, cell adhesion, antimicrobial properties, antiviral properties) of the corresponding medical device. For example, the design of an orthopedic implant may include the shape, size, material, and / or effective stiffness (e.g., lattice density, number of struts, strut positions, etc.) of the implant. In some embodiments, the generated patient-specific medical device design is the design of the entire device. Alternatively, the generated design may be the design of one or more components of the device, rather than the entire device.

[0065] In some embodiments, the design is the design of components for one or more patient-specific devices that can be used with standard commercially available components. For example, in spinal surgery, a pedicle screw kit may include both standard components and patient-specific, customized components. In some embodiments, the generated design is the design of a patient-specific medical device that can be used with standard commercially available supply instruments. For example, the implant (e.g., screw, screw holder, rod) may be designed and manufactured for the patient, while the instrument for supplying the implant may be a standard instrument. This approach allows the implanted components to be designed and manufactured based on the patient's biostructure and / or the surgeon's preferences in order to improve treatment. The patient-specific devices described herein are expected to improve delivery into the patient's body, placement at the treatment site, and / or interaction with the patient's biostructure.

[0066] The system can analyze a design and / or virtual model (e.g., an implant model, an anatomical model, etc.) to determine one or more geometric / shape deviations compared to a reference implant, predicted post-operative metrics that are outside the acceptable range, etc. In some embodiments, system 100 identifies nonconformities for analysis. Criteria for identifying nonconformities can be entered by the user, generated based on patient data and / or the surgical plan, etc. In response to identifying that a nonconformity feature meets a nonconformity risk threshold, system 100 can generate a nonconformity report relating to the implant design, predicted anatomical outcomes, implant manufacturing, etc. Nonconformity reports can be generated at various points in the design and / or manufacturing process and may include virtual model data to show the fit of the virtual model of the implant with one or more anatomical features of the patient. Nonconformity reports may also be generated with respect to instruments or other items disclosed herein.

[0067] In embodiments where a patient-specific treatment plan includes a surgical procedure for implanting a medical device, the treatment planning module 118 may also store various types of implant surgery information, such as implant parameters (e.g., type, dimensions), implant availability, pre-operative planning aspects (e.g., initial implant configuration, detection, and measurement of the patient's biostructure), and FDA requirements for the implant (e.g., specific implant parameters and / or characteristics for compliance with FDA regulations). In some embodiments, the treatment planning module 118 may convert the implant surgery information into a format usable by machine learning-based models and algorithms. For example, the implant surgery information may be tagged with specific identifiers for formulas or converted into a numerical representation suitable for feeding into a trained machine learning model. The treatment planning module 118 may also store information about the patient's biostructure, such as two-dimensional or three-dimensional images or models of the biostructure, as well as information about the biological, morphological, and / or mechanical properties of the biostructure. The biostructure information may be used to inform the design and / or placement of the implant.

[0068] The treatment plan generated by the treatment planning module 118 may be transmitted via the communication network 104 to a digital filing cabinet 180 and / or a client computing device 102 for output to a user (e.g., a clinician, surgeon, healthcare provider, or patient). In some embodiments, the client computing device 102 includes or is operably coupled to a display for outputting the treatment plan. The display may include a graphical user interface (GUI) for visually representing various aspects of the treatment plan. For example, the display may show various aspects of the surgical procedure performed on the patient, such as the surgical technique, treatment level, corrective operation, tissue excision, and / or implant placement. For easier visualization, a virtual model of the surgical procedure may be displayed. As another example, the display may show the design of a medical device to be implanted in the patient, such as a two-dimensional or three-dimensional model of the device design. The display may also show patient information, such as two-dimensional or three-dimensional images or models of the patient's biostructure on which the surgical procedure is performed and / or the patient's biostructure on which the device is implanted. The client computing device 102 may further include one or more user input devices (not shown) that enable the user to modify, select, approve, and / or reject the displayed treatment plan.

[0069] In some embodiments, the medical device design generated by the treatment planning module 118 may be transmitted from the client computing device 102 and / or server 106 to the manufacturing system 124 for the manufacture of the implant or corresponding medical device. The manufacturing system 124 may be located on-site or remotely. The implant may be manufactured by any suitable manufacturing system (e.g., the manufacturing system 124 shown in Figure 1). The digital filing cabinet 180 may store the generated medical device design, manufacturing data (e.g., CAM data, print data, etc.), manufacturing information, data for generating the surgical plan, the surgical plan, the surgical plan report, post-surgical data (e.g., treatment plan, predicted outcome, etc.), and / or other information associated with the medical device.

[0070] Various types of manufacturing systems are suitable for use according to the embodiments described herein. For example, manufacturing system 124 may be configured for additive manufacturing, such as three-dimensional (3D) printing, stereolithography (SLA), digital light processing (DLP), fused deposition modeling (FDM), selective laser sintering (SLS), selective laser melting (SLM), selective heat sintering (SHM), electronic beam melting (EBM), laminated object manufacturing (LOM), powder bed printing (PP), thermoplastic printing, direct material deposition (DMD), inkjet photoresin printing, or similar techniques, or combinations thereof. Alternatively, or in combination, the manufacturing system 124 may be configured for subtractive (conventional) manufacturing, such as CNC machining, electrical discharge machining (EDM), grinding, laser cutting, waterjet machining, manual machining (e.g., milling, turning / rotating), or similar techniques, or combinations thereof. The manufacturing system 124 can manufacture one or more patient-specific medical devices based on manufacturing instructions or data (e.g., CAD data, 3D data, digital blueprints, stereolithography data, or other data suitable for the various manufacturing techniques described herein). Various components of system 100 can generate at least some of the manufacturing data used by the manufacturing system 124.Manufacturing data may include, but is not limited to, manufacturing instructions (e.g., programs executable by additive manufacturing equipment, subtractive manufacturing equipment, etc.), 3D data, CAD data (e.g., CAD files), CAM data (e.g., CAM files), path data (e.g., print head paths, tool paths, etc.), material data, tolerance data, surface finish data (e.g., surface roughness data), regulatory data (e.g., FDA requirements, repayment data, etc.), etc. Based on the received manufacturing data, the manufacturing system 124 can analyze the manufacturability of the implant design. The implant design may be finalized by modifying the shape, surface, etc., and then generating manufacturing instructions. In some embodiments, the server 106 generates at least a portion of the manufacturing data to be sent to the manufacturing system 124.

[0071] The manufacturing system 124 can generate CAM data, printing data (e.g., powder layer printing data, thermoplastic printing data, photoresin data, etc.), and may include additive manufacturing equipment, subtractive manufacturing equipment, heat treatment equipment, etc. Additive manufacturing equipment may be 3D printers, stereolithography devices, digital light processing devices, fused deposition modeling devices, selective laser sintering devices, selective laser melting devices, electron beam melting devices, laminate manufacturing devices, powder layer printers, thermoplastic printers, direct material deposition devices, or inkjet photoresin printers, or similar technologies. Subtractive manufacturing equipment may be CNC machines, electrical discharge machines, grinders, laser cutters, water jet machines, manual machines (e.g., milling machines, lathes, etc.), or similar technologies. Both additive and subtractive technologies may be used to manufacture implants with complex shapes, surface finishes, material properties, etc. The generated manufacturing instructions may be configured to cause the manufacturing system 124 to manufacture patient-specific orthopedic implants that match or are therapeutically identical to a patient-specific design. In some embodiments, patient-specific medical devices may include features, materials, and designs shared throughout the design to simplify manufacturing. For example, deployable patient-specific medical devices for different patients may have similar internal deployment mechanisms but different deployed configurations. In some embodiments, components of a patient-specific medical device may be selected from a set of available off-the-shelf components, and the selected off-the-shelf components may be modified based on manufacturing instructions or data.

[0072] After a patient has been treated according to a treatment plan, the progress of treatment may be monitored over one or more periods to update the data analysis module 116 and / or the treatment planning module 118. Post-treatment data may be added to reference data stored in the database 110 and used for post-operative analysis. Post-treatment data may be used to train machine learning models to develop patient-specific treatment plans, patient-specific medical devices, or a combination thereof.

[0073] System 100 can generate implant data (e.g., design files, manufacturing instructions, etc.) for manufacturing an implant (e.g., implant 150). System 100 can manage access to the implant data based on authentication levels. Authentication levels may include, but are not limited to, authenticating users (e.g., manufacturers, healthcare providers, etc.) based on geographical location, biometric data, blockchain access, tokens, or any authentication method. Based on a determined authentication level of a user requesting access to the implant data, the system may permit the user to access some or all of the implant data. System 100 may include one or more medical digital filing cabinets (e.g., digital filing cabinet 180) for storing implant data, patient information, electronic medical records, and / or additional patient-related information. In some embodiments, the system may share data (e.g., implant data, medical data, patient information, electronic medical records, and / or additional patient-related information) over a network without using digital filing cabinets. Digital filing cabinets may use blockchain and non-fungible token (NFT) technologies to control the collection and access of implant data.

[0074] In some embodiments, system 100 performs quality checks on manufactured implants. System 100 may include one or more implant analyzers 127 that can scan manufactured implants to identify errors on the manufactured implants. The manufacturing system 124, implant analyzers 127, and / or certification manager 119 can communicate with each other directly or via a communication network 104. The implant analyzer 127 may include one or more scanners 129. The implant analyzer 127 may be integrated into the manufacturing system 124 or other components of system 100. For example, the scanner 129 may be a field manufacturing scanner positioned to scan implants during and / or after manufacturing. In some embodiments, the implant analyzer 127 is located away from the manufacturing site. For example, the analyzer 127 may be located at a healthcare provider (e.g., a hospital, clinic, etc.) to enable quality control checks of implants immediately before implant treatment. Based on the identified errors, system 100 may determine adjustments to implants to be remanufactured. System 100 can determine whether an implant can be safely fitted to a patient by analyzing the implant parameters of the manufactured implant (e.g., material composition, temperature, printing speed, manufacturing conditions, printer precision, etc.).

[0075] System 100 may be updated periodically or continuously to expand its capabilities to provide effective treatment for patients with a wide range of symptoms not included in, for example, machine learning training sets or default treatment parameters. System 100 can identify edge cases for human review, compare results between automated and human review, verify analysis results, and / or provide automated annotations / recommendations for human review. System 100 can receive user input and then use this input for one or more of the following: human-assisted implant design, human-assisted simulation, human-assisted prediction, etc. In this way, System 100 may be expanded for effective interventions for patients with, for example, complex edge case conditions, unknown conditions, etc.

[0076] In some embodiments, the system 100 analyzes patient data retrieved from the digital filing cabinet 180, patient data received from the scanner 129, patient data retrieved from the database 110, or patient data retrieved from the treatment planning module 118 (for example, via the data analysis module 116). The system 100 can identify edge case conditions in the patient data that may affect treatment. For example, identified edge case conditions may require human review. System 100 identifies edge-case medical conditions based on one or more parameters, such as modifications of intervertebral fusion, anatomical abnormalities (e.g., hemivertebrae, grade 2+ spondylolisthesis), osteotomy symptoms, implant atypical symptoms, patient tumors, fractures in the patient, torn or partially torn tissue (e.g., ligaments, tendons, or muscles), implants adjacent to or near the patient's spine, transitional vertebrae (e.g., extra vertebrae), or any existing patient conditions in patient data that may affect, for example, implant treatment, implant performance, treatment outcomes, etc.

[0077] System 100 can determine thresholds for each type of edge case parameter. For example, if a tumor is detected in patient data, the threshold could be the size of the tumor or the distance from the implant to the tumor. Based on the type of edge case condition, System 100 can send a notification to Device 102 for a human (e.g., healthcare provider, implant designer, etc.) to review the patient data and the identified edge case condition. In some implementations, System 100 utilizes a machine learning model to identify edge case conditions. The machine learning model may be retrained periodically or continuously using confirmed edge case condition data based on confirmation of results by human review.

[0078] System 100 can analyze patient data and / or implant data to identify whether human review is required at any stage of the patient's specific orthopedic implant procedure (e.g., design, manufacturing, implant treatment, or post-implant treatment). System 100 can generate a patient-linked manufacturing hold to at least temporarily prevent the manufacture of implants unsuitable for implant treatment in the patient. This ensures that System 100 does not inadvertently manufacture implants that are improperly designed for the edge case medical condition.

[0079] A manufacturing hold may be linked to one or more of the following: for example, the patient's electronic medical record, patient account, patient identifier, or other records associated with the patient. Before completing the implant design and manufacturing process, system 100 must clear the manufacturing hold. A manufacturing hold may be cleared, for example, upon completion of a review of the edge case medical condition. This completion may be treatment approval, for example, by a human review officer, a machine learning edge case review module, etc. For some procedures, the decision of whether to remove a manufacturing hold may be based on input from a human review officer who meets the patient's treatment threshold. For example, the input received may indicate that the medical condition identified as an edge case medical condition is unlikely to affect treatment. The treatment threshold may include one or more of the following: for example, a threshold for the likelihood of affecting treatment, the severity of the reduction in outcome, etc.

[0080] It should be understood that the components of system 100 can be configured in various ways. For example, in an alternative embodiment, the database 110, the data analysis module 116, and / or the treatment planning module 118 may be components of a client computing device 102 rather than a server 106. As another example, the database 110, the data analysis module 116, and / or the treatment planning module 118 may be deployed across multiple different servers, computing systems, or other types of cloud computing resources rather than on a single server 106 or client computing device 102.

[0081] Furthermore, in some embodiments, System 100 may be able to operate in conjunction with a number of other computing system environments or configurations. Examples of computing systems, environments, and / or configurations that may be suitable for use with this technology include, but are not limited to, personal computers, server computers, handheld or laptop devices, mobile phones, wearable electronic devices, tablet devices, multiprocessor systems, microprocessor-based systems, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments that include any of these systems or devices.

[0082] Figure 2 shows a computing device 200 suitable for use in relation to the system 100 of Figure 1, according to an embodiment. The computing device 200 may be incorporated into various components of the system 100 of Figure 1, such as a client computing device 102 or a server 106. The computing device 200 includes one or more processors 210 (e.g., a CPU, GPU, HPU, etc.). The processor 210 may be a single processing unit or multiple processing units, either within a single device or distributed across multiple devices. The processor 210 may be coupled to other hardware devices using a bus, such as a PCI bus or a SCSI bus. The processor 210 may be configured to execute one or more computer-readable program instructions, such as program instructions for performing any of the methods described herein.

[0083] The computing device 200 may include, for example, one or more input devices 220 that provide input to the processor 210 in order to notify the processor 210 of actions from the user of the device 200. These actions may be mediated by a hardware controller, which interprets the signals received from the input devices and transmits this information to the processor 210 using a communication protocol. The input devices 220 may include, for example, a mouse, keyboard, touchscreen, infrared sensor, touchpad, wearable input device, camera or image-based input device, microphone, or other user input device.

[0084] The computing device 200 may include a display 230 used to display various types of output, such as text, models, virtual procedures, surgical plans, implants, graphics, and / or images (e.g., images containing voxels showing radiation density units or Hounsfield units representing tissue density at a given location). In some embodiments, the display 230 provides the user with graphical and textual visual feedback. The processor 210 can communicate with the display 230 via the device's hardware controller. In some embodiments, the display 230 includes the input device 220 as part of the display 230, such as when the input device 220 includes a touchscreen or has an eye-direction monitoring system. In alternative embodiments, the display 230 is separate from the input device 220. Examples of display devices include LCD display screens, LED display screens, projected displays, holographic displays, or augmented reality displays (e.g., head-up display devices or head-mounted devices).

[0085] Optionally, other I / O devices 240, such as network cards, video cards, audio cards, USB, FireWire or other external devices, cameras, printers, speakers, CD-ROM drives, DVD drives, disk drives, or Blu-ray devices, may also be coupled to the processor 210. Other I / O devices 240 may also include input ports for information from directly connected medical devices, such as imaging devices including MRI machines, X-ray machines, and CT machines. Other I / O devices 240 may further include input ports for receiving data from these types of machines from other sources, such as over a network, or from previously captured data stored, for example, in a database.

[0086] In some embodiments, the computing device 200 also includes a communication device (not shown) that can communicate with network nodes wirelessly or via wire. The communication device can communicate with other devices or servers over the network, for example, using the TCP / IP protocol. The computing device 200 can utilize the communication device to distribute its operations across multiple network devices, including imaging equipment, manufacturing equipment, and so on.

[0087] The computing device 200 may include memory 250, which may reside within a single device or be distributed across multiple devices. Memory 250 may include one or more of various hardware devices for volatile and non-volatile storage, and may include both read-only and writable memory. For example, memory may include random access memory (RAM), various caches, CPU registers, read-only memory (ROM), and writable non-volatile memory such as flash memory, hard drives, floppy disks, CDs, DVDs, magnetic storage devices, tape drives, and device buffers. Memory is not a signal that propagates independently from the underlying hardware; therefore, memory is non-transient. In some embodiments, memory 250 is a non-transient computer-readable storage medium that stores, for example, programs, software, data, etc. In some embodiments, memory 250 may include program memory 260 that stores programs and software such as an operating system 262, one or more medical data modules 264, and other application programs 266. The application program 266 may include, but is not limited to, authentication programs, enrollment programs, manufacturing programs, diagnostic programs, and report generation programs. The medical data module 264 may include one or more modules configured to perform various methods described herein (e.g., the data analysis module 116 and / or treatment planning module 118 described with respect to Figure 1). The memory 250 may also include a data memory 270 that may be provided to the program memory 260 of the computing device 200 or any other element, and may include, for example, reference data, configuration data, settings, user choices or preferences.

