System architecture and method for analyzing health data across geographic regions by priority using a distributed computing platform - Patents.com

A distributed cloud computing platform for health data analysis addresses the challenge of cross-border data transfer and analysis by setting priorities, decompressing and validating DICOM objects, and anonymizing data, ensuring secure and compliant health data processing across regions.

JP7594538B2Active Publication Date: 2024-12-04HEARTFLOW INC

Patent Information

Application Number
JP2021548691
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Priority Date
2019-02-22
Filing Date
2020-02-21
Publication Date
2024-12-04
Estimated Expiration
2040-02-21

AI Technical Summary

Technical Problem

Hospitals lack the capability to perform advanced health data analysis and face challenges in transferring and analyzing health data across borders due to file size, type, routing, user interface, and privacy regulations, necessitating a decentralized platform for secure, scalable, and flexible health data analysis.

Method used

A distributed cloud computing platform for health data analysis that receives DICOM objects, sets priority levels, decompresses and validates them, and transmits analyses while preserving patient privacy by anonymizing data and maintaining it within regional boundaries.

Benefits of technology

Enables efficient, secure, and prioritized health data analysis across geographic regions, allowing hospitals to access advanced analytics without transferring patient privacy information, ensuring compliance with regional regulations.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007594538000001
    Figure 0007594538000001
  • Figure 0007594538000002
    Figure 0007594538000002
  • Figure 0007594538000003
    Figure 0007594538000003
Patent Text Reader

Abstract

Systems and methods are disclosed for analyzing health data via a distributed cloud computing platform. One method for analyzing health data via a distributed cloud computing platform includes receiving a unique case file containing one or more anonymous DICOM objects for analysis, setting a priority level for the unique case file based on a priority associated with the one or more anonymous DICOM objects, decompressing and validating at least one of the one or more anonymous DICOM objects, and transmitting the completed analysis of at least one of the decompressed and validated anonymous DICOM objects according to the priority level of the unique case file.
Need to check novelty before this filing date? Find Prior Art

Description

[Background technology]

[0001] (Related Applications) This application claims priority to U.S. Provisional Application No. 62 / 809,139, filed February 22, 2019, the entire disclosure of which is incorporated herein by reference in its entirety.

[0002] (introduction) Hospitals generate and store a significant amount of health data records about patients. Traditionally, analysis of health data (e.g., medical records, x-rays, CT scans, MRIs, etc.) is performed by medical professionals at hospitals. However, hospitals may not be equipped to perform predictive health data analysis involving state-of-the-art technology. In these situations, health data may need to be analyzed by a third-party vendor or service.

[0003] Increasingly, various aspects of health data analysis are being performed outside of the hospital. When health data is analyzed using one or more third-party processing systems outside of the hospital, there may be various factors to consider for transferring and analyzing health data outside of the hospital, such as file size, type, routing, file validity (e.g., image quality), application user interface (API) to be used to facilitate selection of files to transfer, and privacy.

[0004] In addition, a third-party processing system may require a robust platform architecture to efficiently analyze and organize data from many different hospitals from different regions (or countries). For example, a specific set of health data analysis, storage, and management capabilities may only exist within a certain region, distinct from the hospital region. Hospitals outside the selected region may not have access to the analysis services because patient privacy regulations may dictate that patient health data cannot be transferred to the selected region.

[0005] Thus, there is a desire for a distributed health data platform that creates a secure, scalable, robust, and flexible infrastructure for end-to-end automated capture, cross-border health data analysis and processing, and analysis reporting. Additionally, there is a desire for the distributed health data platform to provide guided analyst workflows to improve data analysis and reporting to hospitals. The foregoing general description and the following detailed description are exemplary and explanatory only and are not limitations of the present disclosure. Various embodiments of the present disclosure relate to systems and methods that, according to one embodiment, efficiently analyze, manage, and store health data generated within different geographic regions while preserving patient privacy. Summary of the Invention [Means for solving the problem]

[0006] According to one aspect of the present disclosure, a system and method are disclosed for managing cross-border health data via a distributed cloud computing platform.

[0007] A method for analyzing health data via a distributed cloud computing platform, the method including receiving a unique case file containing one or more anonymous DICOM objects for analysis, setting a priority level for the unique case file based on a priority associated with the one or more anonymous DICOM objects, decompressing and validating at least one of the one or more anonymous DICOM objects, and transmitting a completed analysis of at least one of the decompressed and validated anonymous DICOM objects according to the priority level of the unique case file.

[0008] According to another embodiment, a system for analyzing health data via a distributed cloud computing platform comprises a data storage device storing instructions for analyzing health data via a distributed cloud computing platform; and a processor configured to receive a unique case file containing one or more anonymous DICOM objects for analysis, set a priority level for the unique case file based on a priority associated with the one or more anonymous DICOM objects, decompress and validate at least one of the one or more anonymous DICOM objects, and transmit the completed analysis of at least one of the decompressed and validated anonymous DICOM objects in accordance with the priority level of the unique case file.

[0009] According to another embodiment, a non-transitory computer readable medium for use on a computer system containing computer-executable programming instructions for implementing a method for analyzing health data via a distributed cloud computing platform, the method including receiving a unique case file containing one or more anonymous DICOM objects for analysis, setting a priority level for the unique case file based on a priority associated with the one or more anonymous DICOM objects, decompressing and validating at least one of the one or more anonymous DICOM objects, and transmitting a completed analysis of at least one of the decompressed and validated anonymous DICOM objects according to the priority level of the unique case file.

[0010] Additional objects and advantages of the disclosed embodiments will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the disclosed embodiments. The objects and advantages of the disclosed embodiments will be realized and attained by means of the elements and combinations particularly pointed out in the appended claims.

[0011] It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosed embodiments as claimed. The present specification also provides, for example, the following items: (Item 1) 1. A computer-implemented method for analyzing health data via a distributed cloud computing platform, the method comprising: receiving a unique case file containing one or more anonymous DICOM objects for analysis; setting a priority level for the unique case file based on a priority associated with the one or more anonymous DICOM objects; decompressing and validating at least one of the one or more anonymous DICOM objects; transmitting an analysis of at least one of the decompressed and validated anonymous DICOM objects, the analysis being completed according to a priority level of the unique case file; A method comprising: (Item 2) The analysis generating a volumetric image file based on at least one of the anonymous DICOM objects of the unique case file, the volumetric image file representing a three-dimensional (3D) volume; generating an archive file containing analysis data associated with said volumetric image file; generating a patient-specific model based on the volumetric image file and the archive file; 2. The computer-implemented method of claim 1, comprising: (Item 3) Item 3. The computer-implemented method of item 2, wherein the volumetric image file includes a file checksum correlating the volumetric image file to the archive file. (Item 4) monitoring a status of the unique case file; advancing an analysis of the unique case file based on a status and a priority level of the unique case file; 2. The computer-implemented method of claim 1, further comprising: (Item 5) 2. The computer-implemented method of claim 1, further comprising setting the priority level for the unique case file based on an entity that generates one or more DICOM objects associated with the one or more anonymous DICOM objects and / or an entity that analyzes the one or more anonymous DICOM objects. (Item 6) generating a preliminary result report based on the received analysis; and generating an interactive user interface of said interim result report; 2. The computer-implemented method of claim 1, further comprising: (Item 7) receiving modifications to the interim result report; generating a final result report based on said modifications; and storing said final result report, a predetermined identified network address, or a requested web interface; 2. The computer-implemented method of claim 1, further comprising: (Item 8) 7. The computer-implemented method of claim 6, wherein the final result report is pushed to a database, a predefined identified network address, or a request web interface using secure hypertext transfer protocol (HTTPS). (Item 9) 1. A system for analyzing health data via a distributed cloud computing platform, the system comprising: one or more data storage devices storing instructions for analyzing health data via a distributed cloud computing platform; one or more processors, the one or more processors executing the instructions; receiving a unique case file containing one or more anonymous DICOM objects for analysis; setting a priority level for the unique case file based on a priority associated with the one or more anonymous DICOM objects; decompressing and validating at least one of the one or more anonymous DICOM objects; transmitting an analysis of at least one of the decompressed and validated anonymous DICOM objects, the analysis being completed according to a priority level of the unique case file; one or more processors configured to perform a method comprising: A system comprising: (Item 10) The analysis further comprises: generating a volumetric image file based on at least one of the anonymous DICOM objects of the unique case file, the volumetric image file representing a three-dimensional (3D) volume; generating an archive file containing analysis data associated with said volumetric image file; generating a patient-specific model based on the volumetric image file and the archive file; 10. The system according to item 9, comprising: (Item 11) Item 11. The system of item 10, wherein the volumetric image file includes a file checksum correlating the volumetric image file to the archive file. (Item 12) The system further comprises: monitoring a status of the unique case file; advancing an analysis of the unique case file based on a status and a priority level of the unique case file; 10. The system of item 9, configured to: (Item 13) 10. The system of claim 9, further configured to set the priority level for the unique case file based on an entity that generates one or more DICOM objects associated with the one or more anonymous DICOM objects and / or an entity that analyzes the one or more anonymous DICOM objects. (Item 14) The system further comprises: generating a preliminary result report based on the received analysis; and generating an interactive user interface of said interim result report; 10. The system of item 9, configured to: (Item 15) The system further comprises: receiving modifications to the interim result report; generating a final result report based on said modifications; and storing said final result report, a predetermined identified network address, or a requested web interface; 10. The system of item 9, configured to: (Item 16) Item 16. The system of item 15, wherein the final result report is pushed to a database, a predefined identified network address, or a request web interface using secure hypertext transfer protocol (HTTPS). (Item 17) 1. A non-transitory computer readable medium for use on a computer system containing computer executable programming instructions for analyzing health data via a distributed cloud computing platform, the method comprising: receiving a unique case file containing one or more anonymous DICOM objects for analysis; setting a priority level for the unique case file based on a priority associated with the one or more anonymous DICOM objects; decompressing and validating at least one of the one or more anonymous DICOM objects; transmitting an analysis of at least one of the decompressed and validated anonymous DICOM objects, the analysis being completed according to a priority level of the unique case file; 16. A non-transitory computer readable medium comprising: (Item 18) The analysis generating a volumetric image file based on at least one of the anonymous DICOM objects of the unique case file, the volumetric image file representing a three-dimensional (3D) volume; generating an archive file containing analysis data associated with said volumetric image file; generating a patient-specific model based on the volumetric image file and the archive file; 20. The non-transitory computer-readable medium of item 17, comprising: (Item 19) The method further comprises: monitoring a status of the unique case file; advancing an analysis of the unique case file based on a status and a priority level of the unique case file; 20. The non-transitory computer-readable medium of item 17, comprising: (Item 20) 20. The method of claim 17, further comprising setting the priority level for the unique case file based on an entity that generates one or more DICOM objects associated with the one or more anonymous DICOM objects and / or an entity that analyzes the one or more anonymous DICOM objects. [Brief description of the drawings]

