A platform for evaluating medical information and method for using the same

专利摘要:
A platform for evaluating medical information and method for using the same are described. In one embodiment, the method comprises: monitoring, by a medical image management system, for a first indication of a content change at one or more data sources; determining, in response to the first indication, which of a plurality of image analysis engines is to analyze at least one image of the one or more new medical images associated with the content change based on one or both of information accompanying the one or more images or results of applying body part detection on the at least one image; sending a first notification to start image analysis on the at least one image of the one or more medical images, the first notification being sent to each image analysis engine in a set of one or more image analysis engines determined to analyze the at least one image of the one or more new medical images; receiving a second indication that image analysis results are available from the set of one or more image analysis engines; and sending a second notification to subscribers to indicate availability of image analysis results for access and display thereby.
公开号:ES2888523A2
申请号:ES202190057
申请日:2020-02-28
公开日:2022-01-04
发明作者:Brigil Vincent；Keiji Sugihara；Iii William Benjamin Carruthers；Yoshiyuki Kurami
申请人:Fujifilm Medical Systems USA Inc；
IPC主号:

专利说明:

[0002] A PLATFORM TO EVALUATE MEDICAL INFORMATION AND A METHOD FOR
[0004] FIELD OF THE INVENTION
[0006] Embodiments of the present invention relate to the field of medical imaging; more particularly, embodiments of the present invention relate to the use of automated image analysis (eg, artificial intelligence (AI) analysis platform to analyze medical images.
[0008] BACKGROUND
[0010] Physicians and other clinical staff often review all relevant clinical information about a patient when making health care decisions. Clinical information is usually included in health studies and structured reports. These often include information about a patient's history, diagnostic reports from different domains, images, and other clinical data in electronic format.
[0012] A patient health study includes an imaging report that contains parameter values (for example, measurements, readings, etc.) and images of exams or procedures that are typically shared between physicians and clinicians to aid in diagnosis and treatment. treatment.
[0014] Health studies are usually generated in response to a doctor requesting an exam for their patient. The exam is performed and the generated study is often sent to a picture archiving and communication system (PACS). A doctor or clinician can use a medical image management system to obtain a work list containing studies for their patients.
[0016] Various Artificial Intelligence (AI) algorithms have recently been used with Radiology PACS systems. These algorithms automate the image evaluation process in health studies. These algorithms can be applied to individual images or entire studies, and the results will be made available to the interpreting physician as well as other clinical users. Although the results of algorithm are available, the interpreting physician may not be aware of the findings because the AI algorithms are implemented on different platforms and there is no organized storage, access, or data flow for the AI results. Due to these limitations, the interpreting physician may not be able to review automated results in a timely manner, which could further harm a patient or delay treatment if not reviewed with the speed associated with the priority level of findings. .
[0018] SUMMARY OF THE INVENTION
[0020] A platform for evaluating medical information and a method for using the same are described. In one embodiment, the method comprises: monitoring, by a medical image management system, a first indication of a content change in one or more data sources; determine, in response to the first prompt, which of a plurality of image analysis engines will analyze at least one of the one or more new medical images associated with the content change based on one or both of the information accompanying the or the images or the results of applying the body part detection on the at least one image; send a first notification to start image analysis on the at least one image of one or more medical images, the first notification being sent to each image analysis engine in a set of one or more determined image analysis engines to analyze the at least one image of the one or more new medical images; receiving a second indication that image analysis results are available from the set of one or more image analysis engines; and sending a second notification to subscribers to indicate the availability of the image analysis results for access and viewing.
[0022] BRIEF DESCRIPTION OF THE DRAWINGS
[0024] The present invention will be more fully understood from the following detailed description and accompanying drawings of various embodiments of the invention, which, however, should not be taken to limit the invention to specific embodiments, but are only for your explanation and understanding.
[0025] Figure 1 illustrates an example of a medical information computer system environment with which embodiments of the present invention may be implemented.
[0027] Figure 2 is a block diagram showing an embodiment of the architecture of a computer system for analyzing, studying health study information (eg, images).
[0029] Figure 3 is a data flow diagram of one embodiment of a process for analyzing medical images using automated image analysis algorithms (eg, AI algorithms) that have been applied to one or more of the medical images.
[0031] Figure 4A is a block diagram of one embodiment of an AI platform for use in performing image analysis on medical images.
[0033] Figure 4B is a data flow diagram of one embodiment of an AI platform for use in performing image analysis on medical images.
[0035] Figure 5 is a flow diagram of one embodiment of a process for processing medical images using an image analysis platform (eg, AI analysis).