[0088] Figure 3 is a system diagram showing an example of a computing environment in which the disclosed system operates in some embodiments. In some embodiments, the environment 300 includes one or more client computing devices 305A-D, which may host device 200. The client computing devices 305 operate in a networked environment using logical connections to one or more remote computers, such as server computing devices, via the network 330. In some embodiments, the client computing devices 305 may also include medical implants, such as the medical implant 150 described above in relation to Figure 1.

[0089] In some embodiments, device 310 is an edge server that receives client requests and coordinates the fulfillment of those requests through other servers such as servers 320A-C. In some embodiments, server computing devices 310 and 320 comprise a computing system such as device 200. Although each server computing device 310 and 320 appears logically as a single server, each server computing device can be a distributed computing environment encompassing multiple computing devices located in the same physical location or in geographically completely different physical locations. In some embodiments, each server computing device 320 corresponds to a group of servers.

[0090] The client computing device 305 and the server computing devices 310 and 320 each function as a server or client to other servers or client devices. In some embodiments, the servers (310, 320A-C) connect to corresponding databases (315, 325A-C). As previously stated, each server 320 may correspond to a group of servers, each of which may share a database or have its own database. Databases 315 and 325 store (e.g., store) information such as medical information, health records, user biometric information, blockchain transactions including user medical records, and other data. In some embodiments, servers 320A-C may include features of other servers disclosed herein, such as a digital filing cabinet and / or server 106 in Figure 1. Although databases 315 and 325 are logically presented as a single unit, databases 315 and 325 may each be a distributed computing environment encompassing multiple computing devices, located within their corresponding servers, or located in the same physical location or in geographically completely different physical locations.

[0091] Network 330 can be a local area network (LAN) or a wide area network (WAN), but it can also be any other wired or wireless network. In some embodiments, network 330 is the Internet or any other public or private network. Client computing devices 305 connect to network 330 via a network interface, such as by wired or wireless communication. Although the connection between server 310 and server 320 is shown as a separate connection, these connections can be any kind of local area network, wide area network, wired network, or wireless network, including network 330 or separate public or private networks.

[0092] Figure 4 is a block diagram showing a component 400 that may be used in a system employing the disclosed technology in some implementations. Component 400 may be used to store, manage, analyze, and access medical data in a digital filing cabinet. Component 400 includes hardware 402, general software 420, and specialized components 440. As previously stated, a system implementing the disclosed technology may use a variety of hardware, including a processing unit 404 (e.g., CPU, GPU, APU), working memory 406, storage memory 408 (as an interface to local storage or remote storage such as storage 315 or 325), and input and output devices 410. In various implementations, storage memory 408 may be a local device, an interface to a remote storage device, or one or more of these. For example, the storage memory 408 may be a set of one or more hard drives (e.g., a redundant array of independent disks (RAID)) accessible via the system bus, or it may be a cloud storage provider or other network storage (e.g., a network-accessible storage (NAS) device, such as storage provided via storage 315 or another server 320) accessible via one or more communication networks. Component 400 may be implemented within a client computing device such as client computing device 305, or on a server computing device such as server computing device 310 or 320.

[0093] General software 420 may include a variety of applications, including an operating system 422, a local program 424, and a basic input-output system (BIOS) 426. Special components 440 may be subcomponents of general software applications 420, such as the local program 424. Special components 440 may be for providing patient-specific medical data and may include a machine learning module 444, a threshold module 446, a scoring module 448, a confirmation action module 450, a reporting module 452, and components that can be used to provide a user interface, transfer data, and control special components such as interface 442 (e.g., a user interface on a tablet, smartphone, laptop, etc.). In some implementations, components 400 may reside in a computing system distributed across multiple computing devices, or may be an interface to an application based on a server running one or more of the special components 440. Although the special components 440 are shown as separate components, they may be logical distinctions of function or other non-physical distinctions, and / or submodules or code blocks of one or more applications.

[0094] The machine learning module 444 may be configured to analyze a patient's preoperative data (e.g., images such as MRI, CT, CAT, or X-rays, and / or implant data such as implant design files or manufactured implants) to identify when it is appropriate to recommend a human review before implanting the implant to the patient. The machine learning module 444 may be configured to analyze the preoperative data based on at least one machine learning algorithm trained on at least one dataset reflecting approved preoperative data. At least one machine learning algorithm (and model) may be stored locally in a database and / or outside the database (e.g., a cloud database and / or cloud server). A client device may have access to these machine learning algorithms, intelligently analyze the preoperative data, and determine whether a human should review the preoperative data before implantation, based on at least one machine learning model trained on previously approved preoperative data. For example, the machine learning module 444 may perform an analysis of the preoperative data to determine whether the data includes edge-case medical conditions (e.g., tumors, fractures, transitional vertebrae, etc.) and recommend that the data be reviewed by a human. For example, if a patient has thinned vertebrae (e.g., below threshold volume), deformed vertebrae, low-density tissue (e.g., bone, soft tissue, etc.), a tumor within threshold distance from the implant site, a fractured intervertebral disc, or a transitional vertebra, the machine learning module 444 can determine that a human review is required.

[0095] As described herein, a machine-learning (ML) model may mean a predictive or statistical utility or program that can be used to determine a probability distribution over one or more sequences of characters, classes, objects, result sets, or events, and / or to predict a response value from one or more predictors. The model may be based on or incorporate one or more rule sets, machine learning, neural networks, etc. In the example, the ML model may be deployed on a client device, a service device, a network appliance (e.g., a firewall, a router, etc.), or any combination thereof. The ML model may process preoperative data and other data stores of user health indicators to determine when it is appropriate to recommend a human review of the preoperative data before implanting an implant in a patient. Based on the aggregation of data from the user's medical digital filing cabinet, the healthcare provider's preoperative data storage, and other user data stores, at least one ML model may be trained and then deployed to automatically analyze the preoperative data and determine whether the preoperative data contains any edge-case medical conditions that would cause a recommendation for a human review of the preoperative data. A trained ML model may be deployed to one or more devices. For example, instances of a trained ML model may be deployed to a server device and to client devices. An ML model deployed to a server device may be configured, for example, to be used by the client device when the client device is connected to the internet. Conversely, an ML model deployed to a client device may be configured, for example, to be used by the client device when the client device is not connected to the internet. In some examples, the client device may not be connected to the internet but may still be configured to receive satellite signals containing medical data. In such examples, the ML model may be cached locally by the client device.In some implementations, the machine learning module 444 identifies newly added preoperative data in the digital filing cabinet and analyzes the new preoperative data to determine whether the data contains any edge-case medical conditions that require human review.

[0096] Threshold module 446 may be configured to determine edge-case condition thresholds for machine learning module 444 to analyze preoperative data. Thresholds may be based on parameters such as modifications to intervertebral fusion, anatomical abnormalities (e.g., hemivertebrae, grade 2+ spondylolisthesis), osteotomy symptoms, implant atypical symptoms, tumors in the patient (e.g., tumors located near the implant site), vertebral or disc fractures in the patient, torn or partially torn ligaments, tendons, or muscles, implants at adjacent levels of the patient's spine (e.g., fixation mass at the level being operated on), transition vertebrae (e.g., extra vertebrae), or any existing patient conditions that may affect implant placement in the patient. Edge-case condition thresholds are determined so that if the analyzed preoperative data includes a threshold, a human (e.g., healthcare provider, implant manufacturer, implant designer, etc.) is consulted to approve implant placement in the patient. In some implementations, a human review threshold is determined by the healthcare provider to achieve the patient's modified biostructure. The threshold module 446 may be configured to determine (e.g., calculate, retrieve, etc.) a threshold for each parameter or group of parameters (e.g., an aggregated group of parameters). In the first example, if the patient has previously undergone spinal fusion surgery, the threshold may be the distance from the implant site to the fusion level. In the second example, in the case of a rare anatomical anomaly (e.g., an anomaly not adequately represented in the machine learning training set), the threshold may be the amount of deformity in the spine (e.g., 10%, 25%, etc.) or the amount of movement in the spine. In the third example, if the patient information indicates symptoms of a previous osteotomy, the threshold may be the likelihood that the procedure occurred, or the threshold may be the distance from the implant where the osteotomy occurred. In the fourth example, the threshold may be the distance from the implant site to other implanted implants. In the fifth example, if a tumor is detected, the threshold may be the size of the tumor or the distance from the implant site to the tumor.In the sixth example, if a fracture (e.g., an intervertebral disc fracture, a fracture along the vertebrae in any part of the spine) is detected, the threshold can be the size of the fracture, or the threshold can be the distance from the implant to the fractured portion of the spine. In the seventh example, if a transitional vertebra is detected, the threshold can be the distance from the implant to the transitional vertebra.

[0097] The threshold module 446 may be configured to determine an edge case condition threshold based on one or more of the following: statistical analysis, statistical parameters (e.g., numerical values ​​of standard deviation), edge case criteria (e.g., a condition present in less than a percentage such as 1%, 3%, or 5% of a population, reference dataset, etc.), anomalies based on previous patient image analysis, predefined edge case parameters, or a combination thereof. In some embodiments, the threshold module 446 can detect anomalies in preoperative images. The threshold module 446 may search a training set, reference images (e.g., previous patient images), an image and condition database, or other databases to identify anomalies. One or more possible anomalies that affect treatment prediction may be identified. In some embodiments, the edge case criteria may include occurrence thresholds (e.g., less than 1%, less than 2%, less than 5%, less than 10%, less than 15%, etc.) for qualifying as an edge case. The threshold module 446 may determine the incidence of anomalies from available data (e.g., patient examinations, clinical trials, research publications, etc.). This allows the threshold module 446 to identify unknown anomalies as edge cases without accessing specific occurrences of the anomalies. In some embodiments, the threshold module 446 can adaptively identify edge cases without identifying previously matching edge cases. In some embodiments, the threshold module 446 can identify edge cases by matching the current anomaly with known edge cases. The machine learning algorithm can determine thresholds based on analyzed datasets, including training sets, pre-operative datasets, and post-operative datasets. For example, the machine learning algorithm can associate post-operative data with pre-operative parameters to determine whether the pre-operative parameters qualify as edge cases.

[0098] The scoring module 448 may be configured to determine a score based on preoperative data analyzed (for example, by the machine learning module 444). The scoring module 448 may select parameters from the threshold module 446 to determine the score of the preoperative data. If the score reaches a threshold (e.g., an edge-case condition threshold such as a minimum or maximum value), the preoperative data is recommended for human review. The score may indicate confidence that the implant will correct the patient's biostructure. For example, a confidence score indicating that the implant will correct a deformity of the patient's spine. The scoring module 448 may be configured to determine a score based on other data. For example, the scoring module 448 may be configured to determine a score based on an analysis of one or more hypothetical models of the patient's biostructure, a predicted postoperative model, physician input, or a combination thereof. Edge-case condition determination may be based on both preoperative and postoperative data. For example, if a parameter is determined not to affect the postoperative prediction, that parameter is excluded from contributing to edge-case determination. The system can update and modify treatments to limit, prevent, or significantly reduce the therapeutic effect of candidate edge-case parameters. In some embodiments, the scoring module 448 can identify potential edge-case parameters based on identified areas of the body. The scoring module 448 can identify potential parameters of a treatment area based on a series of procedures, such as fusion, decompression, or other spinal surgery. Each procedure may be associated with parameters that may affect that procedure. To provide edge-case scoring based on procedures, procedures and parameters may be grouped and scored. For some procedures, a patient may identify certain procedures as edge-case procedures, while others may not. This allows the system to select procedures based on edge-case conditions with respect to specific procedures.

[0099] The confirmation action module 450 may be configured to use additional imaging (MRI, CT, CAT, or X-ray) to confirm or develop an alternative plan if the preoperative data is determined to include an edge case condition. The confirmation action module 450 may notify the healthcare provider to re-examine the preoperative data or request additional imaging if the score determined by the scoring module 448 exceeds a threshold (e.g., a high-risk edge case). In some implementations, the confirmation action module 450 generates automated requests for condition-specific diagnostic information (e.g., spinal curvature, cancer, tumor, anatomical abnormalities, fractures, implanted implants, etc.) based on the type of implant or target outcome for the patient.

[0100] The confirmation action module 450 can automatically execute postoperative processes, such as scheduling postoperative examinations for patients to acquire postoperative data (e.g., scans or images of the medical condition). The confirmation action module 450 can automatically analyze the patient's medical condition and outcomes after surgery and verify preoperative analysis, postoperative predictions, etc. In some embodiments, the system can send notifications to patients for follow-up examinations. For example, the system can set up automatic payments for follow-up visits for patients to facilitate the collection of postoperative data. Examinations can be scheduled at predetermined times, such as one week or more, one month or more, and / or one year or more after treatment. Post-follow-up schedules can be generated based on surgical procedures, physician input, etc. In some embodiments, the system can automatically acquire data stored behind one or more firewalls, data via hospital record systems, data in metadata, etc. The system can automatically determine search queries to acquire information and can stop searching the database when the target information has been acquired.

[0101] The reporting module 452 may be configured to generate and transmit a report of preoperative data for human review to the device. This report may include automated annotations (e.g., boxes containing text indicating areas to be reviewed), automated requests for optional re-imaging (e.g., regions of interest), and / or automated labeling of annotations and spinal-pelvic parameters.

[0102] Figure 5 is a flowchart illustrating a process 500 for identifying the pathology of an edge case for inspection, according to one or more embodiments of the present technology.

[0103] In step 502, process 500 receives a patient dataset relating to a specific patient requiring treatment. The patient dataset may include data representing the patient's condition, biostructure, medical condition, symptoms, medical history, preferences, and / or any other information or parameters related to the patient. For example, the patient dataset may include surgical intervention data, treatment outcome data, progress data (e.g., surgeon's notes), patient feedback (e.g., quality of life questionnaires, feedback obtained using surveys), clinical data, patient information (e.g., demographics, sex, age, height, weight, type of medical condition, occupation, activity level, tissue information, health assessment, comorbidities, health-related quality of life (HRQL)), vital signs, diagnostic results, drug information, allergies, and diagnostic device information (e.g., manufacturer, model number, specifications, user-selected settings / configurations, etc.). The patient dataset may also include image data such as camera images, MRI images, ultrasound images, CAT scan images, PET images, and X-ray images. In some embodiments, the patient dataset includes data representing one or more of the following: patient identification number (ID), age, sex, BMI, LL, Cobb angle, PI, intervertebral disc height, segmental flexibility, bone quality, rotational displacement, and / or spinal treatment level. The patient dataset may be received by a server, computing device, or other computing system. For example, in some embodiments, the patient dataset may be received by server 106 shown in Figure 1. In some embodiments, the computing system that receives the patient dataset in step 502 also stores one or more software modules (e.g., a data analysis module and / or a treatment planning module, or additional software modules for performing various operations of process 500).

[0104] In some embodiments, the received patient dataset may include disease indicators such as LL, Cobb angle, coronal parameters (e.g., coronal balance, overall coronal balance, coronal-pelvic tilt, etc.), sagittal parameters (e.g., PI, sacral tilt, thoracic kyphosis, etc.), and / or pelvic parameters. Disease indicators may include micromeasurements (e.g., indicators related to specific or individual segments of the patient's spine) and / or macromeasurements (e.g., indicators related to multiple segments of the patient's spine). In some embodiments, disease indicators may not be included in the patient dataset, and process 500 may include determining one or more of the disease indicators based on the patient image data (e.g., automatically determining them).

[0105] In step 504, process 500 creates a virtual model of the patient's innate anatomical structure (also called “preoperative anatomical structure”). The virtual model may be based on image data contained in the patient dataset received in step 502. For example, the same computing system that received the patient dataset in step 502 may analyze the image data in the patient dataset to generate a virtual model of the patient's innate anatomical structure. The virtual model may be a two-dimensional or three-dimensional visual representation of the patient's innate biological structure. The virtual model may include one or more regions of interest and may include some or all of the patient's biological structures within the region of interest (e.g., any combination of tissue types including, but not limited to, bone structures, cartilage, soft tissue, vascular tissue, nerve tissue, etc.). As a non-limiting example, the virtual model may include a visual representation of the patient’s spinal region, including some or all of the sacrum, lumbar region, thoracic region, and / or cervical region. In some embodiments, the virtual model includes soft tissue, cartilage, and other non-bone structures. In other embodiments, the virtual model includes only the patient’s bone structures. Examples of virtual models of the innate anatomical structure are described below with respect to Figures 8A and 8B. In some embodiments, process 500 may optionally omit the creation of a virtual model of the patient's innate biological structure in step 504 and proceed directly from steps 502 to 506.

[0106] In some embodiments, the computing system that generates the virtual model may also determine (e.g., automatically determine or measure) one or more disease indicators of the patient based on the virtual model. For example, the computing system may analyze the virtual model to determine the patient's preoperative LL, Cobb angle, coronal parameters (e.g., coronal balance, overall coronal balance, coronal-pelvic tilt, etc.), sagittal parameters (e.g., PI, sacral tilt, thoracic kyphosis, etc.), and / or pelvic parameters. The disease indicators may include micro-measurements (e.g., indicators related to specific or individual segments of the patient's spine) and / or macro-measurements (e.g., indicators related to multiple segments of the patient's spine).