[0012] [Figure 1A] FIG. 1A depicts an example block diagram of a distributed health data management system according to an example embodiment of the present disclosure.

[0013] [Figure 1B] FIG. 1B depicts an example schematic diagram of the health data IO system of FIG. 1A according to an example embodiment of the present disclosure.

[0014] [Figure 1C] FIG. 1C depicts an exemplary schematic diagram of the health data analysis system of FIG. 1A according to an exemplary embodiment of the present disclosure.

[0015] [Diagram 2] FIG. 2 illustrates a detailed block diagram of the health data analysis system of FIG. 1C, according to an exemplary embodiment of the present disclosure.

[0016] [Diagram 3]FIG. 3 is a block diagram of an example method for balancing network load in the health data IO system 101 of FIG. 1A according to an example embodiment of the disclosure.

[0017] [Figure 4] FIG. 4 is a sequence diagram of a method for receiving, analyzing, and reporting health data according to an exemplary embodiment of the present disclosure.

[0018] [Diagram 5] FIG. 5 is a flow diagram of an example method for managing case processing ordering, according to an example embodiment of the present disclosure.

[0019] [Figure 6] FIG. 6 is a flow diagram of a method for analyzing cross-border health data using a distributed cloud computing platform according to an exemplary embodiment of the present disclosure.

[0020] [Figure 7] FIG. 7 is a flow diagram of an example method for post-processing and reporting result files from a health data analysis system according to an example embodiment of the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0021] Reference will now be made in detail to the exemplary embodiments of the present disclosure, examples of which are illustrated in the accompanying drawings. Wherever possible, the same reference numbers will be used throughout the drawings to refer to the same or like parts.

[0022] As used herein, the term "exemplary" is used in the sense of "example" and not "ideal." Additionally, the terms "a" and "an" do not denote a limitation of quantity herein, but rather denote the presence of one or more of the referenced items. For purposes of this disclosure, a "patient" may refer to any individual or person on whom a diagnostic or therapeutic analysis (e.g., data analysis) is being performed, or to any individual or person associated with one or more individual diagnostic or therapeutic analyses.

[0023] The term "health data" may include medical data collected about a patient at a single practice (e.g., Digital Imaging and Communications in Medicine (DICOM) objects), and health data collected about a patient by any healthcare provider, laboratory, specialist, clinician, etc. Health data may be associated with a particular patient, who is identifiable via patient privacy information. Patient privacy information may be any information that links health data to a particular individual. Patient privacy information may be defined objectively (as information that connects an individual to their health data) and / or by law and regulation. Various domains may include different definitions of the types of information that constitute patient privacy information.

[0024] A "region" may refer to any environment with borders defined by regulations protecting patient privacy. In one embodiment, a "region" may refer to a known geographic area governed by a set of regulations protecting patient privacy / personal data. For example, a "region" may refer to a state (e.g., California), a country (e.g., the United States, Japan, Canada, etc.), or a group of countries (e.g., Europe) that operate under a single set of patient privacy laws. For the United States, patient privacy information may include protected health information (PHI). For the United States and various other regions, patient privacy information may include personal health information, patient identifying information, etc. (e.g., CT, MRI, ultrasound, etc.). Various systems and methods for protecting patient health information while transmitting health data are disclosed, for example, in U.S. Non-Provisional Application No. 15 / 635,127, filed June 27, 2017, and entitled "Systems and Methods for Modifying and Reacting Health Data Across Geographic Regions," which is incorporated herein by reference in its entirety.

[0025] Alternatively or additionally, a "region" may refer to a given physical facility, a computing system of a given entity, or a given cloud platform. In such cases, the various regions between which health data is compiled may still be within a single country, state, or province. For example, a region may include a hospital that may remove patient privacy information from collected health data prior to transferring the health data to another region (e.g., a cloud platform) for data analysis. The hospital and cloud platform may be within the same geographically defined region, e.g., the United States.

[0026] The present disclosure includes systems and methods for managing health data via a distributed cloud computing platform. In one embodiment, the health data may be pre-processed, analyzed, and post-processed for reporting using one or more cloud-based web services. The health data management system may receive health data from one or more hospitals, perform automatic health data capture, prepare the health data for analysis, analyze / convert the health data in preparation for analysis, select or identify files to transfer for analysis, time the transfer of health data files for analysis, determine if health data files are relevant, monitor the progress or status of health data analysis, determine completed analysis of the health data, and otherwise facilitate interaction between hospitals or DICOM modalities and various cloud-based data analysis web services. In one embodiment, the health data may be in the form of DICOM objects.

[0027] In one embodiment, the health data management system may transfer health data (e.g., DICOM objects) from a customer location (e.g., hospital, clinic, etc.) to a cloud computing web service. For example, the system may receive health data directly from a CT scanner, CT workstation, or picture archiving and communication system (PACS) using the DICOM protocol and securely transfer the received health data to the cloud service. The cloud computing service may include a data analysis service that may derive one or more diagnostic metrics or patient recommendations based on the received health data. The system may further retrieve completed analysis of the transferred health data, aggregate analysis results from one or more cloud services, generate reports, and / or push reports to the hospital system.

[0028] In one embodiment, the health data may include patient-specific image data. In one scenario, the image data may include data regarding the geometry of the patient's heart, for example, at least a portion of the patient's aorta, a proximal portion of the main coronary artery (and branches extending therefrom) connected to the aorta, and the myocardium. In one embodiment, the patient-specific image data may be in the form of a DICOM file. The patient-specific anatomical data may be acquired non-invasively, for example, using a non-invasive imaging method. For example, CCTA is an imaging method in which a user may operate a computed tomography (CT) scanner to view and create images of structures, for example, the myocardium, the aorta, the main coronary arteries, and other blood vessels connected thereto. The CCTA data may vary in time, for example, to show changes in blood vessel shape over the cardiac cycle. CCTA may be used to generate images of the patient's heart.

[0029] Alternatively, other non-invasive imaging methods, including magnetic resonance imaging (MRI) or ultrasound (US), or invasive imaging methods such as digital subtraction angiography (DSA) may be used to generate images of the structures of the patient's anatomy. The imaging method may involve injecting a contrast agent intravenously into the patient to allow identification of the structures of the anatomy. Health data accompanying the images (e.g., provided by CCTA, MRI, etc.) may be provided by a third party vendor, such as a radiology lab or cardiologist, by the patient's physician, etc.

[0030] Other patient-specific anatomical data may also be determined non-invasively from the patient. For example, physiological data such as the patient's blood pressure, baseline heart rate, height, weight, hematocrit, stroke volume, etc. may be measured. Blood pressure may be the blood pressure in the patient's brachial artery (e.g., using a pressure cuff), such as maximum (systolic) and minimum (diastolic) pressure. Exemplary health data may therefore include data collected from imaging modalities, wearable devices, blood pressure cuffs, thermometers, fitness trackers, blood glucose meters, heart rate monitors, etc.

[0031] In one embodiment, preparing the file for analysis and retrieving the analyzed data may involve protecting patient privacy. For example, file preparation may include removing patient privacy information from the file. In one such case, the health data file may automatically include the patient privacy information when the file is generated. Exemplary patient privacy information may include protected health information (PHI) or patient privacy data equivalent to PHI. In one scenario, file preparation may include removing patient privacy information from the health data file before the health data is transferred from the hospital to the health data analysis system for analysis. In one embodiment, a result report may be generated comprising the analyzed health data, devoid of patient privacy information. The report may be pushed to the hospital / hospital interface, and the report may be combined with the corresponding patient privacy information. In this manner, data analysis may be performed in the remote domain without transmitting patient privacy information to the remote domain.

[0032] 1A is a block diagram of a distributed health data management system 100 for managing and analyzing health data via a cloud platform according to an exemplary embodiment of the present disclosure. The health data infrastructure may include one or more health data input / output (IO) systems 101 (101a-101n) and a health data analysis system 105, each of which may be a customer of the health data analysis system 105. For example, the health data analysis system 105 may provide a service including generating analyzed data (e.g., diagnostic metrics, treatment simulation, treatment optimization or recommendation, etc.) from health data. The health data input / output system 101 may then transmit the health data to the health data analysis system 105 to perform the service on the provided health data.