[0037] Figure 6 is a flow diagram of another embodiment of a process for processing medical images using an image analysis platform (eg, AI analysis).
[0039] Figure 7 illustrates an embodiment of a logical representation of a medical information and image management system.
[0041] DETAILED DESCRIPTION
[0043] In the following description, numerous details are set forth in order to provide a more complete explanation of the present invention, however, it will be apparent to one skilled in the art that the present invention can be practiced without these specific details. In other cases, known structures and devices are shown in block diagram form, rather than in detail, to avoid obscuring the present invention.
[0045] Embodiments of the present invention are directed to systems and methods for performing an image analysis workflow with a platform for analyzing medical images. In one embodiment, image analysis comprises artificial intelligence (AI) analysis that is used to analyze medical images as part of a workflow (for example, a Radiology workflow, a Cardiology workflow, etc. .). In one embodiment, the medical images are part of health studies and the platform is an artificial intelligence platform that is part of, or associated with, a medical image management system. In one embodiment, the AI Platform is an open API-based platform in which multiple AI algorithms are integrated to establish seamless integration for AI workflow in different branches of medicine. By having multiple AI algorithms available to analyze medical images, there is a greater chance of identifying medical conditions in patients more quickly than in the prior art. This allows healthcare studies to present critical/emerging results at the top of the list, allowing for potentially faster diagnosis. Having briefly described a general description of the present invention, embodiments of the invention will be described with reference to Figures 1-7.
[0047] The object of the embodiments of the present invention is described here with specificity to meet legal requirements. However, the description itself is not intended to limit the scope of this patent. Instead, the inventors have contemplated that the claimed subject matter could also be incorporated in other ways, to include different steps or combinations of steps similar to those described herein, along with other present or future technologies.
[0049] Having briefly described embodiments of the present invention, an example of an operating environment suitable for use in implementing embodiments of the present invention is described below.
[0051] With reference to the drawings in general, and initially to Figure 1 in particular, a medical information computer system environment with which embodiments of the present invention may be implemented is illustrated and designated generally as reference numeral 120. Those skilled in the art will understand and appreciate that the illustrated health information computer system environment 120 is merely an example of a suitable computing environment and is not intended to suggest any limitations as to scope of use or functionality. of the system of the invention. Nor should the medical information computer system environment 120 be construed as having any dependencies or requirements related to any individual component or combination of components illustrated therein.
[0053] Embodiments of the present invention may be operative with numerous general purpose or special purpose computing system environments or configurations. Examples of known computing systems, environments, and/or configurations that may be suitable for use with the present invention include, by way of example only, personal computers, server computers, portable or handheld devices, multiprocessor systems, microprocessor-based systems, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the aforementioned systems or devices, and the like.
[0055] Embodiments of the present invention can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Program modules generally include, but are not limited to, routines, programs, objects, components, and data structures that perform particular tasks or implement particular abstract data types. The present invention can also be practiced in distributed computing environments where tasks are performed by remote processing devices that are connected through a communications network. In a distributed computing environment, program modules may be located in association with local and/or remote computing storage media, including, by way of example only, memory storage devices.
[0057] With continued reference to Figure 1, the exemplary health information computing system environment 120 includes a general purpose computing device in the form of a control server 122. The components of the control server 122 may include, without limitation, a unit processing power, internal system memory, and a system bus suitable for coupling various system components, including database cluster 124, with control server 122. The system bus may be any of several types of bus structures, including a memory bus or memory controller, a peripheral bus, and a local bus, using any of a variety of bus architectures. By way of example and not limitation, such architectures include the Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MCA) bus, Enhanced ISA (EISA) bus, Video Electronic Standards Association (VESA) local bus and Peripheral Component Interconnect (PCI) bus, also known as Mezzanine bus.
[0059] The control server 122 typically includes therein, or has access to, a variety of machine-readable media, for example, database cluster 124. The machine-readable media can be any available media that the server can access. control panel 122, and includes volatile and non-volatile media, as well as removable and non-removable media. By way of example and not limitation, computer readable media may include computer storage media. Computer storage media may include, without limitation, volatile and non-volatile media, as well as removable and non-removable media implemented in any method or technology for the storage of information, such as computer-readable instructions, data structures, program modules or other data. In this sense, computer storage media may include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile discs (DVD) or other optical disc storage, magnetic cassettes, magnetic tape, magnetic disk storage, or other magnetic storage device, or any other medium that can be used to store the desired information and that can be accessed by control server 122. By way of example and not limitation, storage media Communication includes wired media, such as a wired network or direct connection, and wireless media, such as acoustic, RF, infrared, and other wireless media. Combinations of any of the foregoing may also be included within the scope of computer readable media.