[0107] In step 506, process 500 creates a virtual model of the patient's modified anatomical configuration (which may also be referred to herein as the “planned configuration,” “optimized shape,” “postoperative anatomical configuration,” or “target outcome”). For example, a computing system may use the aforementioned analysis procedure to determine a “modified” or “optimized” anatomical configuration of a particular patient that represents an ideal surgical outcome for that particular patient. This may be done, for example, by analyzing multiple reference patient datasets to identify the postoperative anatomical configurations of similar patients who had beneficial postoperative outcomes (based on, for example, the similarity of the reference patient datasets to the patient datasets, and / or whether the reference patients had beneficial treatment outcomes). This may also include applying one or more mathematical rules that define the optimal anatomical outcome (e.g., positional relationships between anatomical elements) and / or target (e.g., acceptable) postoperative indicators / design criteria (e.g., adjusting the biostructure so that the postoperative sagittal vertical axis is less than 7 mm, the postoperative Cobb angle is less than 10 degrees, etc.). Target postoperative indicators may include, but are not limited to, target coronal parameters, target sagittal parameters, target PI angle, target Cobb angle, target shoulder inclination, target iliopsoas angle, target coronal balance, target Cobb angle, target lordosis angle, and / or target intervertebral space height. Differences between the innate anatomical configuration and the modified anatomical configuration may be referred to as “patient-specific modifications” or “target modifications.”

[0108] After the modified anatomical structure has been determined, the computing system can use the modified anatomical structure to generate a two-dimensional or three-dimensional visual representation of the patient's biostructure. Similar to the virtual model created in step 504, the virtual model of the patient's modified anatomical structure may include one or more regions of interest and may include some or all of the patient's biostructure within the region of interest (e.g., any combination of tissue types including, but not limited to, bone structure, cartilage, soft tissue, vascular tissue, nerve tissue, etc.). As a non-limiting example, the virtual model may include a visual representation of the patient's spinal region in the modified anatomical structure, including some or all of the sacrum, lumbar region, thoracic region, and / or cervical region. In some embodiments, the virtual model includes soft tissue, cartilage, and other non-bone structures. In other embodiments, the virtual model includes only the patient's bone structure. Examples of virtual models of the innate anatomical structure are described below with respect to Figures 9A-1 to 9B-2.

[0109] In some implementations, process 500 generates a surgical plan to achieve the modified anatomical configuration shown by the virtual model. The surgical plan may include a preoperative plan, surgical plan, postoperative plan, and / or specific spinal indices relevant to the optimal surgical outcome. For example, the surgical plan may include specific surgical procedures to achieve the modified anatomical configuration. In the context of spinal surgery, the surgical plan may include specific fusion procedures (e.g., PLIF, ALIF, TLIF, LLIF, DLIF, XLIF, etc.) across a specific range of spinal levels (e.g., L1–L4, L1–L5, L3–T12, etc.). Naturally, other surgical procedures to achieve the modified anatomical configuration may be identified, such as non-fusion surgical techniques and orthopedic procedures for other areas of the patient. The surgical plan may also include one or more expected spinal indices (e.g., LL, Cobb angle, coronal parameter, sagittal parameter, and / or pelvic parameter) corresponding to the expected postoperative biostructure of the patient. The surgical plan may be generated by the same or a different computing system that created the virtual model of the modified anatomical configuration. In some embodiments, the surgical plan may also be based, at least in part, on the surgeon's specific preferences and / or outcomes associated with the particular surgeon performing the operation. In some embodiments, two or more surgical plans are generated to provide the surgeon with multiple options. An example of a surgical plan is described below with reference to Figure 10.

[0110] In step 508, process 500 (optionally) designs a patient-specific implant based on the modified anatomical configuration and / or surgical plan (e.g., via the same computing system used to perform steps 502–506). For example, a patient-specific implant may be specifically designed to instruct the patient's biostructure to occupy the modified anatomical configuration when implanted in a particular patient (e.g., converting the patient's biostructure from its original anatomical configuration to the modified anatomical configuration). A patient-specific implant may be designed to cause the patient's biostructure to occupy the modified anatomical configuration for the expected lifespan of the implant (e.g., 5 years or more, 10 years or more, 20 years or more, 50 years or more, etc.) when implanted. In some embodiments, the patient-specific implant is designed based solely on a virtual model of the modified anatomical configuration and / or without reference to preoperative patient images.

[0111] A patient-specific implant may be any implant described herein or any patent reference incorporated herein by reference. For example, a patient-specific implant may include one or more of the following: screws (e.g., bone screws, spinal screws, pedicle screws, small articular surface screws), intervertebral implant devices (e.g., intervertebral implants), cages, plates, rods, intervertebral discs, fixation devices, spacers, rods, expandable devices, stents, brackets, cords, skeletons, fixation devices, anchors, nuts, bolts, rivets, connectors, ropes, fasteners, joint replacements (e.g., artificial intervertebral discs), gluteal implants, etc. A patient-specific implant design may include data representing one or more of the implant's physical properties (e.g., size, shape, volume, material, mass, weight), mechanical properties (e.g., stiffness, strength, modulus of elasticity, hardness), and / or biological properties (e.g., osteointegration, cell adhesion, antimicrobial properties, antiviral properties). For example, the design of an orthopedic implant may include the shape, size, material, and / or effective stiffness of the implant (e.g., lattice density, number of struts, strut position, etc.). Below, with reference to Figures 12A and 12B, an example of a patient-specific implant designed by process 500 is described.

[0112] In some embodiments, patient-specific implants are designed in step 508 only after the surgeon has reviewed and approved a virtual model and surgical plan with a modified anatomical configuration. Thus, in some embodiments, the implant design is not sent to the surgeon along with the surgical plan, nor is it manufactured before receiving the surgeon's approval of the surgical plan. Without being bound by theory, waiting for the surgeon to approve the surgical plan before designing patient-specific implants may improve the efficiency of process 500 and / or reduce the resources required to perform process 500.

[0113] In step 510, process 500 analyzes a virtual model of the patient's innate anatomical structure and / or a virtual model of a modified anatomical structure to identify edge-case conditions that may affect implant placement. For example, process 500 may scan preoperative data (e.g., images such as MRI, CT, CAT, or X-rays, and / or implant data such as implant design files or manufactured implants) to identify edge-case conditions that may require human review.

[0114] In step 512, process 500 models edge-case conditions using a virtual model. Process 500 can detect edge-case conditions based on parameters such as modifications to intervertebral fusion, anatomical abnormalities (e.g., hemivertebrae, grade 2+ spondylolisthesis), osteotomy symptoms, unusual symptoms of implants, tumors in the patient (e.g., tumors located near the implant site may be affected by implant placement), fractures in the patient, torn or partially torn ligaments, tendons, or muscles, implants at adjacent levels of the patient's spine (e.g., fixation mass at the surgical level), transitional vertebrae (e.g., extra vertebrae), or any existing patient conditions that may affect implant placement in the patient. Process 500 can identify edge-case conditions by comparing patient data with multiple reference patient datasets that have identified edge-case conditions. In some cases, edge-case conditions are detected when patient data includes anatomical abnormalities. Process 500 can identify edge-case conditions by digitally analyzing patient datasets. Process 500 can segment patient images to identify anatomical features in the patient's biological structure. When identifying anatomical features, Process 500 can use a trained machine learning algorithm to determine whether the identified features qualify as edge features.

[0115] Process 500 can determine thresholds for each parameter / feature. In the first example, if the patient has undergone interbody fusion of the spine, the threshold may be the distance from the implant site to the removed intervertebral disc. In the second example, in the case of an anatomical abnormality, the threshold may be the amount of deformation in the spine (e.g., 10%, 25%, etc.) or the amount of movement in the spine. In the third example, if the patient information includes symptoms of having undergone an osteotomy, the threshold may be the symptom that the procedure has occurred, or the threshold may be the distance from the implant where the osteotomy occurred. In the fourth example, the threshold may be the distance from the implant site to another implanted implant. In the fifth example, if a tumor is detected, the threshold may be the size of the tumor or the distance from the implant to the tumor. In the sixth example, if a fracture is detected in the intervertebral disc, vertebra, or any part of the spine, the threshold may be the size of the fracture, or the threshold may be the distance from the implant to the fractured portion of the spine. In the seventh example, if a transitional vertebra is detected, the threshold may be the distance from the implant to the transitional vertebra.

[0116] In some implementations, process 500 identifies edge case conditions by scoring the features of one or more candidate edges in the patient dataset. Based on the scoring, process 500 determines an aggregate of edge feature scores and compares the aggregate of edge feature scores to a threshold for edge case conditions to identify at least one edge case condition.

[0117] In step 514, process 500 generates a patient-linked manufacturing hold to prevent the implant design platform from triggering implant manufacturing for the patient when the pathological condition of the edge case is identified. The manufacturing hold may include a design lock to prevent the implant design platform from designing any implants for the patient, a manufacturing data lock to prevent the generation of implant manufacturing data, and / or a transmission lock to prevent the transmission of implant design data to the manufacturing equipment. The number and types of locks may be selected based on other steps in process 500.

[0118] In some embodiments, process 500 may include sending a request from the implant design platform to reconsider the condition of at least one identified edge case. Upon receiving input from the user device regarding the requested reconsideration of the condition of at least one identified edge case, process 500 may, via the implant design platform, determine whether to remove a manufacturing hold based on the received input to enable the automated manufacturing of one or more patient-specific implants for the patient. For example, process 500 may decide to remove a manufacturing hold based on whether the received input meets the patient's treatment threshold. If the received input meets the treatment continuation threshold, process 500 generates a patient-specific implant design (e.g., according to step 508) and sends manufacturing data for manufacturing the patient-specific implant according to the design.

[0119] If an edge case condition is identified, process 500 can send a notification to a human (e.g., a healthcare provider, implant designer, etc.) to review the patient information. In some cases, if any of the parameters are identified in the patient information, process 500 sends a notification to a human to review the patient information. For example, if a tumor is located near the implant site, the physician is notified to verify that the implant placement will not cause the tumor to rupture or become inflamed. Process 500 can display a patient image with annotated edge case parameters via a user interface and can adjust the display of the patient image according to display input from the user reviewing the patient image. Process 500 can dynamically display information on edge case parameters that are viewable and / or selected by the user.

[0120] In some implementations, Process 500 utilizes machine learning to identify edge case morbidities. The machine learning model is retrained using edge case morbidity data identified based on human review results. For example, the machine learning model can determine an edge case morbidity score based on at least one characteristic / parameter of at least one edge case morbidity. This allows for scaling of Process 500 by improving the accuracy of edge case analysis.

[0121] Figure 6A is a flowchart illustrating a process 600 for recommending a human review of a patient-specific implant procedure according to one or more embodiments of the present technology.

[0122] In step 602, process 600 determines a score for the identified edge case malformation (from steps 510 and 512 in Figure 5). In some implementations, the score is based on parameters / characteristics associated with the identified edge case malformation and / or whether the identified edge case malformation prevents the implant from modifying the patient's biostructure. For example, a hypothetical model of the modified anatomical structure (described in step 506 in Figure 5) can show how the identified edge case malformation affects the placement, orientation, and / or function of the implant. The score can indicate confidence that the identified edge case malformation does not affect the implant in modifying the patient's biostructure (e.g., the scored treatment outcome falls below a threshold). Process 600 may determine the score based on thresholds associated with the identified edge case malformation (as described in steps 510 and 512). For example, if the tumor is within the vicinity threshold of the implant site, the score will be lower than if the tumor is outside the vicinity threshold of the implant site. In some implementations, the score is based on comparing the results of different machine learning algorithms to determine whether the identified edge-case medical conditions prevent the implant from modifying the patient's biostructure.

[0123] Process 600 can determine edge-case malformation scores based on scored treatment outcomes of patient-specific orthopedic implants addressing the patient's spinal condition. Scored treatment outcomes may include data representing modified anatomical indicators, presence of fusion, health-related quality of life, activity levels, and / or at least one comorbidity. In some cases, the score represents a statistical correlation between the patient dataset and at least one of several reference patients. Process 600 can dynamically calculate anatomical indicators of a modified anatomical configuration, generated based on anatomical adjustments, the number of implants, or implant configurations entered by the user. Process 600 can display the calculated anatomical indicators and annotate one or more of the calculated anatomical indicators indicating edge-case malformations. In some implementations, Process 600 determines scores by classifying one or more parameters in the patient data and performing edge-case scoring of one or more parameters based on this classification. Based on the classification and edge-case scoring, Process 600 can determine multiple edge-case malformations.

[0124] In step 604, process 600 determines whether the score exceeds a threshold. This threshold may indicate that a condition in any edge case has been identified, or that the condition in the identified edge case requires human review. If the score does not exceed the threshold, in step 616, process 600 sends the user an approval notification regarding the placement of a patient-specific implant in the patient.

[0125] If the score exceeds a threshold, process 600 sends a request to the device for a human to examine the edge case physiology in the patient dataset before the system triggers the manufacture of one or more patient-specific implants for the patient. The request may include automated annotations (e.g., labels, highlighting, etc.) of at least one edge case physiology in the patient dataset. In some cases, process 600 generates an automated request for diagnostic information specific to the physiology (e.g., the target treatment outcome by implanting a patient-specific orthopedic implant in the patient).

[0126] In step 606, process 600 generates a review plan that requests a human review of the patient data. The review plan may include, but is not limited to, edge-case conditions, identified anatomical anomalies, labeled scans (e.g., labeled digital images), statistics, correction routines, optimization routines, reports, automated annotations (e.g., boxes with text indicating areas for review), automated requests for re-imaging of regions of interest (e.g., MRI, CT, CAT, X-ray, etc.), and automated labeling of annotations and spinal-pelvic parameters. Labeled scans may be labeled raw images, labeled pre-processed images, labeled processed images, etc. Labels may be boxes, highlights, annotations, or other types of indicators for identifying features (e.g., identified edge-case conditions) or areas for visual inspection, automated inspection, etc. Labels may include anatomical descriptive labels.

[0127] In some implementations, process 600 may include identifying edge-case malformations in a patient by applying one or more image processing algorithms to at least one digital image of the patient. Process 600 may generate an edge-case malformation review plan that includes annotated images of the patient and edge-case identification information. Based on review feedback from the edge-case malformation review plan, process 600 may design an implant using a virtual model of the patient's biostructure. The edge-case malformation review plan may include machine-executable instructions for machine learning analysis to determine whether to proceed with the implant design process for the patient, computed anatomical indices, and labeled images to associate one or more of the computed anatomical indices with anatomical features indicating edge-case malformations.

[0128] Process 600 may include determining whether an identified area or region is a non-conformity feature that meets the patient's non-conformity risk threshold (e.g., an edge-case medical condition that may be affected by the implant). In response to determining that the identified non-conformity feature meets the non-conformity risk threshold, Process 600 may generate a patient-specific orthopedic implant non-conformity report (e.g., a review plan). The non-conformity report may include, for example, virtual model data to display the fit of a virtual model of the implant to one or more anatomical features of the patient. The user can examine the implant fit to determine whether to change the implant, approve the implant, reject the implant, etc. The non-conformity report may include predicted postoperative patient indicators, labeling of a single non-conformity feature, etc. For example, labels may be applied to each identified non-conformity feature along with relevant postoperative patient indicators resulting from the non-conformity feature (e.g., anatomical changes resulting from implant deformation or incorrect sizing features). Labels may include one or more of the following indicators for identifying the feature or area, for example, for visual and / or automated inspection: boxes, highlights, annotations, or other types of indicators. Labeled non-conformity features and predicted postoperative patient indicators may be displayed (e.g., via display 230, client computing devices 305A-D, interface 442, etc. in Figure 2) to facilitate the review and approval process. The labeling process may include applying different labels to different non-conformity features. If an updated surgical plan is generated, the updated surgical plan can identify the impact of the non-conformity features for human review and / or automated review. Process 600 may send a request to the device to capture one or more images containing edge-case pathologies on or near the patient's spine.

[0129] In step 608, process 600 displays the reconsideration plan via a user interface on the device's electronic screen. This interface may include a patient data window displaying one or more annotated patient images and a simulation window displaying at least one implant simulation labeling one or more anatomical parameters that contribute to identifying the edge case condition. In some implementations, process 600 displays at least one patient image containing one or more annotated edge case parameters and adjusts the display of at least one patient image according to display input from the user. Process 600 can dynamically display information on one or more parameters that are viewable by the user and / or selected by the user. In some implementations, the edge case condition reconsideration plan causes the user interface to display at least one patient image containing one or more annotated edge case parameters and adjusts the display of at least one patient image according to display input from the user. Process 600 can dynamically display information on one or more identifiable parameters that can be viewed by the user.

[0130] In step 610, process 600 receives input from human review. For example, a user can input feedback into the user interface. After receiving input from human review, process 600 can send the implant design to the additive manufacturing apparatus for manufacturing a patient-specific orthopedic implant by additive manufacturing according to the implant design. After receiving human input, process 600 determines, based on the received input, whether to remove the manufacturing hold (e.g., via the implant design platform) to enable automated manufacturing of one or more patient-specific implants for the patient. In response that the received input meets the treatment continuation threshold, process 600 can generate designs for one or more patient-specific implants and send manufacturing data for manufacturing the patient-specific implant according to this design.

[0131] Process 600 receives user input regarding anatomical adjustments, the number of implants, or at least one of the implants, and can modify the anatomical configuration according to an edge-case pathology re-examination plan. Process 600 can determine one or more anatomical indicators within the set of modified anatomical configurations and display the determined one or more anatomical indicators for edge-case analysis. Based on the input received from the user, Process 600 can design a patient-specific implant for the patient, enable one or more design and manufacturing steps to be performed based on the received input, and / or synchronize one or more implant design and manufacturing steps with human re-examination to develop a surgical plan that compensates for the edge-case pathology and virtual models of the implants. In some implementations, Process 600 receives edge-case compensation / modification from human re-examination and determines the surgical plan based on the edge-case compensation / modification.