[0033] In one embodiment, one or more health data IO systems 101 may each be located in a different region than the location of the health data analysis system 105. The one or more health data systems 101 may communicate health data as input to the health data analysis system 105 via the wireless network 103. Alternatively, the health data may be communicated to the health data analysis system 105 via a wired connection. The health data IO system 101 may also receive reports of analyzed health data output from the health data analysis system 105 via the wireless network 103.

[0034] In one embodiment, the health data IO systems 101 may each represent a different pre-determined region where patient privacy information may not be freely transferred from one region to another. In one embodiment, the boundaries of the regions may be defined such that patient privacy information may pass freely within the boundaries of the regions and such that patient privacy information may not be freely transferred once across the boundaries of the regions. For example, the boundaries of the regions may be based on government regulations surrounding the transfer of patient privacy information. In such a case, the regions may refer to physical geographical regions (e.g., Europe, Japan, and the United States) where the governments of each region may dictate different rules for the transfer of patient privacy information. Alternatively or in addition, the health data IO systems 101 may represent regions where the boundaries are dictated by health regulations, insurance policies, hospital policies, and / or local / state / provincial regulations. For example, the health data IO systems 101 may represent various facilities (e.g., distinct hospitals, hospital networks, research facilities, patient populations, etc.). Exemplary distinctions between the health data IO system 101 and the health data analytics system 105 may include embodiments where health data is generated in the health data IO system 101 and there may be technical and legal barriers to transferring the health data to the health data analytics system 105 for analysis.

[0035] In one embodiment, each health data IO system 101 may be associated with a pre-configured priority level. The priority level may dictate the order in which health data is submitted for review (e.g., as illustrated in FIG. 2 ) and transmitted for analysis (e.g., by health data analytics system 105), and the order in which reports of the analyzed data may be consolidated or distributed (e.g., from health data analytics system 105 to health data IO system 101). One example scenario may include health data IO system 101a associated with a “high priority” priority level and health data IO system 101b associated with a “low priority” priority level. In such a case, even if the first set of health data is generated by health data IO system 101a at the same time that the second set of health data is generated by health data IO system 101b, the first set of data may be forwarded to health data analysis system 105, and health data analysis system 105 may communicate a report of the analyzed data back to health data IO system 101a before the second set of health data from health data IO system 101b is forwarded to or analyzed by health data analysis system 105.

[0036] Alternatively or in addition, each health data IO system 101 may be associated with multiple priority levels. For example, health data IO system 101a may include multiple priority settings, organized, for example, by task item (e.g., “action”), priority status (e.g., “low priority”, “high priority”, “urgent status”), or standard process (e.g., “check for updates”, “monitor for updates”, etc.). The priority settings may be pre-set depending on the configuration of the data transmission channel between health data IO system 101a and health data analysis system 105. The priority may refer not only to the priority status for data transmission, but also to the order in which data processing will be performed. Alternatively or in addition, the priority settings may be associated with a device generating the health data, dictated by a clinician / healthcare provider depending on the generation of the health data, dictated by a required analysis, etc. In other words, the priority of a case file of DICOM objects may be based on an entity generating one or more DICOM objects (associated with one or more anonymous DICOM objects) and / or an entity analyzing one or more anonymous DICOM objects of the case file.

[0037] The health data may be analyzed by the health data analysis system 105 in an order according to the priority level of the health data sets within each health data IO system 101 (e.g., relative to other sets of health data generated by the health data IO system 101) and according to the priority level of the health data IO system (e.g., relative to other health data IO systems). In one embodiment, the type of health data may also affect the priority status. For example, if a particular set of health data is found to be associated with a store or pre-created set of health data, the particular set of health data may automatically be considered "prioritized" or "high priority" regardless of the priority level of the store or pre-created set of health data. In other words, the priority setting may consist of and be influenced by various factors, and there may be several tiers of priorities in the health data management system 100.

[0038] In one embodiment, a user may dictate the priority settings and define how various factors affect or change the priorities upon initializing the health data management system 100 or upon initializing communications between the health data IO system 101 and the health data analytics system 105. Alternatively or in addition, the health data analytics system 105 may pre-define at least one or more priority settings such that some priority settings are consistent or used by default across all health data IO system 101 customers of the health data analytics system 105.

[0039] In addition to priority, other properties may be associated with the health data during the upload of the health data from each health data IO system 101 to the health data analysis system 105. Properties associated with the health data may include, for example, imaging modality, referring physician, laboratory test code, laboratory code, billing code, patient health insurance, insurance plan, etc. In one embodiment, properties determined during the health data upload may prompt one or more specific algorithms to be invoked by the health data analysis system 105 either directly in analyzing the data or in an ancillary manner (e.g., when preparing a report of the data or when the data is received or reported).

[0040] It should be understood that the health data IO system 101 and the health data analysis system 105 may include any type or combination of computing systems, such as handheld devices, personal computers, servers, clustered computing machines, and / or cloud computing systems. In an embodiment, the health data IO system 101 and the health data analysis system 105 may comprise an assembly of hardware including a memory, a central processing unit ("CPU"), and / or a user interface. The memory may include any type of RAM or ROM embodied in a physical storage medium such as a magnetic storage device, a semiconductor storage device such as a solid-state disk (SSD) or flash memory, an optical disk storage device, or a magneto-optical disk storage device, including a floppy disk, a hard disk, or a magnetic tape. The CPU may include one or more processors for processing data according to instructions stored in the memory.

[0041] FIG. 1B is a detailed schematic block diagram of the exemplary health data IO system 101 of FIG. 1A in accordance with an exemplary embodiment of the present disclosure. The health data IO system 101 may comprise a hospital interface, including one or more hospital computing devices 106, a data transport system ("connection system") 107, and a mobile user interface 109. An embodiment of the health data IO system 101 may also include a web service 112, a reporting database 113, and a patient health information database (e.g., PHI database 114). The customer support user interface 111 may communicate with the web service 112. The web service 112 may comprise a cloud service. The health data IO system 101 may further include a metadata database. These components of the health data IO system 101 may be located in the same region or country as one or more hospitals.

[0042] The hospital computing devices 106 may include any type of electronic device configured to collect, transmit, and / or receive data, including websites, models, medical data, health records, multimedia content, and the like. Exemplary hospital computing devices may include medical devices, e.g., medical imaging devices, medical monitors, and the like. The hospital computing devices 106 may also include one or more mobile devices, smartphones, personal digital assistants ("PDAs"), tablet computers, or any other type of touch screen enabled device, personal computer, laptop, and / or server. In one embodiment, the hospital computing devices 106 may each have a web browser and / or mobile browser installed to receive and display electronic content received from one or more of the web servers associated with the health data IO system 101. The hospital computing devices 106 may also include client devices, which may have an operating system configured to run a web or mobile browser and any type of application, e.g., a mobile application. In one embodiment, the various hospital computing devices 106 may be configured with a network adapter to communicate data or analyzed reports. Alternatively, or in addition, the various hospital computing devices 106 may be configured to transmit data or receive analyzed data via local network connections.

[0043] In an exemplary embodiment, the hospital device may be a computed tomography (CT) scan workstation, a CT scanner, or a picture archiving and communication system (PACS) system installed in a hospital. In an embodiment, the hospital computing device 106 may upload health data to the connectivity system 107 via a wireless network. The connectivity system 107 may serve as a connection between the health data IO system 101 / hospital computing device 106 and the health data analysis system 105 (e.g., by pushing DICOM objects generated by the hospital computing device 106 to the remote health data analysis system 105).

[0044] In one embodiment, health data may be uploaded from the hospital computing device 106 to the connectivity system 107. For example, a DICOM study (including a CT scan) may be pushed from the DICOM modality hospital computing device 106 to the connectivity system 107. The connectivity system 107 may create a new case file corresponding to the DICOM study. In addition, the connectivity system 107 may prompt the creation of a new case account or file in the health data analysis system 105. In an alternative embodiment, the connectivity system 107 may query hospital computing devices (such as CT workstations, CT scanners, and PACS) and pull health data (e.g., DICOM objects) from such computing devices. In one embodiment, the hospital computing devices may communicate with a web server that stores DICOM objects. The connectivity system 107 may push or pull DICOM objects from the web server.

[0045] In one embodiment, the connection system 107 may communicate with a web server at regular time intervals and pull any DICOM objects available via the server. Alternatively, or in addition, a processor associated with generating DICOM objects may be in local communication with the connection system 107, and the connection system 107 may be prompted to pull DICOM objects as they are generated or stored.

[0046] Alternatively or in addition, the pull mechanism may also be initiated by a request for any task, for example, for analysis of a DICOM object. In one embodiment, a DICOM object may be associated with various data, for example, a priority of analysis, a type of case study to be performed, patient scans / images, patient privacy information, etc. This data may be metadata or tags included within the DICOM object file. Such components of the DICOM object may affect the order in which the files are sent for analysis. In one instance, the DICOM object may be stored until data analysis is requested. The pull mechanism used to transfer cases to the health data analysis system 105 may be governed by a case priority level or case type associated with each DICOM object. For example, health data may be labeled with a "type," which may pull data for analysis based on a rate or order dictated by the data type or priority setting. Exemplary types may include the following: urgent, priority, retrospective, urgent, low priority, testing, research, commercial, standard, demo, training, surveillance, prospective, clinical research. Types may be stacked, for example, an "urgent" case may be further sorted into "urgent" or "priority", where "urgent" may indicate high priority due to a patient's health condition being at risk, and "priority" may indicate a need for immediate analysis for reasons other than the patient's immediate health condition. Case type settings may be set in the hospital computing device 106, the mobile user interface 109, and / or by the connectivity system 107. The pull mechanism may be configured to minimize duplicate data pulls or overloading the health data IO system 101, the connectivity system 107, and the health data analytics system 105.