[0061] The computer storage media described above and illustrated in Figure 1, including database cluster 124, provide storage of instructions, data structures, program modules, and other computer-readable data for control server 122. The server control panel 122 may operate on a computer network 126 using logical connections to one or more remote computers 128. The remote computers 128 may be located in a variety of locations. locations in a medical or research setting, for example, but not limited to, clinical laboratories (for example, molecular diagnostic laboratories), hospitals and other inpatient settings, veterinary settings, outpatient settings, medical and financial billing offices, hospital administration, home health care settings and clinical offices. Clinicians may include, but are not limited to, an attending physician(s), specialists such as intensivists, surgeons, radiologists, cardiologists and oncologists, emergency medical technicians, physician assistants, nurse practitioners, nurses, paramedics, nursing, pharmacists, dietitians, microbiologists, laboratory experts, laboratory technologists, radiologic technologists, researchers, veterinarians, students, and the like. Remote computers 128 may also be physically located in non-traditional health care settings so that the entire health care community can be integrated into the network. Remote computers 128 may be personal computers, servers, routers, network PCs, peer devices, other common network nodes, or the like, and may include some or all of the elements described above in connection with control server 122. Devices can be personal digital assistants or other similar devices.
[0063] Examples of computer networks 126 may include, without limitation, local area networks (LANs) and/or wide area networks (WANs). These network environments are common in offices, company-wide computer networks, intranets, and the Internet. When used in a WAN environment, control server 122 may include a modem or other means of establishing communications over the WAN, such as the Internet. In a network environment, program modules or portions thereof may be stored in association with control server 122, database cluster 124, or any of the remote computers 128. For example, and not by way of limitation, various application programs may reside in memory associated with one or more of the remote computers 128. Those skilled in the art will appreciate that the network connections shown are examples and other means of establishing a communications link between the computers (for example, control server 122 and remote computers 128).
[0065] During operation, a clinician may enter commands and information into the control server 122 or transmit the commands and information to the control server 122 through one or more of the remote computers 128 via communication devices. input, such as a keyboard, pointing device (commonly known as a mouse), trackball, or trackpad. Other input devices may include, without limitation, microphones, scanners, or the like. Commands and information may also be sent directly from a remote healthcare device to monitoring server 122. In addition to a monitor, monitoring server 122 and/or remote computers 128 may include other peripheral output devices, such as speakers and a monitor. printer.
[0067] Although many other internal components of the control server 122 and remote computers 128 are not shown, those skilled in the art will appreciate that such components and their interconnection are well known. Accordingly, additional details related to the internal construction of the control server 122 and remote computers 128 are not described in further detail herein.
[0069] Referring to Figure 2, a block diagram is illustrated showing an example of the architecture of a computer system for performing image analysis (eg, artificial intelligence (AI) analysis on medical images). It will be appreciated that the computer system architecture shown in Figure 2 is merely an example of a suitable computer system and is not intended to have any dependencies or requirements related to any individual module/component or combination of modules/components.
[0071] In one embodiment, the computer system includes a study analysis 200, one or more databases 230 that store and maintain unread health studies or existing health studies that contain new medical images (and potentially other health studies), and one or more databases 231 that store and maintain the findings resulting from the application of one or more automated image analysis algorithms (for example, artificial intelligence (AI) analysis algorithms) on images from unread health studies, such as those, for example, stored in databases 230. In one embodiment, databases 230 and 231 are the same set of databases.
[0073] In one embodiment, health studies include images and study data, such as, for example, values of one or more medical parameters (eg, measurements, etc.) related to the health study. Examples of medical images include radiology images, laboratory images, photographs, cardiology images, such as echocardiography images and other healthcare images. One of skill in the art will appreciate that the databases may be maintained separately or may be integrated. Databases 230 may contain images or other study data (eg, parameter values (eg, measurements)) that are linked to a patient's electronic medical record (EMR), such that the images and/or o Study data may be selected from within the EMR and displayed within a viewer via display component 222 or linked to a VNA (vendor neutral archive) which stores, images, EKG images, notes, etc. As used herein, the acronym "EMR" is not intended to be limiting and may refer broadly to any or all aspects of a patient's medical record in digital format. The EMR is typically supported by systems set up to coordinate the storage and retrieval of individual records with the help of computing devices. As such, a variety of types of health care-related information may be stored and accessed in this manner. In one embodiment, the automated image analysis algorithms are AI analysis algorithms that are performed on one or more healthcare study images. These algorithms can be applied remotely using one or more servers (for example, AI engines and applications) that are in communication with a medical image management platform. These servers receive the studies and their associated images and automatically apply the algorithms to those images. Alternatively, the AI analysis algorithms are integrated into the platform and applied locally on the images of the medical studies by the image analysis component 218 after the studies have been received by the medical image management system. Alternatively, some of the algorithms are performed remotely while others are performed locally.