[0132] In step 612, process 600, after reviewing the medical condition of the identified edge case, determines whether a human has approved the surgery to implant the implant in the patient. If the implant procedure has not been approved by a human, in step 618, process 600 determines whether to make adjustments to the implant or refuse to approve the surgery. Process 600 may determine adjustments to the design of the implant so that the implant does not affect the identified medical condition. For example, the size of the implant may be adjusted so that the implant does not interfere with fracture healing. In some implementations, a computing system may generate one or more reports and send them to the manufacturer, the patient, and / or the healthcare provider. These reports may include one or more of the following: patient data, implant images / scans, anatomical images, identification of anomalies in the patient data, characteristics of implant incompatibility, surgical data, patient data, machine data, adjustments required so that the implant does not affect the medical condition of the identified edge case, modified manufacturing instructions, or any data related to the implant or patient.

[0133] If a human provides approval for the implant procedure, in step 614, process 600 approves the procedure for implanting the implant in the patient. Process 600 may generate and send a notification indicating that the identified edge case medical condition does not affect the implant modifying the patient's biostructure. In response to receiving input from human review (e.g., approval), process 600 may remove the manufacturing hold to allow the submission of the implant design for the additive manufacturing device. For example, process 600 may decide to remove the manufacturing hold based on whether the received input meets the patient's treatment threshold.

[0134] The computing system can train one or more machine learning models based at least in part on the results of human review of identified edge case medical conditions, and then use the newly trained machine learning models to perform analysis of patient data to identify edge case medical conditions and determine whether human review is required. In some embodiments, process 600 can transmit a suitable amount of identified data to the machine learning models. The computing system can determine the amount and type of information to transmit to the training system to train one or more machine learning models for future patient data examinations, etc. The computing system can analyze patient data and / or implant data to determine whether human review is required at any stage of the patient-specific orthopedic implant (e.g., design, manufacturing, implant treatment, or post-implant treatment). The computing system can retrain the machine learning models by comparing postoperative images with predicted results. For example, postoperative images can be used to train the machine learning models to show how the implant affected the identified edge case medical conditions and determine whether human review is required.

[0135] Figure 6B provides a series of images illustrating an example of an edge-case condition report 650, which includes a review plan 652 (described in Figure 6A) and can be sent to a user (e.g., a healthcare provider, implant designer, manufacturer, etc.) for review and approval. The edge-case condition report 650 may include a multi-page report detailing aspects of the review plan 652. A device (e.g., a client computing device 102 shown in Figure 1) can display the report 650 via a user interface on an electronic screen. This interface may include a patient data window displaying one or more annotated patient images and a simulation window displaying at least one implant simulation labeling one or more anatomical parameters (e.g., shown in the pre-operative window 653 and the post-operative window 656). In some implementations, the report 650 displays a patient image with one or more annotated edge-case parameters, and the user can adjust the display of the patient image according to user display input. The device can dynamically display information on one or more parameters that are viewable by the user and / or selected by the user. In some implementations, the edge case condition review plan 652 causes the user interface to display at least one patient image containing one or more annotated edge case parameters and to adjust the display of at least one patient image in accordance with display input from the user. In some embodiments, the report 650 is interactive, and the user can manipulate various aspects of the report 650 (e.g., adjust, zoom in, zoom out, annotate, etc., the display of a virtual model).

[0136] For example, a multi-page report may include labeled and identified edge-case conditions (e.g., tumors or fractures shown in box 654) in images or models of the patient's preoperative biostructure in the preoperative window 653. Reference number 1 identifies a tumor, and reference number 2 identifies a vertebral fracture. The postoperative window 656 shows the planned outcome, including implants and the labeled tumors and fractures. The user can manipulate (e.g., zoom) the images to view and evaluate the biostructure and planned outcome. The edge-case condition report 650 provides some degree of review, designed to confirm effective treatment for various types of rare conditions, thereby enabling safe and effective treatment for larger groups or populations of patients.

[0137] Figures 7A–13 further illustrate selected aspects of providing patient-specific medical care according to, for example, processes 500 and / or 600. For example, Figures 7A–7D show examples of patient datasets 700 (such as those received in step 502 of process 500). The patient dataset 700 may include any of the information previously described with respect to patient datasets. For example, the patient dataset 700 may include patient information 701 (e.g., patient identification number, patient MRN, patient name, sex, age, BMI, date of surgery, surgeon, etc., as shown in Figures 7A and 7B), diagnostic information 702 (e.g., Oswestry Disability Index (ODI), VAS back score, VAS leg score, preoperative pelvic intrinsic angle, preoperative lumbar lordosis, preoperative PI-LL angle, preoperative lumbar coronal cove, etc., as shown in Figures 7B and 7C), and imaging data 703 (e.g., X-ray, CT, MRI, etc., as shown in Figure 7D). In the embodiments shown, the patient dataset 700 is collected by a healthcare provider (e.g., a surgeon, nurse, etc.) using digital reports and / or fillable reports that can be accessed using a computing device. In some embodiments, the patient dataset 700 may be automatically or at least partially automatically generated based on the patient's digital medical records. In any case, after collection, the patient dataset 700 may be sent to a computing system configured to generate a surgical plan for the patient.

[0138] Figures 8A and 8B show examples of virtual models 800 of a patient's innate anatomical structure (such as those created in steps 504 / 506 of process 500). In detail, Figure 8A is a magnified view of the virtual model 800 of the patient's innate biostructure, showing the innate biostructure of the patient's lower spinal cord region. The virtual model 800 is a three-dimensional visual representation of the patient's innate biostructure. In the embodiments shown, the virtual model includes a portion of the spine extending from the sacrum to the L4 vertebral level. Naturally, the virtual model may include other regions of the patient's spine, including the cervical, thoracic, lumbar, and sacrum. While the virtual model 800 shown includes only the skeletal structure of the patient's biostructure, other embodiments may include additional structures such as cartilage, soft tissue, vascular tissue, and nerve tissue.

[0139] Figure 8B shows a virtual model display 850 (sometimes referred to herein as “Display 850”) showing a different view of the virtual model 800. The virtual model display 850 includes three-dimensional views of the virtual model 800, one or more coronal sections 802 of the virtual model 800, one or more axial sections 804 of the virtual model 800, and / or one or more sagittal sections 806 of the virtual model 800. Naturally, other views are possible and may be included in the virtual model display 850. In some embodiments, the virtual model 800 may be interactive, allowing the user to manipulate the orientation or viewpoint of the virtual model 800 (e.g., rotate it), change the depth of the displayed sections, select and isolate specific bone structures, etc.

[0140] Figures 9A-1 to 9B-2 show examples of virtual models of the patient's innate anatomical structure (e.g., as created in step 504 of process 500) and virtual models of the patient's modified anatomical structure (e.g., as created in step 506 of process 500). In detail, Figures 9A-1 and 9A-2 are frontal and lateral views, respectively, of virtual model 910 showing the patient's innate anatomical structure, and Figures 9B-1 and 9B-2 are frontal and lateral views, respectively, of virtual model 920 showing the same patient's modified anatomical structure. Referring first to Figure 9A-1, the frontal view of virtual model 910 shows that the patient has an abnormal curvature of the spine (e.g., scoliosis). This is marked by line X tracing the rostral-caudal axis of the spine. Referring to Figure 9A-1, the lateral view of virtual model 910 shows, marked by ellipse Y, that the patient's intervertebral disc is compressed or the spacing between adjacent vertebral endplates is reduced. Figures 9B-1 and 9B-2 show virtual model 920 modified to account for the abnormal anatomical configuration shown in Figures 9A-1 and 9A-2. For example, Figure 9B-1, a frontal view of virtual model 920, shows the spine of a patient with modified alignment (e.g., reduced abnormal curvature). This modification is also indicated by line X, which follows the rostral-caudal axis of the spine. Figure 9B-2, a lateral view of virtual model 920, shows the spine of a patient with restored intervertebral disc height (e.g., increased spacing between adjacent vertebral endplates), also marked by ellipse Y. Line X and ellipse Y are provided in Figures 9A-1 to 9B-2 to more clearly illustrate the modifications between virtual models 910 and 920 and are not necessarily included in the virtual models generated according to this technique.

[0141] Figure 10 shows an example of a surgical plan 1000 (for example, one generated in step 506 of process 500). The surgical plan 1000 may include preoperative patient indicators 1002, predicted postoperative patient indicators 1004, one or more patient images (e.g., patient images 703 received as part of a patient dataset), a virtual model 910 of the patient's innate anatomical structure (e.g., the patient's biostructure preoperatively) (which may be the model itself or one or more images derived from the model), and / or a virtual model 920 of the patient's modified anatomical structure (e.g., the patient's biostructure postoperatively) (which may be the model itself or one or more images derived from the model). The virtual model 920 of the patient's biostructure postoperatively may optionally include one or more implants 1012 shown to be implanted in the patient's spinal cord region to show how the patient's biostructure will look postoperatively. While the virtual model 920 shows four implants 1012, the surgical plan 1000 may include more or fewer implants 1012, such as one, two, three, five, six, seven, eight, or more implants 1012.

[0142] The surgical plan 1000 may include additional information beyond that shown in Figure 10. For example, the surgical plan 1000 may include pre-operative instructions, surgical instructions, and / or post-operative instructions. The surgical instructions may include one or more specific procedures to be performed (e.g., PLIF, ALIF, TLIF, LLIF, DLIF, XLIF, etc.) and / or one or more specific targets of the surgery (e.g., fusion of vertebral levels L1-L4, fixation screws inserted into the lateral surface of L4, etc.). Although the surgical plan 1000 is shown as a visual report in Figure 10, the surgical plan 1000 may also be encoded in computer-executable instructions, which, when executed by a processor connected to a computing device, cause the surgical plan 1000 to be displayed by the computing device. In some embodiments, the surgical plan 1000 may also include machine-readable surgical instructions for executing the surgical plan. For example, the surgical plan may include surgical instructions for a robotic surgical platform to execute one or more steps of the surgical plan 1000.

[0143] Figure 11 provides a series of images illustrating an example of a patient surgical plan report 1100, which includes a surgical plan 1000 and may be sent to the surgeon for review and approval. The surgical plan report 1100 may include a multi-page report detailing aspects of the surgical plan 1000. For example, a multi-page report may include a first page 1101 showing an overview of the surgical plan 1000 (e.g., as shown in Figure 10), a second page 1102 showing patient images (e.g., patient image 703 shown in Figure 7D, received in step 502), a third page 1103 showing a magnified view of a hypothetical model of the modified anatomical configuration (e.g., hypothetical model 920 shown in Figures 9B-1 and 9B-2), and a fourth page 1104 prompting the surgeon to approve or reject the surgical plan 1000. Naturally, additional information regarding the surgical plan may be presented with the report 1100 in the same or different format. In some embodiments, if a surgeon rejects the surgical plan 1000, the surgeon may be prompted to provide feedback on aspects of the surgical plan 1000 that the surgeon wishes to be adjusted.

[0144] The patient's surgical plan report 1100 may be presented to the surgeon on a digital display of a computing device (e.g., a client computing device 102 shown in Figure 1). In some embodiments, the report 1100 is interactive, allowing the surgeon to manipulate various aspects of the report 1100 (e.g., adjust, zoom in on, zoom out on, and annotate the display of a virtual model). However, even if the report 1100 is interactive, the surgeon typically cannot directly modify the surgical plan 1000. Rather, the surgeon may provide feedback and suggested changes to the surgical plan 1000, which may be sent back to the computing system that generated the surgical plan 1000 for analysis and improvement.

[0145] Figure 12A shows an example of a patient-specific implant 1200 (such as one designed in step 508 of process 500 and manufactured in step 512), and Figure 12B shows the implant 1200 implanted in a patient. The implant 1200 can be any orthopedic implant or other implant specifically designed to cause the patient's body to conform to a previously identified modified anatomical configuration. In the embodiments shown, the implant 1200 is an intervertebral device having a first surface (e.g., an upper surface) 1202 configured to occlude with the lower endplate surface of the upper vertebral body, and a second surface (e.g., an underside surface) 1204 configured to occlude with the upper endplate surface of the lower vertebral body. The first surface 1202 has a patient-specific topography designed to match (e.g., pair) with the topography of the lower endplate surface of the upper vertebral body, and can form an interface between their surfaces that is usually gapless. Similarly, the second surface 1204 has a patient-specific topography designed to coincide with or pair with the topography of the upper endplate surface of the lower vertebral body, and can form an interface between their surfaces that is usually gapless. The implant 1200 may also include a recess 1206 or other features configured to promote intra-bone growth. Because the implant 1200 is patient-specific and designed to cause geometric variations in the patient, the implant 1200 is not necessarily symmetrical and is often asymmetrical. For example, in the embodiments shown, the implant 1200 has a non-uniform thickness such that the plane defined by the first surface 1202 is not parallel to the central longitudinal axis A of the implant 1200. Naturally, because the implants described herein, including the implant 1200, are patient-specific, the art is not limited to any particular implant design or characteristic. Additional features of patient-specific implants that can be designed and manufactured in accordance with this technology are described in U.S. Patent Applications No. 16 / 987,113 and No. 17 / 100,396, the disclosures of which are incorporated herein by reference in their entirety.

[0146] Patient-specific medical procedures described herein may include implanting two or more patient-specific implants in a patient to achieve a modified anatomical configuration (e.g., multi-site procedures). For example, Figure 13 shows a lower spinal cord region with three patient-specific implants 1300a–1300c implanted at different vertebral levels. Implants 1300a–1300c are similar to the implants described in relation to Figures 2–3B and may include one or more of their features. More specifically, the first implant 1300a is implanted between the L3 and L4 vertebral bodies, the second implant 1300b is implanted between the L4 and L5 vertebral bodies, and the third implant 1300c is implanted between the L5 vertebral body and the sacrum. Implants 1300a–c together may cause the patient's spinal cord region to assume a previously identified modified anatomical configuration (e.g., transforming the patient's biostructure from a diseased configuration pre-operatively to an optimized configuration post-operatively). In some embodiments, more or fewer implants are used to achieve a modified anatomical configuration. For example, in some embodiments, one, two, four, five, six, seven, eight, or more implants are used to achieve a modified anatomical configuration. In embodiments involving two or more implants, the implants do not necessarily have the same shape, size, or function. In practice, multiple implants often have different shapes and topography to correspond to the target spinal level in which they are implanted. As also shown in Figure 13, patient-specific medical procedures described herein may include treating a patient in multiple target areas (e.g., multiple spinal levels).

[0147] In addition to designing patient-specific medical care based on a reference patient dataset, the system and methods of this technology can also design patient-specific medical care based on disease progression in a particular patient. Therefore, in some embodiments, the technology includes software modules (e.g., machine learning models or other algorithms) that can be used to analyze, predict, and / or model disease progression in a particular patient. The machine learning model may be trained on multiple reference patient datasets, including disease progression indicators for each of the reference patients, in addition to the patient data described with respect to Figure 1. Progression indicators may include measurements of disease indicators over a period of time. Appropriate indicators may include spinal-pelvic parameters (e.g., LL, pelvic tilt, sagittal vertical axis (SVA), Cobb angle, coronal offset, etc.), physical disability scores, functional ability scores, flexibility scores, VAS pain scores, etc. Progression of indicators for each reference patient may correlate with other patient information of the particular reference patient (e.g., age, sex, height, weight, activity level, diet, etc.).

[0148] In some embodiments, the technology includes a disease progression module that comprises an algorithm, machine learning model, or other software analysis tool for predicting disease progression in a particular patient. The disease progression module may be trained on a reference patient dataset that includes patient information (e.g., age, sex, height, weight, activity level, diet) and disease indicators (e.g., diagnosis, LL, pelvic tilt, SVA, Cobb angle, coronal offset, and other spinal-pelvic parameters, disability score, functional ability score, flexibility score, VAS pain score, etc.). Disease indicators may include values ​​over a period of time. For example, the reference patient data may include values ​​of disease indicators on a daily, weekly, monthly, bimonthly, yearly, or other basis. By measuring indicators over a period of time, changes in the indicator values ​​can be tracked as estimates of disease progression and can correlate with other patient data.

[0149] Therefore, in some embodiments, the disease progression module can estimate the rate of disease progression for a particular patient. This progression may be estimated by providing an estimated change in one or more disease indicators over a period of time (e.g., an annual increase of X% in the disease indicator). This rate can be constant (e.g., an annual increase of 5% in pelvic tilt) or variable (e.g., a 5% increase in pelvic tilt in year 1, a 10% increase in pelvic tilt in year 2, etc.). In some embodiments, the estimated rate of progression can be sent to a surgeon or other healthcare provider who can review and update the estimate as needed.

[0150] As a non-limiting example, a particular patient, a 55-year-old male, may have an SVA value of 6 mm. The disease progression module can analyze the patient reference dataset to identify disease progression in individual reference patients who share one or more similarities with the particular patient (e.g., individual patients among the reference patients who have an SVA value of approximately 6 mm and whose age, weight, height, and / or sex are nearly the same). Based on this analysis, the disease progression module can predict the rate of disease progression if no surgical intervention occurs (e.g., if no surgical intervention occurs, the patient's VAS pain score may increase by 5%, 10%, or 15% per year; if no surgical intervention occurs, the SVA value may continue to increase by 5% per year, etc.).