[0047] Alternatively, or in addition, a hospital user or clinician may upload health data to the connected system 107 using the mobile user interface 109. The mobile user interface 109 may be a customer-facing web interface (e.g., to a hospital user). The mobile user interface 109 may present a login interface to one or more users, perform password management, display case lists, search, display case details, and display user profiles. The mobile user interface 109 may also be used to access reports containing analyzed data associated with a patient. The mobile user interface 109 may comprise a mobile application.

[0048] In one embodiment, data generated by the hospital computing device 106 may be associated with patient privacy information, and the patient privacy information in the DICOM object may be available to the web service 112, but is not available or accessible to entities outside the health data IO system 101. For example, the connection system 107 may extract the patient privacy information (e.g., PHI data) from the DICOM object and store the patient privacy information so that it is available to the web service 112 (e.g., PHI database 114) in the health data IO system 101. For example, the connection system 107 may split the DICOM tag containing the patient privacy information from the DICOM image. The DICOM tag containing the patient privacy information may be replaced with a predefined string. As an alternative embodiment, the web service 112 may receive the DICOM object from the connection system 107 and extract the patient privacy information from the DICOM object. The web service 112 may further store the patient privacy information in a database (e.g., PHI database 114). In an exemplary embodiment, domain-specific customer service providers may have access to patient privacy information (PHI) and patient outcome reporting using the customer support user interface 111.

[0049] In one embodiment, the connection system 107 may create a case file for data analysis using an application programming interface (API) deployed in the health data IO system 101 (e.g., an area associated with a hospital). Patient privacy information removed from the DICOM file of the case may be submitted to the API, which may store the submitted patient privacy information in a patient privacy information database (e.g., PHI database 114). Once the case is processed (e.g., data in the DICOM file is analyzed by the health data analysis system 105), case / analysis completion may trigger a reporting process deployed in the health data IO system 101, meaning a report can be generated using the analyzed data and the patient privacy information stored in the patient privacy information database (e.g., PHI database 114). In one embodiment, the patient privacy information may be retrieved using the API.

[0050] Patient privacy information may always remain in the domain where the health data for the case (e.g., DICOM object) was generated or where the health data was pushed. Patient privacy information may be stored in storage service buckets and / or patient privacy information databases within each of the domains (e.g., health data IO system 101 of FIG. 1A).

[0051] In one embodiment, all of the patient privacy information may be encrypted in transit and at rest. In one embodiment, the storage service bucket for patient privacy information may enable server-side encryption so that encryption may occur each time a new object is stored. The patient privacy information database (e.g., PHI database 114) may also provide an encryption service whereby patient privacy information may be encrypted upon storage and each time the patient privacy information is accessed.

[0052] In one embodiment, patient privacy information cannot be accessed outside the region in which it is stored. For example, a user who submitted a case in Japan may be able to access various web services while in another region (e.g., Europe), but the user may not be able to access the patient privacy information in the health data management system 100 while in the other region.

[0053] In an exemplary embodiment, the web service 112 may receive a report of the analyzed data from the health data analysis system 105. For example, the web service 112 may determine that a completed report is available in the report database 113. In one embodiment, the completed report may be associated with a unique case identifier. In one embodiment, the web service 112 may use the unique case identifier to retrieve patient privacy information from the PHI database 114 associated with the retrieved case results. The web service 112 may further integrate the PHI data with the results report and / or store the report in the report database 113. The web service 112 may further store the report in the report database 113 either with the patient privacy information or with the unique case identifier. In one embodiment, the health data IO system 101 may further include a metadata database. The metadata database may include metadata associated with each report or set of health data. The metadata may include data associated with the health data known at the time the data was created or imaged, insurance plan, billing information, etc. Metadata may also include data created in the process of analysis of the health data (e.g., by the health data analysis system 105), such as notes or markings from a physician or reviewer, pins from health data pre-processing (e.g., when an anatomical model is constructed from the health data), etc.

[0054] In one embodiment, the web services 112 may include a report queue monitor that may apply an order or priority in which reports are accessed or made available for use (e.g., by clinicians, medical personnel, patients, etc.). The priority may be set by a user or by the web service (e.g., the web services 112). In one embodiment, the report queue may manage cases in order of case priority (as dictated by case “type”). The health data IO system 101 may further include a customer support user interface 111. The customer support user interface 111 may be accessible to a customer support provider (e.g., customer support personnel). In one embodiment, the patient privacy information and / or patient reports may be accessed via a virtual private network (VPN) configured for its use by the customer support provider. For example, the customer support personnel may have the ability to log into each area individually by utilizing a VPN service. The VPN service may extend access to the patient privacy information of a specific area to the customer support provider even when the customer support provider is not accessing the patient privacy information from the specific area in which the health data was generated.

[0055] 1C is a schematic block diagram of a health data analysis system 105 according to an exemplary embodiment of the present disclosure. The health data analysis system 105 may comprise a case processor 121, one or more analytic web services 123, an anonymous temporary storage 124a, an anonymous permanent storage 124b, a synchronization processor 125, and a modeling workstation 127. The health data analysis system 105 may further provide a user interface for analysts or customer support personnel to review and / or access data from the one or more analytic web services 123. In one embodiment, the user interface may include a generation user interface 129 and / or a customer support user interface 131.

[0056] The health data analysis system 105 may serve as one or more cloud platforms for analyzing anonymous health data (e.g., via the connection system 107) from at least one cross-border region. More specifically, the analysis web service 123 may calculate results, generate reports (e.g., 3D models, PDFs, etc.), and present the analysis and results to the patient, clinician, or user. For example, the one or more analysis web services 123 may calculate blood flow in coronary arteries using received image data from a cardiac CT scan. Various embodiments of diagnostic analysis using image reconstruction and anatomical models are disclosed, for example, in U.S. Patent No. 8,315,812, issued November 20, 2012, and entitled "Method and System for Patient-Specific Modeling of Blood Flow," which is incorporated by reference in its entirety. The one or more analysis web services 123 may provide reports of the results of the analysis performed by the analysis web service 123. For the analysis described above, the report may include an interactive 3D model of the coronary arteries and / or a PDF report showing the calculated fractional flow reserve (FFR) values. Exemplary reports may include surface models, graphs, charts, color coding, indicators, interactive displays, alerts, signals, or any known graphical objects. Various embodiments of graphical representations and displays are disclosed, for example, in U.S. Patent No. 8,706,457, issued April 22, 2014, and entitled "Method and System for Providing Information from a Patient-Specific Model of Blood Flow," which is incorporated by reference in its entirety.

[0057] In an exemplary embodiment, one or more processors associated with the case processor 121 may generate one or more intermediate results from the received DICOM objects. In an exemplary embodiment, a synchronization processor 125 may synchronize the intermediate results from one or more processors, which may be accessed for review via a modeling workstation 127. In an exemplary embodiment, the synchronization processor 125 may be an application launched from a user interface accessible by people (e.g., technicians or analysts) processing the cases. In one embodiment, the synchronization processor 125 may also launch a modeling workstation 127 for each specific case. The synchronization processor 125 may ensure proper communication between one or more analysis web services 123 and the modeling workstation 127. In an exemplary embodiment, an analyst may submit a request to view intermediate results from case processing, update results generated by the case processor 121, or reanalyze a case via the case processor 121.

[0058] In an exemplary embodiment, the health data analytics system 105 may include multiple front ends, including a generation user interface 129 and a customer support user interface 131. In an exemplary embodiment, the generation user interface 129 may serve as an internal front end for case analysts and quality control of the analytics web service 123. The generation user interface 129 may also serve as a portal to internal processing applications for the health data analytics system 105. In one scenario, the generation user interface 129 may provide analysts with the ability to process cases, manage priorities, and report and review cases before they are delivered to customers. The generation user interface 129 may also facilitate analysts validating analysis results, for example, by identifying analysis results that may have relatively low statistical confidence and prompting the analyst to review (and prioritize the review of) the analysis results. The customer support user interface 131 may serve as an internal front end for site creation, user management, and dashboarding. The customer support user interface 131 may be used by support staff to monitor and manage locations and facilities, for example, by the connection system 107 (shown in FIG. 1B). In one embodiment, the support user interface 131 may also manage data quarantine, for example, for product testing, development, or other tool development functions. For example, a case may have one or more tags, for example, quarantine, restricted, or general. These tags may be generated when a case is created. Tags may be used or updated after each case report is completed to see if raw or analyzed data can be used for future development / analysis. The customer support user interface 131 may also generate reports of cases that are processed over a given or selected period of time for internal use or accounting to trigger reports of cases that are generated free of charge, by contract, for payment, for research, etc.

[0059] 2 illustrates a detailed block diagram 200 of the health data analysis system 105 of FIG. 1C in accordance with an exemplary embodiment of the present disclosure. In an exemplary embodiment, the connection system 107 may extract patient privacy information from the DICOM objects of the case files such that the case files are anonymous. The connection system 107 may then transmit the anonymous case files to the health data analysis system 105 using HTTP or HTTPS Secure Hypertext Transfer Protocol (HTTPS). In one embodiment, the connection system 107 may create a unique case identifier for the anonymous case files such that the anonymous case files may be tracked using the unique case identifier rather than using the patient privacy information.