[0151] The systems and methods described herein can also generate models / simulations based on the estimated rate of disease progression, thereby modeling various outcomes over a desired period of time. Furthermore, the models / simulations can consider any number of additional diseases or conditions to predict the patient's overall health, mobility, etc. These additional diseases or conditions can be used in combination with other patient health factors (e.g., height, weight, age, activity level, etc.) to generate a patient health score that reflects the patient's overall health. The patient health score can be presented for review by the surgeon and / or incorporated into the estimation of disease progression. Thus, the technology can generate one or more virtual simulations of predicted disease progression to show how the patient's biostructure is expected to change over time. Physician input may be used to generate or modify the virtual simulations. The technology can generate one or more post-treatment virtual simulations based on the physician input received for review by healthcare providers, patients, etc.

[0152] In some embodiments, the technology can also predict, model, and / or simulate disease progression based on one or more possible surgical interventions. For example, the disease progression module may simulate what a patient's biostructure might look like one, two, five, or ten years after surgery for several surgical intervention options. The simulations may also incorporate non-surgical factors, such as the patient's age, height, weight, sex, activity level, and other health conditions, as previously mentioned. Based on these simulations, the system and / or surgeon can select the surgical intervention that is best suited for long-term effectiveness. These simulations may also be used to determine patient-specific modifications to compensate for predicted disease progression.

[0153] Therefore, in some embodiments, multiple (e.g., two, three, four, five, six, or more) disease progression models are simulated to provide disease progression data for multiple different surgical intervention options or other situations. For example, a disease progression module can generate a model predicting disease progression after surgery for each of three different surgical interventions. A surgeon or other healthcare provider can review the disease progression model and, based on the review, select the option from the three surgical interventions that is most likely to yield the best long-term outcome for the patient. Naturally, as described herein, the selection of the optimal intervention can be fully automated or semi-automated.

[0154] Based on modeled disease progression, the systems and methods described herein can also (i) identify the optimal timing for surgical intervention and / or (ii) identify the optimal type of surgical procedure for a patient. Accordingly, in some embodiments, the technique includes an intervention timing module that includes an algorithm, machine learning model, or other software analysis tool for determining the optimal timing for surgical intervention in a particular patient. This determination may be performed, for example, by analyzing patient reference data that includes (i) preoperative disease progression indicators for an individual reference patient, (ii) disease indicators at the time of surgical intervention for an individual reference patient, (iii) postoperative disease progression indicators for an individual reference patient, and / or (iv) scored surgical outcomes for an individual reference patient. The intervention timing module can compare the disease indicators of a particular patient with the reference patient dataset to determine, for similar patients, the point in disease progression at which surgical intervention produced the most beneficial outcome.

[0155] As a non-limiting example, a reference patient dataset may include data associated with the reference patient's SVA. The data may include (i) the individual patient's SVA value over a period prior to surgical intervention (e.g., the rate and extent of change in the SVA value), (ii) the individual patient's SVA at the time of surgical intervention, (iii) the change in SVA after surgical intervention, and (iv) the degree of success of the surgical intervention (e.g., based on pain, quality of life, or other factors). Based on the aforementioned data, the intervention timing module can identify, based on a particular patient's SVA value, at what point in time a surgical intervention is most likely to yield the most beneficial results. Of course, the aforementioned indicators are provided merely as examples, and the intervention timing module may incorporate other indicators (e.g., LL, pelvic tilt, SVA, Cobb angle, coronal offset, physical disability score, functional ability score, flexibility score, VAS pain score) instead of or in combination with SVA to predict the timing at which a surgical intervention is most likely to yield the most beneficial results for a particular patient.

[0156] The intervention timing module may also incorporate one or more mathematical rules based on thresholds for various disease indicators. For example, the intervention timing module may indicate that surgical intervention is necessary if one or more disease indicators exceed a predetermined threshold or meet any other criteria. Typical thresholds indicating that surgical intervention may be necessary include an SVA value greater than 7 mm, a discrepancy greater than 10 degrees between lumbar lordosis and pelvic intrinsic angle, a Cobb angle greater than 10 degrees, and / or a combination of a Cobb angle greater than 20 degrees and a discrepancy between LL / PI. Naturally, other thresholds and indicators can be used, and the above are provided merely as examples and do not limit the disclosure in any way. In some embodiments, the aforementioned rules may be adjusted to suit a specific patient population (for example, for men over 50 years of age, an SVA value greater than 7 mm indicates the need for surgical intervention). If a particular patient does not exceed a threshold indicating that surgical intervention is recommended, the intervention timing module may provide an estimate of when the patient's indicators exceed one or more thresholds, thereby providing the patient with an estimate of when surgical intervention may be recommended.

[0157] The technology may also include a treatment planning module that can identify the most appropriate type of surgical procedure for a patient based on the progression of the patient's disease. The treatment planning module may be an algorithm, machine learning model, or other software analysis tool that is trained or otherwise based on multiple reference patient datasets, as described above. The treatment planning module may also incorporate one or more mathematical rules for identifying surgical procedures. As a non-limiting example, if the LL / PI mismatch is between 10 and 20 degrees, the treatment planning module may recommend anterior fixation, but if the LL / PI mismatch is greater than 20 degrees, the treatment planning module may recommend both anterior and posterior fixation. As another non-limiting example, if the SVA value is between 7 mm and 15 mm, the treatment planning module may recommend posterior fixation, but if the SVA is greater than 15 mm, the treatment planning module may recommend both posterior and anterior fixation. Naturally, other rules may be used, and the above are provided merely as examples and do not limit the disclosure in any way.

[0158] Integrating disease progression modeling into patient-specific medical procedures described herein, without being constrained by theory, can further enhance the effectiveness of those procedures. For example, surgery after a patient's disease has progressed to an irreversible or unstable state can often be disadvantageous. However, surgery performed too early, before a patient's disease causes symptoms and / or before the patient's disease can progress further, can also be disadvantageous. Therefore, the disease progression module and / or intervention timing module can help identify the time frame in which surgical intervention in a particular patient is most likely to yield beneficial results for the patient.

[0159] Figure 14 shows the spine 2030 of a patient with intervertebral implants positioned at individual levels. Implants 2000a–g (collectively, “Implant 2000”) can be adjusted to conform to anatomical features at individual levels. Non-invasive postoperative spinal adjustments can be performed by using inter-implant communication, a remote device or controller 2049 that controls the operation of implants 2000, or a combination thereof. Inter-implant communication can be wireless communication over a network maintained by implants 2000. The remote device 2049 can communicate wirelessly with selected implants or all implants. Implants 2000 can be the implants described in relation to Figures 2–4, or other implants disclosed herein.

[0160] The implant 2000a is implanted at level 2031, which has a normal endplate free from surface topology defects. The endplate of implant 2000a may have a convex shape that matches the indicated concave endplate of the adjacent vertebra at level 2031. The implant 2000a may have an actuation mechanism 2001 that can be powered by an external applied field (e.g., a magnetic field or another field) supplied by a remote device 2049. The actuation mechanism 2001 may include an inductively rechargeable power supply, actuation elements, a processor, a transmitter / receiver, etc. The location, number, and capabilities of the actuation mechanisms (e.g., 2001 or 2004) may be selected based on available adjustability (e.g., range expansion / contraction, driving force, etc.).

[0161] The remote device 2049 can communicate with one or more of the implants 2000. To install new software, the remote device 2049 can wirelessly transmit the software to at least one of the implants 2000. Implant 200 can be programmed to wirelessly receive the software, install and run the received software, and / or wirelessly transmit the software to another implant 2000. In some implementations, the remote device 2049 can communicate with the implants 2000 via wireless protocols, including mesh protocols (e.g., Zigbee protocol, Z-wave protocol, etc.), ad-hoc network protocols, etc. In some embodiments, the implants 2000 can receive wireless updates distributed upon authorization. If an implant 2000 detects any adverse event, the implant 2000 can send a request for additional software or instructions. The remote device 2049 or another device can receive the request and, accordingly, send instructions, wireless updates, etc., based on the request. In some implementations, the remote device 2049 may take the form of a smartphone, tablet, or computer that communicates with the implant 2000 via a local wireless connection such as Bluetooth or Wi-Fi.

[0162] Implant 2000 may have patient-specific features. Implant 2000b is implanted at level 2032, which has a severe concave shape in the upper and lower vertebrae. Implant 2000b has a large convex contour that matches the corresponding concave shape of the upper vertebra. Implant 2000c is implanted at level 2033, which has an upper endplate containing a local defect 2054 adjacent to, but not on, the longitudinal lateral surface of, the upper vertebra. Implant 2000d includes an upper endplate 2052 with a contour feature 2056 that generally corresponds to the local defect in order to better fit the upper endplate. Local defects in the patient's spine may range from relatively small cavities (e.g., as shown in level 2033) to relatively large valleys (e.g., as shown in level 2034). Furthermore, local defects may include projections (not shown) of extra bone and / or cartilage accumulations, which may require concave contour features in the endplate of the implant to match them. Implant 2000e is implanted in level 2035, which has corner defects in the upper and lower vertebrae. The corner defects are located at least partially on the longitudinal lateral surface of the vertebra. The corner defects can include missing corners cut off at various angles, protrusions at the corners (not shown), and / or rough topologies at the corners (e.g., at the missing corners, at the protrusions, and / or otherwise at the normal surface of the corner). Implant 2000e includes an upper endplate 2052 having an outer contour 2058 configured to fit the corner defect in the upper endplate, and a lower endplate 2053 having an outer contour 2060 configured to fit the corner defect in the lower endplate. Other adjacent levels, such as level 2036, may be formed by endplates having relatively smooth, planar, or linear topologies. In such embodiments, implant 2000f having a relatively smooth contour may be implanted in level 2036.

[0163] Implant 2000g is implanted at level 2037, which has an upper vertebra containing an erosive defect on the lower surface of the upper vertebra. An external device 2049 can be commanded to move implant 2000g to the target position. As shown in the figure, the erosive defect may spread across the entire surface of the vertebra and include multiple valleys and peaks. In some patients, the erosive defect may be contained to localized areas of the surface and / or corner areas. In some patients, the erosive defect may include one or more deep valleys and / or one or more high peaks. As shown in the figure, implant 2000g may have an upper endplate 2052 configured to pair with the erosive defect in the upper vertebra.

[0164] Figure 15 shows an exemplary corrective plan 2100 relating to a patient-specific surgical procedure that may be used and / or generated in connection with the methods described herein, according to an embodiment. The corrective plan 2100 may be an adjustable implant corrective plan that incorporates all or part of the surgical plans or other plans disclosed herein. The corrective plan 2100 may include, but is not limited to, intraoperative and / or preoperative patient indicators (e.g., preoperative patient indicators 1002 as described in connection with Figures 10-13), predicted postoperative patient indicators (e.g., predicted postoperative patient indicators 1004 as described in connection with Figures 10-13), adjustment indicators 2110, and adjustment configurations (e.g., implant adjustment configurations, biostructure adjustment configurations, spinal adjustment configurations, etc.).

[0165] The adjustment index 2110 may include any number of planned adjustments for an adjustable spinal implant. The shown correction plan 2100 includes planned adjustments 2120a, 2120b, and 2120c (collectively, "adjustment 2120"). Each adjustment 2120 may include associated adjusted indices that can be reviewed by the physician. For example, the physician can review and approve these indices by selecting an approve button. The computing system can then design an adjustable implant based on the approved adjustments (for example, an adjustable implant can be designed to have an adjustable range of motion that can accommodate the approved adjustments). If the physician wishes to change an adjustment, the physician can select a change button. The physician can then input one or more parameters or indices for the adjustment. The computing system can update the spinal model according to the input parameters or indices. Arrows (for example, arrows 2130a, 2130b, and 2130c) may indicate adjustments such as range of motion and adjustment values. In some environments, arrows 2130a-b and / or indicator 2120a may indicate the degrees of freedom and range of motion of a particular implant. The user can modify or approve the adjustability of the implant based on the arrows. Adjustments 2 and 3 include adjustment indicators (indicated by arrows) that show planned adjustments, such as the adjustments described in relation to Figures 16A-16D. The physician can approve / select individual target intraoperative and / or postoperative configurations for various loading conditions.

[0166] Planned adjustments 2120a, 2120b, and 2120c may include configurations for implantable patient devices that can operate autonomously in a coordinated manner to position the patient's anatomical elements to achieve a target anatomical configuration. The patient device may detect at least one value and then determine an anatomical adjustment based on the detected at least one value and a correction plan. These values ​​may be entered by the user or from the correction plan. One or more of the patient devices may modify their configuration to result in the determined anatomical adjustment. The correction plan (or part thereof) 2110 is sent to the implant. The correction plan 2110 may include protocols for device detection, determined communication details, data exchange, data processing, etc. This allows newly available devices to join a wireless network within the body. The implant may detect other implants for networking purposes and use a predetermined protocol for detection. For example, if a new device is implanted or coupled to a patient, another implant may detect this newly available device. The implant may perform a detection routine to authenticate the newly available device and allow it to join the network.

[0167] Figures 16A–16D show the patient's spine in different orientations that result in different loads on the implants. Figure 16A shows the patient's spine in a generally horizontal orientation. For example, the implant may be implanted when the patient's body is generally horizontal so that the spine is generally not subjected to load. During surgery, it can be difficult to determine how closely the load on the spine corresponds to the predicted load. Therefore, the implant may be reconfigured postoperatively to move the spine to the target postoperative configuration. Figures 16B–16D show postoperative adjustments to the patient's spine in a vertical orientation (e.g., sitting or standing), although postoperative adjustments may be performed on patients in other orientations as well. This allows for postoperative adjustments based on postoperative load, dynamic visualization, etc. Each implant can monitor the load and communicate with other implants to determine, for example, load along the vertebral segments, multi-level load relationships, and other parameters disclosed herein. Examples of parameters include intervertebral space height, lumbar lordosis, Cobb angle, coronal parameters (e.g., coronal balance, overall coronal balance, coronal pelvic tilt, etc.), sagittal parameters (e.g., pelvic intrinsic angle, sacral tilt, thoracic kyphosis, etc.), pelvic parameters, or combinations thereof.

[0168] In spinal fusion surgery, implants may be adjusted immediately after surgery (e.g., a few hours or days later) to position them in the vertebrae for fixation. In spinal alignment surgery, implants may be periodically adjusted to compensate for patient improvement, disease progression, etc. For example, adjustments 1-3 in Figures 16B-16D may be performed monthly, annually, or at intervals determined by the physician. The number of adjustment sessions, the intervals between adjustment sessions, and the changes to the spine may be selected based on the treatment plan, patient recovery, etc. For example, the system in Figure 1 and the computer device in Figure 2 may be used to generate modification and adjustment plans. For example, the computer system may be used to determine the patient's modified anatomical configuration to achieve the target treatment outcome. The computer system can use at least one machine learning model to predict disease progression of the disease affecting the patient's spine based on the patient's patient dataset. The computer system can identify operable implants configured to be implanted in the patient to achieve the modified anatomical configuration. The operable implants are movable between multiple configurations to compensate for predicted disease progression based on the target treatment outcome. At least one machine learning model can determine, based on the adjusted images, whether to reconfigure at least one device. The adjusted images may include dynamic sitting / standing X-ray images, and in some adjustment procedures, the spine may be visualized (e.g., using fluoroscopy) while the workable implant is invasively or non-invasively activated.

[0169] In some embodiments, one or more anatomical modifications are generated for a patient based on pre-adjustment images and a patient-specific pre-operative modification plan. The computer system can generate a series of modified anatomical models representing anatomical changes over a period of time, based on patient-specific modifications to the innate biostructure and predicted disease progression. The modified anatomical models can be viewed and modified by the user as part of the pre-operative modification plan. The pre-operative modification plan can be generated by comparing the patient dataset with multiple reference patient datasets and identifying one or more similar patient datasets from among the multiple reference patient datasets, each similar patient dataset corresponding to (a) a reference patient with spinal condition data similar to the patient, and / or (b) a reference patient treated with a post-operatively adjustable orthopedic implant. In some embodiments, a virtual model of the spine is generated. Predicted disease progression uses the virtual model. An operable implant can be designed to fit the virtual model throughout the predicted disease progression. The simulation can be modified and rerun based on post-operative adjustments (see Figures 16A–16D). Additional implants configured to work in conjunction with the operable implant may be designed to achieve the target therapeutic outcome and configured for multiple levels of adjustment. The plans disclosed herein can provide results from simulations of multiple levels of adjustment (e.g., level-by-level analysis, overall spinal correction score, etc.).

[0170] The networked systems and networked devices disclosed herein may include data storage elements for storing patient-specific data, and retrieval functions for accessing patient-specific data. A data storage module containing memory for storing data, and a retrieval module configured to transmit patient-specific surgical plans from the data storage module to a surgical platform may be configured to perform one or more aspects of the patient-specific surgical plan. Thus, patient-specific data is linked to patient-specific implants. The data may be accessed after the implant has been implanted. The data may be used to verify aspects of the implant / surgery (e.g., whether the implant is properly positioned) and may be combined, aggregated, and analyzed with post-implant treatment data (e.g., implant status data, configuration data, sensor data, etc.). U.S. Patent Application No. 16 / 990,810 discloses features, systems, devices, materials, and methods that may be incorporated into or used with the networked systems and networked devices disclosed herein. U.S. Patent Application No. 16 / 990,810 is incorporated herein by reference in its entirety.