[0060] In an exemplary embodiment, the case processor 121 may process the DICOM objects of each case using a preload processor 202 (modeling workstation 127), a preanalysis processor 204, an analysis processor 206, a postanalysis processor 207, and a report processor 208. In an exemplary embodiment, cases may be processed through the preload processor 202, the preanalysis processor 204, the analysis processor 206, the postanalysis processor 207, and the report processor 208 according to the set priority of the cases. In an exemplary embodiment, a hospital clinician or administrator may input priority settings for one or more DICOM object source CT scanners, CT workstations, or PACS using a web interface. In an alternative embodiment, the priority may be set automatically according to the location of the hospital devices within the hospital. The priority may be transmitted to the case processor 121 using a stateless protocol. At each stage or process, cases may be queued according to priority, with higher priority cases being processed before lower priority cases. Alternatively or in addition, case priority may be edited at any stage of the process.

[0061] In one embodiment, the preload processor 202 may prepare or verify that the received DICOM file is ready for analysis. For example, the preload processor 202 may decompress (or decompress) the DICOM file received from the connectivity system 107. In one embodiment, the connectivity system 107 may also have the ability to decompress the DICOM file, but the preload processor 202 may provide a more secure environment for the connectivity system 107 to decompress the DICOM file. The preload processor 202 may also perform a verification of the quality of the DICOM file as meeting a minimum threshold required for the requested analysis to be performed by the health data analysis system 105. For example, a blood flow analysis may involve CT image quality meeting a minimum pre-determined threshold for the resulting blood flow metric to be clinically accurate. A plaque detection analysis may have a lower pre-determined threshold of image quality for the results to be useful to the patient. In one embodiment, the connectivity system 107 may automatically reject the received health data. As another embodiment, the preload processor 202 may also have the ability to automatically reject the received health data. For example, the pre-load processor 202 may reject a case if the health data has missing DICOM images, incorrect image spacing, incorrect z-positioning, etc. If the DICOM object fails to meet the quality assessment of the pre-load processor 202, the connectivity system 107 may receive a notification that the DICOM file is rejected for analysis or that analysis cannot be performed using the uploaded DICOM object. In an embodiment, the health data IO system 101 may be prompted to generate or upload a new or updated DICOM file for analysis. For example, the mobile user interface 109 may be prompted to display the case as "returned." As another example, the "returned" case may include an image quality rejection report that includes the reason the health data is rejected for analysis.If acceptable new health data is submitted, a new case entity may be created and any associated data may be copied to the new case / duplicate case.

[0062] In one embodiment, if the uncompressed DICOM file has an accuracy or image quality level that meets a pre-determined threshold accuracy or image quality level, the uncompressed DICOM file may be passed directly to the pre-analysis processor 204. If the file does not meet the pre-determined accuracy or image quality threshold level, the file may be transferred to the modeling workstation 127 for manual processing, for example, using a workstation to create a mesh (anatomical model). In some embodiments, the mesh model created by the modeling workstation 127 may then be transferred to the pre-load processor 202.

[0063] In one embodiment, the pre-analysis processor 204 may perform preliminary calculations and convert the image files into a form to be used for the requested analysis. For example, the blood flow analysis of the health data analysis system 105 may be based on a three-dimensional ("3D") model of the patient's vasculature. The pre-analysis processor 204 may convert the raw DICOM image files into a 3D solid model for blood flow analysis. In one embodiment, the pre-analysis processor 204 may further generate a mesh of tetrahedral elements and / or identify initial boundary conditions to be used for the analysis. In one embodiment, the pre-analysis processor 204 may communicate with the modeling workstation 127 and render the model for verification by the analyst of the health data analysis system 105. If the requested analysis includes blood flow simulation, the pre-analysis processor 204 may determine initial boundary conditions along various locations of the 3D model of the patient's vasculature.

[0064] In one embodiment, the analysis processor 206 may launch an analysis of the prepared model rendered by the pre-analysis processor 204. For example, the analysis processor 206 may generate a blood flow simulation through a 3D model generated by the pre-analysis processor 204. In one embodiment, the analysis processor 206 may interface with an analysis web service 123 to perform the analysis. In one embodiment, the analysis processor 206 may include a post-analysis processor that may aggregate the analysis performed by the analysis processor 206 and package the aggregated analysis for the reporting processor 208. For example, the post-analysis processor may identify a previous or stored analysis that is related to the current analysis performed by the analysis processor 206. The analyses may be related via the type of analysis performed (e.g., blood flow simulation or perfusion simulation), the analysis linked by association with the patient or patient demographics, etc. (The data may be anonymized (e.g., dissociated from patient privacy information), but various sets of data may still be identified as associated with the same individual by tracking or hashing a unique ID associated with the anonymized data.) The analysis may include user-specific images and models of blood flow and possible treatment outcomes. In one embodiment, the report processor 208 may generate views, displays, pins, PDFs, and / or interactive interfaces and models that may constitute a final report of the assessment performed by the analysis processor (and post-analysis processor) 206. The report processor 208 may allow review and revisions to be made to the analysis generated by the analysis processor 206.

[0065] In one embodiment, each component of the case processor may have a corresponding queue monitor 201 (e.g., pre-load queue processor 203 corresponding to pre-load processor 202, pre-analysis queue processor 205 corresponding to pre-analysis processor 204, analysis queue processor 207a and post-analysis queue processor 207b corresponding to analysis processor 206, and report queue processor 209 corresponding to report processor 208). The queue monitor 201 may progress the cases through each of their corresponding components. For example, pre-load queue processor 203 may dictate the progress of each image dataset through pre-load processor 202. In one embodiment, an API may be used to capture the status of each case as it progresses through the states of analysis.

[0066] Each of the queue monitors may periodically poll for indications of cases available from their respective components of the case processor and download the relevant data from storage. For example, the preload queue processor 203 may detect that a new DICOM data set has been uploaded. The preload queue processor 203 may then prompt the preload processor 202 to download the new data set and verify that the data set meets image quality requirements. The preload queue processor 203 may detect from the preload processor 202 that the data set meets image quality standards and initiate an action from the pre-analysis processor 204 to generate a geometric or volumetric model based on the data set. Similarly, the analysis queue processor 207a may detect that a geometric or volumetric model is available and suitable for analysis and therefore prompt the analysis processor 206 (and analysis web services 123) to initiate an analysis that simulates blood flow through a geometric or volumetric model. In one embodiment, the post-analysis queue processor 207b may detect the completion of an analysis by the analysis processor 206. The post-analysis queue processor 207b may then prompt the post-analysis processor 207 to validate the results. In some embodiments, this validation may include submitting the results to the modeling workstation 127 for an analyst to verify the analysis outcome. The report queue processor 209 may trigger the creation of a report for the report processor 208 from the validated results.

[0067] In one embodiment, the queue monitor 201 may further process one or more received cases (e.g., DICOM objects) according to one or more priorities set for the received cases. For example, the priorities may be set by the health data IO system 101 and / or the priorities may be set in association with the analysis requested for the health data. The queue monitor may manage each stage of the data analysis according to a priority assigned to each case. For example, a low priority case in the pre-analysis processor stage may remain in the pre-analysis processor stage while a subsequent high priority case is routed to the analysis processor stage before the low priority case. The queue monitor further monitors the bandwidth of the analysis web service 123 to ensure that cases are processed at a steady pace.

[0068] The supervisory processor 210 of FIG. 2 may be responsible for monitoring the status of case processing and analysis. The supervisory processor 210 may monitor the status of the case (e.g., success, failure, etc.) at each level as the case is processed through the preload processor 202, the pre-analysis processor 204, the analysis processor 206, and the reporting processor 208. In an exemplary embodiment, the supervisory processor 210 may be invoked or disabled by one or more authorized users using remote commands. The supervisory processor 210 may execute commands under the context of a user that is specifically created with limited rights for the sole purpose of executing remote commands. The supervisory processor 210 may notify one or more APIs associated with the front-end user interface about the status of the case (e.g., DICOM object analysis). The supervisory processor 210 may generate one or more automatic processing events using one or more status messages (e.g., success, failure, etc.) generated by a unique case at one or more stages of processing.

[0069] FIG. 3 is an exemplary block diagram 300 of load balancing for bridging communications between a front end (clinical user interface (e.g., including mobile user interface 109 or customer support user interface 111, web services 112, analytics web services 123, or any cloud, local, analytics web services)), database, and back end (e.g., web services or case processors) according to an exemplary embodiment of the present disclosure. In one embodiment, the health data management system 100 may horizontally scale its APIs 307 (e.g., APIs 307a-307c). For example, the health data management system 100 may employ an elastic load balancer 301 that may use thin multiple proxy machines 303 provided behind a load balancer 305 that may efficiently balance network load across the APIs 307. Alternatively or in addition, the health data management system 100 may include advanced machine hardware to handle the load encountered by the health data management system 100.

[0070] 4 depicts a sequence diagram of an exemplary method 400 of receiving, analyzing, and reporting health data according to an exemplary embodiment of the present disclosure. In an exemplary embodiment, a DICOM object may be pushed from one or more hospital computing devices 106 to the connectivity system 107. Once the connectivity system 107 receives the DICOM object (e.g., step 401), the connectivity system 107 may create a new case file associated with the DICOM object. This case file may be monitored throughout the progress of the analysis of the DICOM object. The case file may also be monitored throughout the analysis until a final report is available.

[0071] In one embodiment, the connection system 107 may extract patient privacy information (e.g., PHI data) from the DICOM object (e.g., step 403). In doing so, the DICOM object may be anonymized. In one embodiment, the patient privacy information may be sent to a domain-specific cloud platform (e.g., web services 112) using a service that may allow stateless operation. For example, the patient privacy information may be sent to the domain-specific cloud platform via a Representational State Transfer (REST) ​​Application Programming Interface (API). The domain-specific cloud platform may store the patient privacy information in a local relational database (e.g., a database that is structured to recognize relationships between stored items of information). The PHI database 114 is one example of such a local relational database. Various systems and methods for anonymizing DICOM objects are disclosed in U.S. Non-Provisional Application No. 15 / 635,127, filed June 27, 2017, and entitled "Systems and Methods for Modifying and Reacting Health Data Across Geographic Regions," which is incorporated herein by reference in its entirety.

[0072] In one embodiment, step 405 may include transmitting the de-identified DICOM object to a pre-load processor (e.g., pre-load processor 202). In one embodiment, either the connection system 107 and / or the pre-load processor 202 may unpack the DICOM object and validate it according to pre-configured validation criteria. The validation criteria may depend on the analysis to be performed on the DICOM object.

[0073] In one embodiment, the preload processor 202 may further prepare a volumetric image file from the received DICOM objects (e.g., step 407). In one embodiment, the preload processor 202 may further generate a compressed archive file that may be associated with one or more volumetric image files. In one embodiment, the compressed archive file may contain information corresponding to the volumetric image files. The information may be related to the analysis of the volumetric image files. For example, the compressed archive file may include a user log, task status, types of processing that may be performed using the volumetric image file, and the like. The user log may include a history of the file, including previous processing performed on the volumetric image file, actions performed on the model by the user, and the like. The task status may include the status of each task for generating the volumetric image file, and the time spent on each task. Exemplary tasks may include the following: identifying the aorta, identifying vascular landmarks, identifying myocardial tissue, identifying the midline, labeling the vasculature, refining the lumen geometry, finalizing the model, and the like. The type of processing that may be performed using the volumetric image file may be dictated by the health data IO system 101, a particular user of the health data IO system 101, or the health data analysis system 105, or by the type of data received. For example, step 407 may include correcting misregistration in the received data, or having a pre-configured action, e.g., further processing the file if the received data contains only one object phase. In one embodiment, operations on the validated DICOM object or volumetric image file prepare the validated DICOM object for analysis (e.g., cleaning by the health data analysis system 105 or by the modeling workstation 127). In one embodiment, the processed image may be sent to the modeling workstation 127 for additional checks of image quality. This additional check may lead to a denial of image validation.In one embodiment, the connection system 107 may perform a first level of validation of the received images, and the pre-load processor 202 may perform image-related validation (e.g., checking whether any anatomical intervals or thicknesses are incorrect in the received data). In one embodiment, the modeling workstation 127 may further review the image data, for example, for whoosh sounds or image artifacts. In one embodiment, a correspondence or association between the compressed archive file and the volumetric image file may be maintained using a checksum. The checksum may be calculated by the pre-load processor 202, the connection system 107, the API, or any component of the system 100. In one embodiment, the volumetric image files of the pre-load processor 202 may be presented to an analyst, for example, via the modeling workstation 127. For example, step 409 may include presenting one or more volumetric image files using a user interface of the modeling workstation 127. In one embodiment, the modeling workstation 127 may be used to generate a patient-specific anatomical model (e.g., step 411). The model may include a three-dimensional (3D) geometry of the patient's vasculature (e.g., a 3D mesh model of the patient's coronary artery tree). In one embodiment, the patient-specific 3D model may be generated from volumetric image files using one or more automated image analysis algorithms.

[0074] In one embodiment, the modeling workstation 127 may present the 3D model to the analyst for review (e.g., via the generating user interface 129). The modeling workstation 127 may present the 3D model in a series of verification screens and prompt the analyst to approve or disapprove various portions of the model. In one embodiment, the modeling workstation 127 may receive input from the analyst and store the analyst's responses in a compressed archive file (e.g., as part of a user log). The input may include case notes from the analyst. In an exemplary embodiment, the modeling workstation 127 may prompt for a re-upload of DICOM objects if the model is not approved. For example, one or more hospital computing devices 106 may receive a prompt (e.g., via the mobile user interface 109) to generate additional DICOM objects. In an alternative embodiment, the modeling workstation 127 may communicate a request to the pre-load processor 202 to regenerate the volumetric image file. In an exemplary embodiment, the modeling workstation 127 may convert the 3D model and case notes, for example, into an Extensible Markup Language (XML) file. Output from the modeling workstation 127 may be used in the analysis (eg, to simulate blood flow and generate one or more blood flow metrics).

[0075] Once the 3D model is approved, a pre-analysis processor (e.g., pre-analysis processor 204) may perform one or more operations to further prepare it for analysis by the health data analysis system 105. In one embodiment, the pre-analysis processor may receive the approved 3D model (e.g., step 413) and generate boundary conditions for blood flow simulation through the 3D model (e.g., step 415). In one embodiment, the pre-analysis processor may use one or more analysis algorithms to generate a mesh of tetrahedral elements from the 3D model (which may be in an XML file). The pre-analysis processor 204 may calculate a set of initial boundary conditions to be used as input to the blood flow simulation calculation. The pre-analysis processor 204 may then communicate the initial boundary conditions to the analysis processor (e.g., step 417).

[0076] In one embodiment, an analysis processor (e.g., analysis processor 206) may perform the analysis (e.g., employing analysis web services 123). For example, analysis processor 206 may perform a blood flow simulation (e.g., step 419) using a three-dimensional (3D) mesh model of tetrahedral elements and initial boundary conditions. In one scenario, analysis processor 206 may calculate one or more blood flow metrics based on the blood flow simulation. The blood flow metrics may include fractional flow reserve (FFR), blood pressure, blood flow velocity, perfusion, plaque vulnerability metrics, etc.

[0077] In one embodiment, the analysis results of the analysis processor 206 may be received and compiled by a post-analysis processor (e.g., post-analysis processor 207). For example, the analysis processor 206 may send the calculated blood flow metrics from the blood flow simulation analysis as results to the post-analysis processor (e.g., step 421). In one embodiment, the post-analysis processor (e.g., post-analysis processor 207) may aggregate one or more files generated by the analysis processor 206 (e.g., step 423). The post-analysis processor 207 may further remove items from the analysis files that are not desired in the final report, including notes or pins in the simulation that may not contribute to the final results.

[0078] Alternatively or in addition, a reporting processor (e.g., reporting processor 208) may aggregate the results data and prepare a report for the completed analysis (e.g., step 423). For example, reporting processor 208 may generate a report of the completed analysis including an initial view and a PDF file with automatic pins pointing to areas of interest. For example, reporting processor 208 may identify areas of stenosis in the modeled vasculature (e.g., based on an analysis including blood flow simulation or calculation).

[0079] In one embodiment, steps 427 and 429 may include presenting the generated report and pins for review and editing (e.g., via the generating user interface 129). Annotations may be automatically generated by the report processor 208. In one embodiment, the 3D model report may be modified, for example, in the generating user interface 129. For example, the generating user interface 129 may prompt the analyst to review the report and adjust the location of one or more pins for relevant areas (e.g., main vessels) in various model views. In addition, the generating user interface 129 may provide the ability to add more pins or notes to the report, create relevant views, and generate a PDF report containing the details of the analysis (e.g., blood flow simulation). The PDF report may be provided to the health data IO system 101 for storage (e.g., in the report database 113 of FIG. 1B) and access (e.g., via one or more mobile user interfaces 109). In addition, analysis may be prompted if the analyst is not satisfied with the results. In an exemplary embodiment, re-triggering the analysis may transition the case back to any of the preload, modeling, pre-analysis, analysis, post-analysis, or reporting states (e.g., step 431). In one embodiment, manual intervention (e.g., via the modeling workstation 127) may be bypassed if the case of received raw data meets a preset required accuracy or image quality level. After passing through the pre-analysis processor 204, analysis processor 206, and post-analysis processor 207 operations, a report may be delivered to the customer (e.g., by the report processor 208). Such an embodiment may bypass further scrutiny (e.g., of the report or pins within the report by the modeling workstation 127).

[0080] In an exemplary embodiment, the case monitor (e.g., case processor 121) 429 may notify the API of the health data IO system 101 that the report is complete. In one embodiment, the report may be matched with patient privacy information (e.g., PHI in the PHI database 114) and the patient privacy information may be integrated into the report. The report may then be pushed to a requesting hospital workstation, PACS, or HL7 engine / broker to be accessed by the patient and / or healthcare personnel.

[0081] FIG. 5 is a flow diagram of an exemplary method for managing the order of case processing (by the health data analysis system 105) according to an exemplary embodiment of the present disclosure. In an embodiment, steps 501 and 503 may include receiving a first de-identified DICOM object and identifying a priority level to be associated with the first de-identified DICOM object. The de-identified DICOM object may include one or more image files that are devoid of patient privacy information. In an embodiment, the priority level may be associated with the health data IO system 101 that generated the DICOM object. For example, various priority levels may be configured upon establishment of communication between the health data IO system 101 and the health data analysis system 105. The priority level may further be set by the analysis or by a healthcare provider requesting the analysis. For example, the health data IO system 101 may prompt a user to set a priority level to be associated with the generated DICOM object.