[0171] In some embodiments, the technology can also predict, model, and / or simulate disease progression. For example, the disease progression module may simulate what a patient's biostructure might look like one, two, five, or ten years after surgery for several surgical intervention options. The simulations may also incorporate non-surgical factors, such as the patient's age, height, weight, sex, activity level, and other health conditions, as previously mentioned. Based on these simulations, the system and / or surgeon can select the surgical intervention best suited for long-term effectiveness. These simulations may also be used to determine patient-specific modifications to compensate for predicted disease progression. Networked systems and devices can generate data for monitoring and predicting disease progression. In some embodiments, one or more of the implantable devices include a disease progression module for local analysis of the data. In other embodiments, a remote computing device may include a disease progression module. The disease progression module can predict disease progression continuously or periodically, while the implanted networked system adjusts the modifications.

[0172] The systems disclosed herein may also include multiple (e.g., two, three, four, five, six, or more) disease progression models that are simulated to provide disease progression data for multiple different surgical intervention options or other situations. For example, a disease progression module may generate a model predicting postoperative disease progression for each of three different surgical interventions. A surgeon or other healthcare provider can review the disease progression model and, based on the review, select the option from the three surgical interventions that is most likely to yield the best long-term outcome for the patient. Naturally, as described herein, the selection of the optimal adjustment can be fully automated or semi-automated. The implanted networked system may be programmed using multiple disease progression models. The disease progression models may be modified based on collected data and healthcare providers, etc.

[0173] In some embodiments, networked implants can be used to correct a number of different diseases in a variety of situations, including spinal surgery, hand surgery, shoulder and elbow surgery, whole joint reconstruction (arthroplasty), skull reconstruction, pediatric orthopedics, foot and ankle surgery, musculoskeletal oncology, surgical sports medicine, or orthopedic trauma. The implants can dynamically correct irregular spinal curvature, such as (excessive or insufficient) scoliosis, lordosis, or kyphosis, and irregular vertebral displacement (e.g., spondylolisthesis). Thus, the correction can change over time to compensate for disease progression and patient growth (e.g., devices implanted when the patient is not yet fully grown). Networked devices can be designed to treat osteoarthritis, lumbar or cervical degenerative disc disease, lumbar stenosis, or cervical stenosis.

[0174] The networked systems and networked devices disclosed herein may include data storage elements for storing patient-specific data, and retrieval functions for accessing patient-specific data. A data storage module containing memory for storing data, and a retrieval module configured to transmit patient-specific surgical plans from the data storage module to a surgical platform may be configured to perform one or more aspects of the patient-specific surgical plan. Thus, patient-specific data is linked to patient-specific implants. The data may be accessed after the implant has been implanted. The data may be used to verify aspects of the implant / surgery (e.g., whether the implant is properly positioned) and may be combined, aggregated, and analyzed with post-implant treatment data (e.g., implant status data, configuration data, sensor data, etc.). U.S. Patent Application No. 16 / 990,810 discloses features, systems, devices, materials, and methods that may be incorporated into or used with the networked systems and networked devices disclosed herein. U.S. Patent Application No. 16 / 990,810 is incorporated herein by reference in its entirety.

[0175] In some embodiments, the technology can also predict, model, and / or simulate disease progression. For example, the disease progression module may simulate what a patient's biostructure might look like one, two, five, or ten years after surgery for several surgical intervention options. The simulations may also incorporate non-surgical factors, such as the patient's age, height, weight, sex, activity level, and other health conditions, as previously mentioned. Based on these simulations, the system and / or surgeon can select the surgical intervention best suited for long-term effectiveness. These simulations may also be used to determine patient-specific modifications to compensate for predicted disease progression. Networked systems and devices can generate data for monitoring and predicting disease progression. In some embodiments, one or more of the implantable devices include a disease progression module for local analysis of the data. In other embodiments, a remote computing device may include a disease progression module. The disease progression module can predict disease progression continuously or periodically, while the implanted networked system adjusts the modifications.

[0176] The systems disclosed herein may also include multiple (e.g., two, three, four, five, six, or more) disease progression models that are simulated to provide disease progression data for multiple different surgical intervention options or other situations. For example, a disease progression module may generate a model predicting postoperative disease progression for each of three different surgical interventions. A surgeon or other healthcare provider can review the disease progression model and, based on the review, select the option from the three surgical interventions that is most likely to yield the best long-term outcome for the patient. Naturally, as described herein, the selection of the optimal adjustment can be fully automated or semi-automated. The implanted networked system may be programmed using multiple disease progression models. The disease progression models may be modified based on collected data and healthcare providers, etc.

[0177] As those skilled in the art will understand, any of the previously described software functions may be combined with or distributed among one or more software functions or devices to perform the operations described herein. Therefore, any of the operations described herein may be performed by any of the computing devices or systems described herein unless otherwise explicitly mentioned.

[0178] Networked implants can be used to correct a number of different diseases in a variety of situations, including spinal surgery, hand surgery, shoulder and elbow surgery, whole joint reconstruction (arthroplasty), skull reconstruction, pediatric orthopedics, foot and ankle surgery, musculoskeletal oncology, surgical sports medicine, and orthopedic trauma. Implants can dynamically correct irregular spinal curvature, such as (excessive or insufficient) scoliosis, lordosis, or kyphosis, and irregular vertebral displacement (e.g., spondylolisthesis). Therefore, the correction can change over time to compensate for disease progression and patient growth (e.g., devices implanted when the patient is not yet fully grown). Networked devices can be designed to treat osteoarthritis, lumbar or cervical degenerative disc disease, lumbar spinal stenosis, and cervical spinal stenosis. Networked implants can include orthopedic implants (e.g., artificial hip joints, fracture repair structures, alignment implants, spinal support structures, and corresponding attachment mechanisms), sensory / neurological implants, transplant organs, and assistive mechanisms (e.g., pacemakers, defibrillators, valves, and stents). Other devices, such as attachable or wearable devices (e.g., blood glucose monitors, cardiac monitors), may be attached to or worn by a patient for extended periods. Even more devices (e.g., personal devices such as mobile phones, smartwatches, and / or other personal health monitors) may be carried by the patient or within a certain distance of the patient for a significant portion of the day. These devices can be part of the network. [Examples]

[0179] This technology is presented according to various embodiments, for example, as described below. Various embodiments of this technology are described as numbered embodiments (1, 2, 3, etc.) for convenience. These are provided as embodiments and do not limit the technology. Note that any of the dependent embodiments may be combined in any suitable manner and placed in each independent embodiment. Other embodiments may be presented in a similar manner. 1. An optional additive manufacturing apparatus capable of manufacturing orthopedic implants, An implant manufacturing system comprising an implant design platform configured to communicate with an additive manufacturing device and to design orthopedic implants specific to one or more patients by virtual anatomical modeling, wherein the implant design platform is One or more processors, The implant design platform includes one or more memories that store instructions, and when these instructions are executed by one or more processors, The patient dataset associated with the patient is input into at least one trained machine learning model, and the patient dataset is input into at least one trained machine learning model, Identify at least one edge case of a patient condition in the patient dataset. Determine the score for the condition of at least one edge case based on at least one characteristic of the condition of at least one edge case. In response to the score exceeding the threshold, the system will initiate a request for human review of the condition in at least one edge case. An implant manufacturing system comprising a process that includes receiving input from human review, transmitting the implant design to an additive manufacturing device, which is configured to manufacture a patient-specific orthopedic implant by additive manufacturing according to the implant design. 2. The process is To prevent the implant design platform from triggering implant manufacturing for patients, it generates manufacturing holds and, The implant manufacturing system according to Example 1 further includes removing a manufacturing hold to enable transmission of the implant design to an additive manufacturing apparatus in response to receiving input from human review. 3. The process is This further includes determining the condition score for at least one edge case based on the scored treatment outcomes of patient-specific orthopedic implants addressing the patient's spinal condition, The scored treatment outcomes include data representing at least one of the following: modified anatomical indicators, presence of fixation, health-related quality of life, activity level, or at least one complication. An implant manufacturing system according to any one of Examples 1 to 2, wherein the score represents a statistical correlation between a patient dataset and at least one of several reference patients. 4.1 or more processors, A system comprising one or more memories storing instructions, wherein when these instructions are executed by one or more processors, the system causes the system to execute a process for identifying the condition of an edge case for human examination, and this process is Train at least one machine learning model using images that display reference pathologies. The patient dataset associated with the patient is input into at least one trained machine learning model, and the patient dataset is input into at least one trained machine learning model, By comparing the patient dataset with multiple reference patient datasets, we can identify at least one edge case of a patient condition in the patient dataset. Determine the score for the condition of at least one edge case based on at least one characteristic of the condition of at least one edge case. A system that, in response to a score exceeding a threshold, causes at least one device to send a request for a human to examine the medical condition of at least one edge case in a patient dataset before the system triggers the manufacture of one or more patient-specific implants for the patient. 5. The process is The method further includes determining a score for at least one edge case of a patient's condition based on the scored treatment outcomes of one or more patient-specific implants addressing the patient's spinal condition. The scored treatment outcomes include data representing one or more of the following: modified anatomical indicators, presence of fixation, health-related quality of life, activity levels, or complications. The system described in the example, wherein the score represents a statistical correlation between a patient dataset and at least one of several reference patient datasets. 6. The process is The system according to any one of Examples 4-5, further comprising sending a user an approval notification regarding the placement of one or more patient-specific implants to the patient in response to the score falling below a threshold. 7. The process is The system according to any one of Examples 4 to 6, further comprising sending a second request to the device to capture one or more images of the patient's spine, wherein one or more images include a pathological condition of at least one edge case. 8. The process is The system according to any one of Examples 4 to 7, further comprising generating an automated request for diagnostic information specific to a medical condition, wherein the diagnostic information specific to the medical condition includes a target treatment outcome by implanting one or more patient-specific implants in the patient. 9. The system according to any of Examples 4 to 8, wherein the requirement includes at least one automated annotation of the pathological condition of at least one edge case in the patient dataset. 10. The system according to any one of Examples 4 to 9, wherein at least one edge case of pathology includes intervertebral fusion, anatomical abnormalities, osteotomy symptoms, fractures in vertebral features, tumors, or transitional vertebrae. 11. Transmitting a patient dataset to an implant design platform configured to communicate with a manufacturing device and design an orthopedic implant specific to one or more patients by virtual anatomical modeling, wherein the implant design platform inputs the patient dataset into at least one trained machine learning model, and the implant design platform inputs the patient dataset into at least one trained machine learning model, Identify at least one edge case of a patient condition in the patient dataset. Determine the score for the condition of at least one edge case based on at least one characteristic of the condition of at least one edge case. The system is configured to send a request for human review of the condition in at least one edge case in response to the score exceeding a threshold, and to transmit a request for human review of the condition in at least one edge case. For human review, display the pathology of at least one identified edge case, A computer-aided method comprising transmitting user input relating to the medical condition of at least one identified edge case, the input being received by an implant design platform for the design of a human-assisted orthopedic implant specific to one or more patients. 12. Sending a manufacturing hold from at least one user device to prevent the implant design platform from initiating implant manufacturing for a patient, The computer implementation method according to Example 11 further includes transmitting input from at least one user device to remove a manufacturing hold from human review. 13. Further comprising determining a score for at least one edge case of a condition based on the scored treatment outcomes of one or more patient-specific orthopedic implants addressing the patient's spinal condition, The scored treatment outcomes include data representing at least one of the following: modified anatomical indicators, presence of fixation, health-related quality of life, activity level, or at least one complication. A computer-aided method according to any one of Examples 11 to 12, wherein the score represents a statistical correlation between a patient dataset and at least one of several reference patients. 14. A computer-aided method for identifying the condition of an edge case for human examination, wherein this method is Train at least one machine learning model using images that display reference medical conditions. The patient dataset associated with the patient is input into at least one trained machine learning model, and the patient dataset is input into at least one trained machine learning model, By comparing the patient dataset with multiple reference patient datasets, we can identify at least one edge case of a patient condition in the patient dataset. Determine the score for the condition of at least one edge case based on at least one characteristic of the condition of at least one edge case. A computer implementation method comprising, in response to a score exceeding a threshold, causing the system to send a request to at least one device for a human to examine the medical condition of at least one edge case in a patient dataset before the system triggers the manufacture of one or more patient-specific orthopedic implants for the patient. 15. Further comprising determining a score for at least one edge case of a condition based on the scored treatment outcomes of one or more patient-specific orthopedic implants addressing the patient's spinal condition, The scored treatment outcomes include data representing one or more of the following: modified anatomical indicators, presence of fixation, health-related quality of life, activity levels, or complications. The computer implementation method according to Example 14, wherein the score represents a statistical correlation between a patient dataset and at least one of several reference patient datasets. 16. A computer-aided method according to any one of Examples 14-15, further comprising sending a user an approval notice regarding the placement of one or more patient-specific orthopedic implants to the patient in response to the score falling below a threshold. 17. A computer-aided method according to any one of Examples 14 to 16, further comprising sending a second request to the device to capture one or more images of the patient's spine, wherein one or more images include a pathological condition of at least one edge case. 18. A computer-aided method according to any one of Examples 14 to 17, further comprising generating an automated request for diagnostic information specific to a medical condition, wherein the diagnostic information specific to the medical condition includes a target treatment outcome by implanting one or more patient-specific orthopedic implants in the patient. 19. A computer implementation method according to any of Examples 14 to 18, wherein the requirement includes at least one automated annotation of the pathological condition of at least one edge case in the patient dataset. 20. A computer-aided method according to any one of Examples 14-19, wherein at least one edge case of a medical condition includes intervertebral fusion, anatomical abnormalities, osteotomy symptoms, fractures in spinal features, tumors, or transitional vertebrae. 21. A non-transient computer-readable medium storing instructions, wherein, when executed by a computing system, these instructions cause the computing system to perform actions to identify the condition of an edge case for human examination, and these actions Train at least one machine learning model using images that display reference medical conditions. The patient dataset associated with the patient is input into at least one trained machine learning model, and the patient dataset is input into at least one trained machine learning model, By comparing the patient dataset with multiple reference patient datasets, we can identify at least one edge case of a patient condition in the patient dataset. Determine the score for the condition of at least one edge case based on at least one characteristic of the condition of at least one edge case. A non-transient computer-readable medium, which includes causing the system to send a request to at least one device for a human to examine the medical condition of at least one edge case in a patient dataset before the system triggers the manufacture of one or more patient-specific orthopedic implants for the patient, in response to a score exceeding a threshold. 22. The operation is, The method further includes determining a score for at least one edge case of a patient's condition based on the scored treatment outcomes of one or more patient-specific orthopedic implants addressing the patient's spinal condition. The scored treatment outcomes include data representing one or more of the following: modified anatomical indicators, presence of fixation, health-related quality of life, activity levels, or complications. A non-transient, computer-readable medium according to Example 21, wherein the score represents a statistical correlation between a patient dataset and at least one of several reference patient datasets. 23. The operation is, A non-transient computer-readable medium according to any one of Examples 21-22, further comprising sending a user an approval notice regarding the placement of one or more patient-specific orthopedic implants to a patient in response to a score falling below a threshold. 24. The operation is, A non-transient computer-readable medium according to any of Examples 21 to 23, further comprising sending a second request to the device to capture one or more images of the patient's spine, wherein one or more images include a pathological condition of at least one edge case. 25. The operation is, A non-transient computer-readable medium according to any of Examples 21 to 24, further comprising generating an automated request for diagnostic information specific to a medical condition, wherein the diagnostic information specific to the medical condition includes a target treatment outcome by implanting one or more patient-specific orthopedic implants in the patient. 26. A non-transient computer-readable medium according to any of Examples 21-25, wherein the request includes at least one automated annotation of the pathological condition of at least one edge case in the patient dataset. 27. Identifying at least one edge case medical condition in a patient's patient dataset using an implant design platform, wherein the implant design platform is configured to digitally analyze the patient dataset and design a patient-specific orthopedic implant for automated manufacturing. In response to identifying the pathological condition of at least one edge case, To prevent the implant design platform from causing implant manufacturing for patients, the implant design platform generates patient-linked manufacturing holds, From the implant design platform, submit a request to reconsider the pathology of at least one identified edge case, The system receives input from the user device regarding a requested re-examination of the medical condition of at least one identified edge case, A computer-aided method comprising determining whether to remove a manufacturing hold via an implant design platform, based on received input, in order to enable the automated manufacturing of one or more patient-specific orthopedic implants for a patient. 28. The computer implementation method according to Example 27, wherein the determination of whether to remove the manufacturing hold includes determining whether the received input meets the patient's treatment threshold. 29. Via an electronic screen, A patient data window that displays one or more annotated patient images, A computer-aided method according to any one of Examples 27 to 28, further comprising displaying an edge case graphical interface, which includes a simulation window displaying at least one implant simulation that labels one or more anatomical parameters that contribute to the identification of a pathological condition in at least one edge case. 30. In response to the received input meeting the treatment continuation threshold, To generate at least one design of an orthopedic implant specific to one or more patients, A computer-aided method according to any one of Examples 27 to 29, further comprising transmitting manufacturing data for manufacturing one or more patient-specific orthopedic implants according to at least one design. 31. A computer-aided implementation method according to any of Examples 27-30, wherein the manufacturing hold includes one or more of the following: a design lock to prevent the implant design platform from designing any implant for a patient; a manufacturing data lock to prevent the generation of implant manufacturing data; and / or a transmission lock to prevent the transmission of implant design data to a manufacturing device. 32. Using an implant design platform, Segmenting patient images, Identifying one or more features within a segmented image, A computer-aided method according to any one of Examples 27 to 31, further comprising digitally analyzing a patient dataset by determining whether one or more identified features qualify as edge features using a machine learning algorithm trained on a reference dataset. 33. Identifying the pathology of at least one edge case is Scoring the features of one or more candidate edges in a patient dataset, Based on the scoring, the aggregation of edge feature scores is determined, A computer implementation method according to any one of Examples 27 to 32, comprising comparing the aggregated edge feature scores with a threshold for the pathology of an edge case to identify the pathology of at least one edge case. 34. Classifying one or more parameters within a patient dataset, Based on this classification, edge case scoring is performed for one or more parameters, A computer-aided method according to any one of Examples 27 to 33, further comprising determining the pathological condition of multiple edge cases based on classification and edge case scoring. 35. Displaying at least one patient image containing one or more annotated edge-case parameters, Adjust the display of at least one patient image according to the display input from the user, A computerized implementation according to any one of Examples 27 to 34, further comprising dynamically displaying information for one or more annotated edge case parameters that are viewable by the user and / or selected by the user. 36. A computer-aided method according to any of Examples 27 to 35, further comprising dynamically calculating anatomical indices of a modified anatomical configuration, generated based on anatomical adjustments, the number of implants, or the implant configuration entered by the user. 37. Displaying calculated anatomical indicators, A computer-aided method according to any one of Examples 27 to 36, further comprising annotating one or more of the calculated anatomical features to indicate at least one edge-case pathological condition. 38. Identifying edge-case medical conditions of a patient by applying one or more image processing algorithms to at least one digital image of the patient, To generate an edge-case condition review plan including at least one annotated image of the patient and edge-case identification information, A computer-aided edge case detection method, comprising designing one or more implants using a virtual model of the patient's biostructure based on feedback from an edge case condition reassessment plan. 39. A computer-implemented edge case detection method according to Example 38, comprising machine-executable instructions for machine learning analysis to determine whether the edge case condition review plan should proceed with the patient's implant design process. 40. The plan to re-examine the symptoms of edge cases is A computer-implemented edge case detection method according to any one of Examples 38 to 39, comprising calculated anatomical indices and images labeled to associate one or more of the calculated anatomical indices with anatomical features indicating the pathology of an edge case. 41. The edge case symptom review plan is now available in a graphical interface. Displaying at least one patient image containing one or more annotated edge-case parameters, Adjust the display of at least one patient image according to the display input from the user, A computer-implemented edge case detection method according to any one of Examples 38 to 40, which causes the computer to dynamically display information of one or more annotated edge case parameters that can be displayed by the user. 42. The edge case condition review plan includes a set of anatomical indicators for edge case analysis, and the method is as follows: Receiving user input regarding anatomical adjustments, the number of implants, or at least one of the implants, Modify the anatomical structure in accordance with the edge case disease reassessment plan, To determine one or more anatomical landmarks within the set of altered anatomical structures, A computer-implemented edge case detection method according to any of Examples 38 to 41, further comprising displaying one or more determined anatomical indicators for edge case analysis. 43. Using an image analysis program, analyze one or more images of a patient to identify at least one anatomical abnormality in one or more images. To determine whether at least one anatomical anomaly requires re-examination by a human, The process involves generating a re-examination report for display through human re-examination, wherein the re-examination report includes at least one image of at least one anatomical anomaly, and at least one anatomical anomaly is labeled by a label. A computer-aided method, including designing a patient-specific implant for a patient in response to receiving input from human review. 44. The computer-aided method described in Example 43, wherein the label includes an anatomical description label. 45. A computer implementation method according to any one of Examples 43 to 44, further comprising enabling the execution of one or more design and manufacturing steps based on the received input in response to receiving input from a human review. 46. ​​A computer-aided method according to any one of Examples 43 to 45, wherein at least one anatomical anomaly indicates the pathology of the edge case. 47. A computer-aided method according to any one of Examples 43 to 46, further comprising synchronizing one or more implant design and manufacturing steps with human review to develop a surgical plan that compensates for the edge case condition and a patient-specific virtual model of the implant. 48. Receiving compensation for edge cases from human review, A computer-aided method according to any one of Examples 43 to 47, further comprising determining the surgical plan based on compensation for edge cases. 49. Displaying a review report for user review on at least one user device, wherein the review report includes information generated by an image analysis program that analyzes one or more images of a patient to identify at least one anatomical anomaly requiring human review, and the at least one anatomical anomaly is labeled and displayed. A computer-aided implementation method comprising transmitting user input via at least one user device to an implant design platform programmed to design a patient-specific implant for a patient based on one or more virtual simulations. 50. The computer-aided method described in Example 49, wherein the label includes an anatomical description label. 51. The computer implementation method according to Examples 49-50, further comprising enabling the execution of one or more design and manufacturing steps based on the received input in response to receiving input from a human review. 52. The computer-aided method according to Examples 49-51, wherein at least one anatomical abnormality indicates the pathology of the edge case. 53. The computer-aided method according to Examples 49-52, further comprising synchronizing one or more implant design and manufacturing steps with human review to develop a surgical plan that compensates for the edge case condition and a patient-specific virtual model of the implant. 54. Receiving compensation for edge cases from human review, The computer-aided method according to Examples 49-53 further includes determining the surgical plan based on compensation for edge cases. 55. Displaying, by at least one user device, one or more quantified characteristics of the patient's condition in at least one edge case and at least one identified anatomical abnormality in the patient for user review, A computer-aided implementation method comprising transmitting user input from at least one user device to an implant design platform programmed to analyze the pathological condition of at least one edge case and design one or more treatments and / or implants. 56. The computer-aided method according to Example 55, wherein one or more quantified characteristics include one or more vertebral indicators. 57. A virtual model of the pathology of at least one edge case, One or more annotated images representing the condition of at least one edge case, or The computer implementation method according to Examples 55-56, further comprising displaying at least one of one or more quantified characteristics. 58. A computer-aided implementation according to Examples 55-57, further comprising displaying an interactive report and / or an image of the pathological condition of at least one edge case configured to operate on one or more models by at least one user device, wherein one or more quantified characteristics in the interactive report are modified based on this operation. 59. The computer implementation method according to Example 58, wherein the interactive report includes one or more selectable menus or buttons. 60.1 or more processors, A computing system comprising one or more memories storing instructions, wherein when these instructions are executed by one or more processors, the computing system causes the computing system to execute one of the processes described in Examples 11-20 and / or 27-59. 61. A non-transient computer-readable medium storing instructions, wherein, when executed by a computing system, these instructions cause the computing system to perform one of the operations described in Examples 11-20 and / or 27-59.