[0082] In one embodiment, steps 505 and 507 may include receiving a second de-identified DICOM object and identifying a priority level associated with the second de-identified DICOM object. Step 509 may include determining which case to process first (e.g., in the health data analysis system 105). In one embodiment, step 509 may include comparing a priority level associated with the first de-identified DICOM object against a priority level associated with the second de-identified DICOM object. If the priority level of the first de-identified DICOM object is higher than the priority level associated with the second de-identified DICOM object, analysis may be initiated on the first de-identified DICOM object prior to analysis of the second de-identified DICOM object (e.g., step 511). For example, step 511 may include a queue monitor 201 (e.g., of FIG. 2) queuing the first de-identified DICOM object for processing through various components of the case processor 121 prior to pushing the second de-identified DICOM object through the progression of steps of the case processor 121. For example, the first de-identified DICOM object may undergo one of the steps of validation by the pre-load processor 202, preparation for analysis by the pre-analysis processor 204, analysis by the analysis processor 206, or a combination thereof prior to the second de-identified DICOM object. A report of the analysis based on the first de-identified DICOM object may also be generated and made available to the health data IO system 101 prior to completion of the final report based on the second de-identified DICOM object.

[0083] If the priority level of the first de-identified DICOM object is lower than the priority level associated with the second de-identified DICOM object, analysis may be initiated on the second de-identified DICOM object prior to analysis of the first de-identified DICOM object (e.g., step 513). For example, the second de-identified DICOM object may undergo at least one of processing through the pre-load processor 202, the pre-analysis processor 204, the analysis processor 206, the reporting processor 208, or a combination thereof prior to the first de-identified DICOM object.

[0084] 6 is a flow diagram of an example method for efficiently analyzing cross-border health data via a cloud computing platform, according to an example embodiment of the present disclosure. In one embodiment, step 601 may include receiving the health data as a unique case file containing one or more DICOM objects. The case file may include one or more priority levels associated therewith, which may dictate the order in which the cases go through analysis in the health data analysis system 105. In one embodiment, a unique case identifier may be generated for the received health data (e.g., DICOM objects).

[0085] In one embodiment, step 603 may include decompressing and extracting DICOM images from the received data. In one embodiment, step 603 may include generating a compressed archive file from the extracted DICOM images for each unique case. The compressed archive file may include a customized file containing a header file, a set of selection files, a history file (e.g., user log), a task status file, a process ID file, a set of folders (n>=1) representing a set of calculations that have been processed, etc. The compressed archive file may also include one or more volumetric image files depending on the health data received. For example, if the health data is generated by a CT scanner, the compressed archive file may include one or more volumetric image files comprising a CT scan.

[0086] In one embodiment, step 603 may include validating one or more DICOM objects of the received unique case. For example, a DICOM object may be considered invalid if the DICOM object is missing a DICOM image, if the DICOM object's DICOM image size is outside a defined range, or if the DICOM image does not meet a defined quality criterion. In one embodiment, step 603 may include setting one or more quality standards to be associated with each analysis available from the health data analysis system 105 (e.g., via the analysis web service 123). For example, if the analysis provided by the health data analysis system 105 includes performing a blood flow simulation based on an anatomical model constructed from the DICOM object, step 603 may include a minimum image quality requirement for constructing the anatomical model that will render a simulation result that can be used clinically. Step 603 may further include generating a confidence parameter based on the image quality parameter. For example, step 603 may include assessing the image quality of the received DICOM object and comparing the image quality score of the received DICOM object against pre-determined image quality requirements. Step 603 may further include determining that the image quality score of the DICOM object meets the predetermined image quality requirement and generating a metadata tag including a confidence score for the blood flow simulation result based on determining the image quality score of the DICOM object. Step 603 may further include determining that the received DICOM object is invalid. Step 603 may then include generating a prompt to recapture or resubmit health data for the use case. The prompt may be sent directly to the hospital computing device 106 or the prompt may be sent to the user via a user interface.

[0087] In one embodiment, step 605 may include pre-processing the DICOM object. For example, step 605 may include determining an analysis to be performed using the received DICOM object, and step 605 may further include pre-processing the DICOM object in preparation for the determined analysis. For example, if the analysis includes blood flow simulation, the pre-processing step may include generating an anatomical model based on the DICOM object. The anatomical model may include a 3D mesh model. In one scenario, the pre-processing may include extracting a DICOM image from the received DICOM object and generating a 3D volumetric model from the extracted DICOM image. The blood flow simulation analysis may then be performed by simulating blood flow through the generated 3D mesh model. Various systems and methods for simulating blood flow are disclosed, for example, in U.S. Patent No. 8,315,812, filed Jan. 25, 2011, entitled "Method and System for Patient-Specific Modeling of Blood Flow," which is incorporated herein by reference in its entirety. In one embodiment, step 605 may include determining a desired pre-processing for each analysis available from the health data analysis system 105. Step 605 may further include determining a use for the results of the pre-processing. For example, the generated 3D mesh model may be used for analyses other than blood flow simulation. Alternatively, or in addition, a 3D mesh model used for an analysis related to one metric (e.g., blood pressure) may be used in an analysis related to a different metric (e.g., plaque vulnerability). In one scenario, step 605 may also include determining a degree of pre-processing desired for an analysis.For example, a DICOM object may be used to complement a 3D mesh model associated with a case's unique case ID, or a DICOM object may be used to generate a new 3D mesh model if the unique case identifier does not already have a model associated with it.

[0088] In one embodiment, the 3D mesh model may be in the form of a 3D volumetric file created using the Representational State Transfer (REST) ​​HTTP protocol. In an exemplary embodiment, the volumetric image file may include a header, binary data, a custom header (e.g., a tracking mechanism or identifier supported by analysis web services 123), etc. The header may include a set of unique identifications (UIDs). The UIDs may be retrieved from a DICOM object header associated with the case file under analysis.

[0089] In one embodiment, each volumetric image file of image data may correspond to a compressed archive file that stores information about a 3D model generated based on the image data of the volumetric image file. The compressed archive file may be further converted to a file that includes a representation of the 3D model (e.g., an Extensible Markup Language (XML) file) that may be both human-readable and machine-readable. The file may include the 3D model mesh and other information including midlines, vessel boundaries, vessel diameters / radii, tissue boundaries, pins (e.g., markings made by the analyst), image quality scores, confidence metrics, etc.

[0090] In one embodiment, each volumetric image file may contain a file checksum. The file checksum may be used to correlate a set of UIDs with model data (e.g., lumen segmentation and / or myocardial mass), to associate the compressed archive file with one or more volumetric image files (e.g., for longitudinal studies or data tracking), and to ensure that one or more volumetric image files are not corrupted over time. In one embodiment, the checksum may be calculated using an off-the-shelf algorithm in conjunction with the raw data portion of the volumetric image file. An identifier for the checksum algorithm may be contained in the checksum itself. This sequence may be used to accommodate different versions or upgrades of the checksum algorithm. In one embodiment, the compressed archive file may include a volumetric image file checksum to maintain a consistent link between the compressed archive file and the image data. In one embodiment, various models may be created from the image data, for example, using different segmentation algorithms, focusing on different portions of the image data, etc.

[0091] In one embodiment, step 607 may include prompting for analysis of the results of the pre-processing. For example, if the pre-processing of step 605 generated a 3D model, step 607 may include prompting for analysis using the 3D model. In one embodiment, step 607 may be performed according to a set priority of the case file as per FIG. 5. In one embodiment, step 607 may include sending one or more pre-processed anonymous DICOM objects to the analysis web service 123 for analysis. In an exemplary embodiment, the modeling workstation may calculate a checksum for the volumetric image file if the checksum is not already included in the transmitted volumetric image file. The checksum may be used to match the volumetric image file to the compressed archive file. For example, the volumetric image file may be named based on its unique checksum, thus facilitating correlation with information contained within the compressed archive file. For example, the modeling workstation may retrieve a list of volumetric image files associated with a given compressed archive file.

[0092] FIG. 7 is a flow diagram of an exemplary method 700 for post-processing a result file from a health data analysis system 105, according to an exemplary embodiment of the present disclosure. In an exemplary embodiment, step 701 may include receiving an interim result report from the analysis web service 123. In an embodiment, a pre-determined format for presenting the final result report may be pre-determined, and any items in the interim result report that are not part of the final result report may be removed. For example, the analysis web service 123 may return several simulations of stent insertion to treat a stenosis in the patient's vasculature. The multiple simulation results may be removed such that the interim result report shows only the most conservative simulation (e.g., the stent insertion that is the least invasive and may generate the highest hemodynamic impact). In an embodiment, step 701 may further include identifying whether there are previous results from the analysis web service 123 associated with the interim result report, and aggregating the interim result report with the previous results.

[0093] In one embodiment, step 703 may include presenting the interim / aggregated report to an analyst for review (using an interactive user interface, e.g., modeling workstation 127 or generation user interface 129 of FIG. 1C). In one embodiment, step 705 may include receiving one or more changes / modifications to the interim report, e.g., edits to views of the interim report, pins, simulations, etc. In one embodiment, step 707 may include re-triggering the analysis and generation of the interim report.

[0094] In an exemplary embodiment, step 707 may include generating a final outcome report incorporating one or more changes / modifications provided by the analyst via the user interface. In an exemplary embodiment, the final outcome report may include a PDF document and / or an interactive 3D model with pins or marks highlighting relevant areas of the model and solutions from the analysis. For example, the final outcome report may include a 3D model of the patient's anatomy, blood flow metrics calculated based on the 3D model, a comparison of the hemodynamic impact of various geometric modifications to the 3D model (e.g., from stent or graft insertion, changes in physiological conditions, medications, etc.). In one embodiment, the final outcome report may be stored (e.g., in anonymous persistent storage 124b of FIG. 1C ) for machine learning purposes for longitudinal studies and / or various analyses of the analysis web service 123.

[0095] In one embodiment, step 709 may include pushing the final result report to the health data IO system 101 (e.g., via a predefined network address of the hospital system or a web interface of the health data IO system 101). In an exemplary embodiment, step 709 may include retrieving the predefined network address from one or more DICOM objects used to generate the final result report. In an exemplary embodiment, a user (e.g., a hospital technician) may request a result report for a case using a web application (e.g., an online portal) and a unique case identifier. The result report may be pushed to the health data IO system 101 in a specific format (e.g., PDF format or 3D model format). The health data IO system 101 may receive a request for a result report from a user through HTTP protocol.

[0096] In an alternative embodiment, the health data analysis system 105 may push the result report to a hospital computing system (e.g., PACS or CT workstation) when the result report is ready. For example, the case status may be monitored (e.g., by a queue monitor) and the result may be pushed to a hospital system or clinical user interface (e.g., of the health data IO system 101) once the case is transitioned to a completed state. In an exemplary embodiment, pushing the health data report to at least one of the hospital computing device, the hospital electronic record server, and the preconfigured destination network address may include sending an HL7 standard message to at least one of the hospital computing device, the hospital electronic record server, and the preconfigured destination network address. In an embodiment, the HL7 standard message may contain the health data report as an encapsulated data type or a URL.

[0097] The present disclosure includes systems and methods for analyzing, storing, and managing patient health data in different geographic regions while protecting patient privacy data. The present disclosure further includes systems and methods for sending anonymous health data to an analytics cloud platform outside the hospital's region and transmitting analysis results reports. Thus, data analysis can be provided across regions while avoiding the transfer of privacy data from one region to another. Any of the described embodiments may be modified to include, for example, variations in data that may be kept within a region. The disclosed systems and methods may be modified to model and assess any range of changes to circulation.

[0098] The functions of one or more processors associated with one or more cloud-based web services may be provided by a single dedicated processor or by multiple processors. Additionally, the processor may include, but is not limited to, digital signal processor (DSP) hardware, or any other hardware capable of executing software. The program aspects of the technology may be considered as a "product" or "article of manufacture" typically in the form of executable code and / or associated data carried on or embodied in some type of machine-readable medium. "Storage" type media includes any or all of the tangible memory of a computer, processor, or equivalent, or its associated modules, such as various semiconductor memories, tape drives, disk drives, and the like, that may provide non-transitory storage at any time for software programming.

[0099] All or part of the software may be communicated from time to time through the Internet or various other telecommunications networks. Such communication may, for example, enable loading of the software from one computer or processor into another, for example, from a management server or host computer of a mobile communications network into a computer platform of a server, and / or from a server into a mobile device. Thus, other types of media that may bear software elements include optical, electrical, and electromagnetic waves, such as those used across physical interfaces between local devices, through wired and optical landline networks, and via various air links. Physical elements that carry such waves, such as wired or wireless links, optical links, or the like, may also be considered media that bear the software. As used herein, unless limited to non-transitory tangible "storage" media, terms such as computer or machine "readable medium" refer to any medium that participates in providing instructions to a processor for execution.

[0100] Other embodiments of the invention will be apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the invention being indicated by the following claims.

Claims

1. 1. A computer-implemented method for analyzing health data via a distributed cloud computing platform, the method comprising: receiving a unique case file containing one or more anonymous DICOM objects for analysis; setting a priority level for the unique case file based on a priority associated with the one or more anonymous DICOM objects; decompressing and validating at least one of the one or more anonymous DICOM objects; transmitting an analysis of at least one of the unpacked and validated anonymous DICOM objects, the analysis being completed according to a priority level of the unique case file; A method comprising:

2. The analysis generating a volumetric image file based on at least one of the anonymous DICOM objects of the unique case file, the volumetric image file representing a three-dimensional (3D) volume; generating an archive file containing analysis data associated with said volumetric image file; generating a patient-specific model based on the volumetric image file and the archive file; The computer-implemented method of claim 1 , comprising:

3. The computer-implemented method of claim 2 , wherein the volumetric image file includes a file checksum that correlates the volumetric image file to the archive file.

4. monitoring a status of the unique case file; advancing an analysis of the unique case file based on a status and a priority level of the unique case file; The computer-implemented method of claim 1 , further comprising:

5. 2. The computer-implemented method of claim 1, further comprising setting the priority level for the unique case file based on an entity that generates one or more DICOM objects associated with the one or more anonymous DICOM objects and / or an entity that analyzes the one or more anonymous DICOM objects.

6. generating a preliminary result report based on the received analysis; and generating an interactive user interface of said interim result report; The computer-implemented method of claim 1 , further comprising:

7. receiving modifications to the interim result report; generating a final result report based on said modifications; and storing said final result report, a predetermined identified network address, or a requested web interface; The computer-implemented method of claim 6 , further comprising:

8. 8. The computer-implemented method of claim 7, wherein the final result report is pushed to a database, a pre-identified network address, or a requested web interface using secure hypertext transfer protocol (HTTPS).

9. 1. A system for analyzing health data via a distributed cloud computing platform, the system comprising: one or more data storage devices storing instructions for analyzing health data via a distributed cloud computing platform; one or more processors, the one or more processors executing the instructions; receiving a unique case file containing one or more anonymous DICOM objects for analysis; setting a priority level for the unique case file based on a priority associated with the one or more anonymous DICOM objects; decompressing and validating at least one of the one or more anonymous DICOM objects; transmitting an analysis of at least one of the unpacked and validated anonymous DICOM objects, the analysis being completed according to a priority level of the unique case file; one or more processors configured to perform a method including: A system comprising:

10. The analysis further comprises: generating a volumetric image file based on at least one of the anonymous DICOM objects of the unique case file, the volumetric image file representing a three-dimensional (3D) volume; generating an archive file containing analysis data associated with said volumetric image file; generating a patient-specific model based on the volumetric image file and the archive file; The system of claim 9 , comprising:

11. The system of claim 10 , wherein the volumetric image file includes a file checksum that correlates the volumetric image file to the archive file.

12. The system further comprises: monitoring a status of the unique case file; advancing an analysis of the unique case file based on a status and a priority level of the unique case file; The system of claim 9 , configured to:

13. 10. The system of claim 9, further configured to set the priority level for the unique case file based on an entity that generates one or more DICOM objects associated with the one or more anonymous DICOM objects and / or an entity that analyzes the one or more anonymous DICOM objects.

14. The system further comprises: generating a preliminary result report based on the received analysis; and generating an interactive user interface of said interim result report; The system of claim 9 , configured to:

15. The system further comprises: receiving modifications to the interim result report; generating a final result report based on said modifications; and storing said final result report, a predetermined identified network address, or a requested web interface; The system of claim 14 configured to:

16. 16. The system of claim 15, wherein the final results report is pushed to a database, a predefined identified network address, or a requested web interface using secure hypertext transfer protocol (HTTPS).

17. 1. A non-transitory computer readable medium for use on a computer system containing computer executable programming instructions that cause operations to be performed for analyzing health data via a distributed cloud computing platform, the operations comprising: receiving a unique case file containing one or more anonymous DICOM objects for analysis; setting a priority level for the unique case file based on a priority associated with the one or more anonymous DICOM objects; decompressing and validating at least one of the one or more anonymous DICOM objects; transmitting an analysis of at least one of the unpacked and validated anonymous DICOM objects, the analysis being completed according to a priority level of the unique case file; 16. A non-transitory computer readable medium comprising:

18. The analysis generating a volumetric image file based on at least one of the anonymous DICOM objects of the unique case file, the volumetric image file representing a three-dimensional (3D) volume; generating an archive file containing analysis data associated with said volumetric image file; generating a patient-specific model based on the volumetric image file and the archive file; 20. The non-transitory computer readable medium of claim 17, comprising:

19. The operation further comprises: monitoring a status of the unique case file; advancing an analysis of the unique case file based on a status and a priority level of the unique case file; 20. The non-transitory computer readable medium of claim 17, comprising:

20. 20. The non-transitory computer-readable medium of claim 17, wherein the operations further include setting the priority level for the unique case file based on an entity that generates one or more DICOM objects associated with the one or more anonymous DICOM objects and / or an entity that analyzes the one or more anonymous DICOM objects.

Citation Information

Patent Citations

  • Electronic clinical recording system

    JP1998079770A

  • Analysis device and analysis system

    JP2017213287A

  • Systems and methods for anonymization of health data and transmition of health data for analysis across geographic regions

    WO2018005562A1

Cited By

  • Systems, methods, and devices for medical image analysis, diagnosis, risk stratification, decision making and / or disease tracking

    US12396695B2

  • Systems, devices, and methods for non-invasive image-based plaque analysis and risk determination

    US12440180B2

  • Systems, methods, and devices for medical image analysis, diagnosis, risk stratification, decision making and / or disease tracking

    US12499539B2

  • Systems, methods, and devices for medical image analysis, diagnosis, risk stratification, decision making and / or disease tracking

    US12555228B2

  • Systems, methods, and devices for medical image analysis, diagnosis, risk stratification, decision making and / or disease tracking

    US12558048B2