[0180] As those skilled in the art will understand, any of the previously described software modules may be combined into a single software module to perform the operations described herein. Similarly, software modules can be distributed across any combination of computing systems and devices described herein, and are not limited to the specific arrangements described herein. Thus, any of the operations described herein may be performed by any of the computing devices or systems described herein unless otherwise explicitly mentioned.

[0181] The detailed descriptions above illustrate various embodiments of devices and / or processes using block diagrams, flowcharts, and / or examples. To the extent that such block diagrams, flowcharts, and / or examples include one or more functions and / or operations, it will be understood by those skilled in the art that each function and / or operation included in such block diagrams, flowcharts, or examples can be implemented individually and / or collectively by a wide range of hardware, software, firmware, or virtually any combination thereof. In some embodiments, multiple parts of the subject matter described herein may be implemented by application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), or other integrated forms. However, a person skilled in the art will recognize that some aspects of the embodiments disclosed herein can be equivalently implemented in an integrated circuit, in whole or in part, as one or more computer programs running on one or more computers (e.g., as one or more programs running on one or more computer systems), as one or more programs running on one or more processors (e.g., as one or more programs running on one or more microprocessors), as firmware, or in virtually any combination thereof, and that designing circuits and / or writing software and / or firmware code is well within the skill of a person skilled in the art in light of this disclosure. In addition, a person skilled in the art will understand that the mechanisms of the subject matter described herein can be distributed as program products in various forms, and that the examples of embodiments of the subject matter described herein apply regardless of the particular type of signaling medium used to actually carry out the distribution.Examples of signal transmission media include, but are not limited to, recordable media such as floppy disks, hard disk drives, CDs, DVDs, digital tapes, and computer memory, as well as transmission media such as digital and / or analog communication media (e.g., fiber optic cables, waveguides, wired communication links, wireless communication links, etc.).

[0182] Those skilled in the art will recognize that, within the art, it is common to describe devices and / or processes in the manner shown herein and then to integrate such described devices and / or processes into a data processing system using technical methods. That is, at least some of the devices and / or processes described herein can be integrated into a data processing system by a reasonable amount of experimentation. Those skilled in the art will recognize that a typical data processing system usually includes one or more of the following: a system unit housing, a video display device, memory such as volatile and non-volatile memory, a processor such as a microprocessor and a digital signal processor, computing entities such as an operating system, drivers, a graphical user interface, and application programs, one or more interaction devices such as a touchpad or touchscreen, and / or control systems including feedback loops and control motors (e.g., feedback for detecting position and / or velocity, control motors for moving and / or adjusting components and / or quantities). A typical data processing system may be implemented using any suitable commercially available components, such as components commonly found in data computing / communication systems and / or network computing / communication systems.

[0183] The subject matter described herein sometimes involves various components that are contained within or connected to other different components. It should be understood that such shown architectures are merely examples, and that in practice, many other architectures can be implemented to achieve the same functionality. Conceptually, any arrangement of components to achieve the same functionality is effectively “associated” in such a way that the desired functionality is achieved. Thus, any two components in this specification that are combined to achieve a particular functionality can be considered “associated” with each other, regardless of the architecture or intermediate components, in such a way that the desired functionality is achieved. Similarly, any two such associated components can be considered “operably connected” or “operably coupled” with each other to achieve the desired functionality, and any two components that can be associated in such a way can be considered “operably coupled” with each other to achieve the desired functionality. Specific examples of operatically coupled components include, but are not limited to, components that can be physically connected and / or physically interact, as well as components that can / or wirelessly interact, and / or wirelessly interact, and / or logically interact, and / or logically interactable.

[0184] The embodiments, features, systems, devices, materials, methods, and techniques described herein may, in some embodiments, be similar to one or more of the embodiments, features, systems, devices, materials, methods, and techniques described below. U.S. Patent Application No. 16 / 048,167, "SYSTEMS AND METHODS FOR ASSISTING AND AUGMENTING SURGICAL PROCEDURES," filed on July 27, 2017. U.S. Patent Application No. 16 / 242,877, "SYSTEMS AND METHODS OF ASSISTING A SURGEON WITH SCREW PLACEMENT DURING SPINAL SURGERY," filed on January 8, 2019. U.S. Patent Application No. 16 / 207,116, "SYSTEMS AND METHODS FOR MULTI-PLANAR ORTHOPEDIC ALIGNMENT," filed on December 1, 2018. U.S. Patent Application No. 16 / 352,699, "SYSTEMS AND METHODS FOR ORTHOPEDIC IMPLANT FIXATION," filed on March 13, 2019. U.S. Patent Application No. 16 / 383,215, "SYSTEMS AND METHODS FOR ORTHOPEDIC IMPLANT FIXATION," filed on April 12, 2019. U.S. Patent Application No. 16 / 569,494, "SYSTEMS AND METHODS FOR ORTHOPEDIC IMPLANTS," filed on September 12, 2019. U.S. Patent Application No. 62 / 773,127, "SYSTEMS AND METHODS FOR ORTHOPEDIC IMPLANTS," filed on November 29, 2018. U.S. Patent Application No. 62 / 928,909, "SYSTEMS AND METHODS FOR DESIGNING ORTHOPEDIC IMPLANTS BASED ON TISSUE CHARACTERISTICS," filed on October 31, 2019. U.S. Patent Application No. 16 / 735,222, “PATIENT-SPECIFIC MEDICAL PROCEDURES AND DEVICES, AND ASSOCIATED SYSTEMS AND METHODS,” filed on January 6, 2020. U.S. Patent Application No. 16 / 987,113, “PATIENT-SPECIFIC ARTIFICIAL DISCS, IMPLANTS AND ASSOCIATED SYSTEMS AND METHODS,” filed on August 6, 2020. U.S. Patent Application No. 16 / 990,810, "Linking Patient-Specific Medical Devices with Patient-Specific Data, and Associated Systems, Devices, and Methods," filed on August 11, 2020. U.S. Patent Application No. 17 / 678,874, "NON-FUNGIBLE TOKEN SYSTEMS AND METHODS FOR STORING AND ACCESSING HEALTHCARE DATA," filed on February 23, 2022. U.S. Patent Application No. 17 / 085564, "SYSTEMS AND METHODS FOR DESIGNING ORTHOPEDIC IMPLANTS BASED ON TISSUE CHARACTERISTICS," filed on October 30, 2020. U.S. Patent Application No. 17 / 100,396, “PATIENT-SPECIFIC VERTEBRAL IMPLANTS WITH POSITIONING FEATURES,” filed on November 20, 2020. U.S. Patent Application No. 17 / 531,417, "PATIENT-SPECIFIC JIG FOR PERSONALIZED SURGERY," filed on November 19, 2021. U.S. Patent Application No. 17 / 835,777, “PATIENT-SPECIFIC EXPANDABLE SPINAL IMPLANTS AND ASSOCIATED SYSTEMS AND METHODS,” filed on June 8, 2022. International patent application PCT / US22 / 42188, "BLOCKCHAIN ​​MANAGED MEDICAL IMPLANTS," filed on August 31, 2022. U.S. Patent Application No. 17 / 851,487, “PATIENT-SPECIFIC ADJUSTMENT OF SPINAL IMPLANTS, AND ASSOCIATED SYSTEMS AND METHODS,” filed on June 28, 2022. U.S. Patent Application No. 17 / 856,625, "MESHED NETWORK OF MEDICAL IMPLANTS," filed on July 1, 2022. U.S. Patent Application No. 17 / 867,621, “PATIENT-SPECIFIC SACROILIAC IMPLANT, AND ASSOCIATED SYSTEMS AND METHODS,” filed on July 18, 2022. U.S. Patent Application No. 17 / 842,242, “PATIENT-SPECIFIC ANTERIOR PLATE IMPLANTS,” filed on June 16, 2022. U.S. Patent Application No. 17 / 978,673, “SPINAL IMPLANTS AND SURGICAL PROCEDURES WITH REDUCED SUBSIDENCE, AND ASSOCIATED SYSTEMS AND METHODS,” filed on November 1, 2022. U.S. Patent Application No. 17 / 868,729, "SYSTEMS FOR PREDICTING INTRAOPERATIVE PATIENT MOBILITY AND IDENTIFYING MOBILITY-RELATED SURGICAL STEPS," filed on July 19, 2022. U.S. Patent Application No. 17 / 978,746, “PATIENT-SPECIFIC SPINAL INSTRUMENTS FOR IMPLANTING IMPLANTS AND DECOMPRESSION PROCEDURES,” filed on November 1, 2022. International patent application PCT / US22 / 48729, “PATIENT-SPECIFIC ARTHROPLASTY DEVICES AND ASSOCIATED SYSTEMS AND METHODS,” filed on November 2, 2022. U.S. Patent Application No. 18 / 113,573, “PATIENT-SPECIFIC IMPLANT DESIGN AND MANUFACTURING SYSTEM WITH A DIGITAL FILING CABINET MANAGER,” filed on February 23, 2023. U.S. Patent Application No. 17 / 878,633, "NON-FUNGIBLE TOKEN SYSTEMS AND METHODS FOR STORING AND ACCESSING HEALTHCARE DATA," filed on August 1, 2022. U.S. Patent No. 11,806,241, "SYSTEM FOR MANUFACTURING AND PRE-OPERATIVE INSPECTING OF PATIENT-SPECIFIC IMPLANTS," issued on November 7, 2023. U.S. Patent Application No. 18 / 120,979, "MULTI-STAGE PATIENT-SPECIFIC SURGICAL PLANS AND SYSTEMS AND METHODS FOR CREATING AND IMPLEMENTING THE SAME," filed on March 13, 2023. U.S. Patent Application No. 18 / 455,881, "SYSTEMS AND METHODS FOR GENERATING MULTIPLE PATIENT-SPECIFIC SURGICAL PLANS AND MANUFACTURING PATIENT-SPECIFIC IMPLANTS," filed on August 25, 2023. U.S. Patent No. 11,793,577, "Techniques to Map Three-Dimensional Human Anatomy Data to Two-Dimensional Human Anatomy Data," issued on October 24, 2023, and U.S. Patent Application No. 63 / 437,975, "SYSTEM FOR MODELING PATIENT SPINAL CHANGES," filed on January 9, 2023.

[0185] All of the above identified patents and applications are incorporated in their entirety by reference. In addition, the embodiments, features, systems, devices, materials, methods, and techniques described herein may, in some embodiments, be applied to or used in connection with one or more of the embodiments, features, systems, devices, or other methods.

[0186] The scope disclosed herein includes all overlaps, partial scopes, and combinations thereof. Language such as “at most,” “at least,” “greater than,” “less than,” and “between” includes the numbers listed. Numbers preceded by terms such as “approximately,” “about,” and “substantially,” when used herein, include the numbers listed (e.g., approximately 10% = 10%) and also represent amounts close to the stated amount that still perform the desired function or achieve the desired result. For example, the terms “approximately,” “about,” and “substantially” may refer to amounts within a range of less than 10%, less than 5%, less than 1%, less than 0.1%, and less than 0.01% of the stated amount.

[0187] From the above, it will be understood that various embodiments of the Disclosure are described herein for illustrative purposes only, and that various modifications can be made without departing from the scope and spirit of the Disclosure. Accordingly, the various embodiments disclosed herein are not intended to be limiting.

Claims

1. Additive manufacturing equipment capable of manufacturing orthopedic implants, An implant manufacturing system comprising an implant design platform configured to communicate with the additive manufacturing apparatus and to design orthopedic implants specific to one or more patients by virtual anatomical modeling, wherein the implant design platform is One or more processors, The implant design platform includes one or more memories storing instructions, and when the instructions are executed by the one or more processors, The patient dataset associated with the patient is input into at least one trained machine learning model, and the at least one trained machine learning model is, In the patient dataset, identify at least one edge case of the disease state. Based on at least one characteristic of the pathological condition of the at least one edge case, the score of the pathological condition of the at least one edge case is determined. In response to the score exceeding a threshold, the system will perform the action of sending a request for human review of the condition of at least one edge case. An implant manufacturing system that, after receiving input from the aforementioned human review, transmits the implant design to the additive manufacturing apparatus, which is configured to manufacture a patient-specific orthopedic implant by additive manufacturing according to the implant design, and performs a process including transmitting the design.

2. The aforementioned process, To prevent the implant design platform from triggering the manufacture of implants for the patient, a manufacturing hold is generated. The implant manufacturing system according to claim 1, further comprising removing the manufacturing hold in response to receiving the input from the human review, in order to enable transmission of the implant design to the additive manufacturing apparatus.

3. The aforementioned process, The method further includes determining the score of the condition of at least one edge case based on the scored treatment outcomes of the patient-specific orthopedic implants that address the spinal condition of the patient, The scored treatment outcome includes data representing at least one of the following: modified anatomical indicators, presence of fixation, health-related quality of life, activity level, or at least one comorbidity. The implant manufacturing system according to claim 1, wherein the score represents a statistical correlation between the patient dataset and at least one of a plurality of reference patients.

4. One or more processors, A system comprising one or more memories storing instructions, wherein when an instruction is executed by one or more processors, the system causes the system to execute a process for identifying the condition of an edge case for human examination, and the process Train at least one machine learning model using images that display reference medical conditions. The patient dataset associated with the patient is input into at least one trained machine learning model, and the at least one trained machine learning model is, By comparing the patient dataset with multiple reference patient datasets, at least one edge case of a patient condition is identified in the patient dataset. Based on at least one characteristic of the pathological condition of the at least one edge case, the score of the pathological condition of the at least one edge case is determined. A system that, in response to the score exceeding a threshold, causes at least one device to send a request for a human to examine the condition of at least one edge case in the patient dataset before the system causes the manufacture of one or more patient-specific implants for the patient.

5. The aforementioned process, The method further includes determining the score of the condition of at least one edge case based on the scored treatment outcomes of the one or more patient-specific implants that address the spinal condition of the patient, The aforementioned scored treatment outcomes include data representing one or more of the following: modified anatomical indicators, presence of fixation, health-related quality of life, activity levels, or complications. The system according to claim 4, wherein the score represents a statistical correlation between the patient dataset and at least one of the plurality of reference patient datasets.

6. The aforementioned process, The system according to claim 4, further comprising sending a notification to the user regarding the placement of one or more patient-specific implants to the patient in response to the score falling below the threshold.

7. The aforementioned process, The system according to claim 4, further comprising transmitting a second request to the device to capture one or more images of the patient's spine, wherein the one or more images include the pathology of at least one edge case.

8. The aforementioned process, The system according to claim 4, further comprising generating an automated request for diagnostic information specific to a medical condition, wherein the diagnostic information specific to the medical condition includes a target treatment outcome by implanting one or more patient-specific implants in the patient.

9. The system according to claim 4, wherein the requirement includes an automatic annotation of at least one disease condition of the at least one edge case in the patient dataset.

10. The system according to claim 4, wherein the pathological condition of the at least one edge case includes intervertebral fusion, anatomical abnormalities, osteotomy symptoms, fractures in vertebral features, tumors, or transitional vertebrae.

11. The patient dataset is transmitted to an implant design platform configured to communicate with a manufacturing device and design orthopedic implants specific to one or more patients by virtual anatomical modeling, wherein the implant design platform inputs the patient dataset into at least one trained machine learning model, and the at least one trained machine learning model... In the patient dataset, identify at least one edge case of the disease state. Based on at least one characteristic of the pathological condition of the at least one edge case, the score of the pathological condition of the at least one edge case is determined. The system is configured to send a request for human review of the condition of at least one edge case in response to the score exceeding a threshold, and to transmit To allow for human review, the pathological condition of at least one identified edge case is displayed, A computer-aided method comprising transmitting user input relating to the medical condition of at least one identified edge case, wherein the input is received by the implant design platform for the design of a human-assisted orthopedic implant specific to the one or more patients.

12. From at least one user device, transmit a manufacturing hold to prevent the implant design platform from initiating the manufacture of an implant for the patient, The computer implementation method according to claim 11, further comprising transmitting the input from at least one user device to remove the manufacturing hold from the human review.

13. The method further includes determining the score of the condition of at least one edge case based on the scored treatment outcomes of the one or more patient-specific orthopedic implants that address the spinal condition of the patient, The scored treatment outcome includes data representing at least one of the following: modified anatomical indicators, presence of fixation, health-related quality of life, activity level, or at least one comorbidity. The computer implementation method according to claim 11, wherein the score represents a statistical correlation between the patient dataset and at least one of a plurality of reference patients.

14. A computer-aided method for identifying the condition of an edge case for human examination, wherein the method is Train at least one machine learning model using images that display reference medical conditions. The patient dataset associated with the patient is input into at least one trained machine learning model, and the at least one trained machine learning model is, By comparing the patient dataset with multiple reference patient datasets, at least one edge case of a patient condition is identified in the patient dataset. Based on at least one characteristic of the pathological condition of the at least one edge case, the score of the pathological condition of the at least one edge case is determined. A computer implementation method comprising, in response to the score exceeding a threshold, causing the system to send a request to at least one device for a human to examine the condition of the at least one edge case in the patient dataset before the system causes the manufacture of one or more patient-specific orthopedic implants for the patient.

15. The method further includes determining the score of the condition of at least one edge case based on the scored treatment outcomes of the one or more patient-specific orthopedic implants addressing the spinal condition of the said patient, The aforementioned scored treatment outcomes include data representing one or more of the following: modified anatomical indicators, presence of fixation, health-related quality of life, activity levels, or complications. The computer implementation method according to claim 14, wherein the score represents a statistical correlation between the patient dataset and at least one of the plurality of reference patient datasets.

16. The computer implementation method according to claim 14, further comprising sending a notification to the user regarding the placement of one or more patient-specific orthopedic implants to the patient in response that the score falls below the threshold.

17. The computer implementation method according to claim 14, further comprising transmitting a second request to the device to capture one or more images of the patient's spine, wherein the one or more images include the pathology of at least one edge case.

18. The computer-aided method according to claim 14, further comprising generating an automated request for diagnostic information specific to a medical condition, wherein the diagnostic information specific to the medical condition includes a target treatment outcome by implanting one or more patient-specific orthopedic implants in the patients.

19. The computer implementation method according to claim 14, wherein the requirement includes an automatic annotation of at least one pathological condition of the at least one edge case in the patient dataset.

20. The computer-aided method according to claim 14, wherein the pathological condition of the at least one edge case includes intervertebral fusion, anatomical abnormalities, osteotomy symptoms, fractures in vertebral features, tumors, or transitional vertebrae.

21. A non-transient computer-readable medium storing instructions, wherein, when the instructions are executed by a computing system, the computing system causes the computing system to perform an action to identify the condition of an edge case for human examination, and the action is Train at least one machine learning model using images that display reference medical conditions. The patient dataset associated with the patient is input into at least one trained machine learning model, and the at least one trained machine learning model is, By comparing the patient dataset with multiple reference patient datasets, at least one edge case of a patient condition is identified in the patient dataset. Based on at least one characteristic of the pathological condition of the at least one edge case, the score of the pathological condition of the at least one edge case is determined. A non-transient computer-readable medium, which in response to the score exceeding a threshold, causes the system to send a request to at least one device for a human to examine the condition of the at least one edge case in the patient dataset before causing the manufacture of one or more patient-specific orthopedic implants for the patient.

22. The aforementioned operation, The method further includes determining the score of the condition of at least one edge case based on the scored treatment outcomes of the one or more patient-specific orthopedic implants addressing the spinal condition of the said patient, The aforementioned scored treatment outcomes include data representing one or more of the following: modified anatomical indicators, presence of fixation, health-related quality of life, activity levels, or complications. The non-transient computer-readable medium according to claim 21, wherein the score represents a statistical correlation between the patient dataset and at least one of the plurality of reference patient datasets.

23. The aforementioned operation, The non-transient computer-readable medium according to claim 21, further comprising sending a notification to the user of approval for the placement of one or more patient-specific orthopedic implants to the patient in response to the score falling below the threshold.

24. The aforementioned operation, The non-transient computer-readable medium according to claim 21, further comprising transmitting a second request to the device to capture one or more images of the patient's spine, wherein the one or more images include the pathology of at least one edge case.

25. The aforementioned operation, A non-transient computer-readable medium according to claim 21, further comprising generating an automated request for diagnostic information specific to a medical condition, wherein the diagnostic information specific to the medical condition includes a target treatment outcome by implanting an orthopedic implant specific to the patient in the patient.

26. The non-transient computer-readable medium according to claim 21, wherein the requirement includes at least one automated annotation of the pathological condition of the at least one edge case in the patient dataset.

27. Identifying at least one edge case medical condition in a patient's patient dataset using an implant design platform, wherein the implant design platform is configured to digitally analyze the patient dataset and design a patient-specific orthopedic implant for automated manufacturing. In response to identifying the pathological condition of at least one edge case, To prevent the implant design platform from causing implant manufacturing for the patient, the implant design platform generates a patient-linked manufacturing hold, The implant design platform transmits a request to reconsider the pathology of the identified at least one edge case. Receiving input from the user device regarding the requested re-examination of the medical condition of the identified at least one edge case, A computer-aided method comprising determining, via the implant design platform, whether to remove the manufacturing hold in order to enable the automated manufacturing of one or more patient-specific orthopedic implants for the patient, based on the received input.

28. The computer implementation method according to claim 27, wherein the determination of whether to remove the manufacturing hold includes determining whether the received input meets the treatment threshold for the patient.

29. via electronic screen, A patient data window that displays one or more annotated patient images, The computer implementation method according to claim 27, further comprising displaying an edge case graphical interface, which includes a simulation window displaying at least one implant simulation that labels one or more anatomical parameters that contribute to identifying the pathological condition of the at least one edge case.

30. In response to the received input meeting the treatment continuation threshold, To generate at least one design of an orthopedic implant specific to the one or more patients, The computer-aided method according to claim 27, further comprising transmitting manufacturing data for manufacturing an orthopedic implant specific to one or more patients according to the at least one design.

31. The computer implementation method according to claim 27, wherein the manufacturing hold includes one or more of the following: a design lock to prevent the implant design platform from designing any implant for the patient; a manufacturing data lock to prevent the generation of implant manufacturing data; and / or a transmission lock to prevent the transmission of implant design data to a manufacturing device.

32. Using the aforementioned implant design platform, Segmenting the patient's image, Identifying one or more features within the segmented image, The computer-aided method according to claim 27, further comprising digitally analyzing the patient dataset by determining whether one or more of the identified features qualify as edge features using a machine learning algorithm trained on a reference dataset.

33. Identifying the pathological condition of at least one of the edge cases is Scoring the features of one or more candidate edges in the aforementioned patient dataset, Based on the aforementioned scoring, the aggregation of edge feature scores is determined, The computer implementation method according to claim 27, further comprising comparing the aggregated edge feature scores with a threshold for the pathological condition of an edge case to identify the pathological condition of at least one edge case.

34. Classifying one or more parameters within the aforementioned patient dataset, The process involves performing edge-case scoring of one or more parameters based on the aforementioned classification, The computer implementation method according to claim 27, further comprising determining the pathological condition of a plurality of edge cases based on the classification and edge case scoring.

35. Displaying at least one patient image containing one or more annotated edge-case parameters, Adjusting the display of at least one patient image according to the display input from the user, The computer implementation method according to claim 27, further comprising dynamically displaying information of one or more annotated edge case parameters that are displayable by the user and / or selected by the user.

36. The computer implementation method according to claim 27, further comprising dynamically calculating anatomical indicators of a modified anatomical configuration, generated based on anatomical adjustments, the number of implants, or the implant configuration entered by the user.

37. Displaying the calculated anatomical indicators, The computer-aided method according to claim 36, further comprising annotating one or more of the calculated anatomical indicators with a description of the pathology of at least one edge case.

38. By applying one or more image processing algorithms to at least one digital image of the patient, the patient's edge case pathological condition is identified. To generate an edgecase condition review plan including at least one annotated image of the patient and edgecase identification information, A computer-aided edge case detection method comprising designing one or more implants using a virtual model of the patient's biological structure based on feedback from a reassessment plan for the edge case condition.

39. The computer-implemented edge case detection method according to claim 38, comprising machine-executable instructions for machine learning analysis to determine whether the edge case condition review plan should proceed with the implant design process for the patient.

40. The aforementioned edge case symptom reassessment plan, A computer-implemented edge case detection method according to claim 38, comprising calculated anatomical indices and images labeled to associate one or more of the calculated anatomical indices with anatomical features indicating the pathology of the edge case.

41. The aforementioned edge case symptom reassessment plan is displayed in a graphical interface. Displaying at least one patient image containing one or more annotated edge-case parameters, Adjusting the display of at least one patient image according to the display input from the user, The computer-implemented edge case detection method according to claim 38, further comprising the function of dynamically displaying information of one or more annotated edge case parameters that can be displayed by the user.

42. The aforementioned edge case condition review plan includes a set of anatomical indicators for edge case analysis, and the method is Receiving user input regarding anatomical adjustments, the number of implants, or at least one of the implants, In accordance with the aforementioned edge case disease reassessment plan, the anatomical configuration will be modified, To determine one or more anatomical features within the set of the modified anatomical configurations, The computer-implemented edge case detection method according to claim 38, further comprising displaying one or more of the determined anatomical indicators for the edge case analysis.

43. Using an image analysis program, analyze one or more images of a patient to identify at least one anatomical abnormality in said one or more images, To determine whether the aforementioned at least one anatomical anomaly requires re-examination by a human, The process involves generating a re-examination report for display through human re-examination, wherein the re-examination report includes at least one image of the at least one anatomical anomaly, and the at least one anatomical anomaly is labeled by a label. A computer-aided method comprising designing a patient-specific implant for the patient in response to receiving input from the aforementioned human review.

44. The computer-aided method according to claim 43, wherein the label includes an anatomical description label.

45. The computer implementation method according to claim 43, further comprising enabling the execution of one or more design and manufacturing steps based on the received input in response to receiving input from the human review.

46. The computer-aided method according to claim 43, wherein the at least one anatomical anomaly indicates a pathological condition of an edge case.

47. The computer-aided method according to claim 43, further comprising synchronizing one or more implant design and manufacturing steps with human review to develop a surgical plan that compensates for the edge case condition and a virtual model of the implant specific to the patient.

48. Receiving compensation for edge cases from the aforementioned human review, The computer-aided method according to claim 43, further comprising determining a surgical plan based on the compensation for the edge case.

49. Displaying a review report for user review on at least one user device, wherein the review report includes information generated by an image analysis program that analyzes one or more images of a patient to identify at least one anatomical anomaly requiring human review, and the at least one anatomical anomaly is labeled and displayed. A computer-aided implementation method comprising transmitting user input via at least one user device to an implant design platform programmed to design a patient-specific implant for the patient based on one or more virtual simulations.

50. The computer-aided method according to claim 49, wherein the label includes an anatomical description label.

51. The computer implementation method according to claim 49, further comprising enabling the execution of one or more design and manufacturing steps based on the received input in response to receiving input from the human review.

52. The computer-aided method according to claim 49, wherein the at least one anatomical anomaly indicates a pathological condition of an edge case.

53. The computer-aided method according to claim 49, further comprising synchronizing one or more implant design and manufacturing steps with human review to develop a surgical plan that compensates for the edge case condition and a virtual model of the implant specific to the patient.

54. Receiving compensation for edge cases from the aforementioned human review, The computer-aided method according to claim 49, further comprising determining a surgical plan based on compensation for the edge case.

55. At least one user device displays, for user review, one or more quantified characteristics of the condition of at least one edge case of a patient and at least one identified anatomical abnormality of the patient, A computer-aided implementation method comprising transmitting user input from at least one user device to an implant design platform programmed to analyze the pathological condition of at least one edge case and design one or more treatments and / or implants.

56. The computer-aided method according to claim 55, wherein one or more quantified characteristics include one or more vertebral indices.

57. A virtual model of the pathology of at least one edge case, One or more annotated images representing the condition of at least one edge case, or The computer implementation method according to claim 55, further comprising displaying at least one of the one or more quantified characteristics.

58. The computer implementation method according to claim 55, further comprising displaying an interactive report and / or an image of the condition of the at least one edge case configured for manipulating one or more models on at least one user device, wherein the one or more quantified characteristics in the interactive report are modified based on the manipulation.

59. The computer implementation method according to claim 58, wherein the interactive report includes one or more selectable menus or buttons.

Images