[0075] AI algorithms (or other image analysis algorithms) produce findings that specify the results of applying the algorithms to images. In one embodiment, these AI algorithms produce textual findings indicating possible conditions for a patient identified by the algorithm. In one embodiment, the findings include a score (eg, abnormality score, numerical indication of the confidence level associated with the analysis results, etc.) prepared by the automated image analysis algorithm. For example, an abnormality score with a magnitude indicating the likelihood that the patient has an abnormality based on analysis performed on the images (for example, the higher the score, the higher the probability). Note that in alternate embodiments, other scores, such as diagnostic confidence levels, may be included in the results of the algorithms.
[0077] The study analysis module 200 includes a monitoring component 210 that monitors the data sources for changes in content, such as the storage of new health studies or health studies that have new medical images. These data sources may be PACS, VNA, or other repository or database systems. These health studies may come from more than one source (for example, database). Study analysis module 200 may reside on one or more computing devices, such as, for example, control server 122 described above with reference to Figure 1. By way of example, in one embodiment, control server 122 includes a computer processor and may be a server, personal computer, desktop computer, laptop computer, handheld device, mobile device, consumer electronic device, or the like.
[0079] In one embodiment, study analysis module 200 comprises selection component 212, orchestration component 214, notification component 216, image analysis (eg, AI analysis) component 218, and generation component. of results 220 of image analysis (for example, AI analysis). In various embodiments, the study analysis module 200 includes a display component 222, a history component 224, an information component 226, and a manipulation component 228. It will be appreciated that while receiving health studies stored in databases 230, study analysis module 200 can receive health studies from multiple sources, including databases distributed across multiple facilities and/or multiple locations, as well as findings resulting from the application of one or more automated image analysis algorithms (for example, AI analysis algorithms) in images from health studies. It will also be appreciated that study analysis module 200 may receive health studies with their images and/or findings resulting from automated image analysis algorithms (eg, AI analysis algorithms) from the sources described above via links within a patient's EMR.
[0081] In response to the monitoring component 210 determining that new content exists in one or more of the data sources, the selection component 212 each health study that is new or has new medical images is selected and the data sources are accessed to obtain said studies. In one embodiment, the health study comprises one or more series of images and one or more parameter values (eg, measurements, findings, impressions, patient demographics and history/risk factors, etc.). In one embodiment, each series comprises one or more images depicting the subject of the image from various angles. A list perspective within a media manager provides a list of available studies (including unread studies), images, and other media.
[0083] Orchestration component 214 determines which image analysis algorithms should be applied to analyze the new medical images. In one embodiment, orchestration component 214 determines which images of the new medical images are to be analyzed and which AI algorithms should perform the analysis. In one embodiment, the AI algorithms are performed by AI engines or AI applications of the image analysis component (eg, AI analysis) 218. In one embodiment, an AI engine comprises circuitry that implements logic and/or run software. The software may be an application. Based on the determination, the orchestration component 214 assigns the AI algorithms of the image analysis component 218 the images to evaluate.
[0085] The notification component 216 notifies the AI engines or systems (eg, servers) executing the AI algorithms regarding their assignments regarding the analysis of the new medical images. In one embodiment, the notification component 216 sends a study identifier and information about which images to evaluate as part of the notifications, and in response thereto, the image analysis component 218 retrieves the images from the study source for evaluation. data in which they are stored. Alternatively, the notification component 216 sends the images to be analyzed.
[0087] Once the images are obtained, the image analysis component 218 performs AI or other image analysis on the images to produce findings. In one embodiment, this includes activating one or more AI algorithms running on one or more AI engines or servers. One or more of the AI engines or servers may be internal to the AI platform or external to the AI platform. In one embodiment, multiple AI engines or servers may be performing their analysis at the same time on the same or different medical images. Each of the AI algorithms produces results that are indicative of the algorithm's findings. As part of the findings, the AI algorithms may include images, or parts thereof, that are relevant to the diagnosis or condition of the patient associated with the analyzed image. The image analysis result generation component 220 takes the findings of the image analysis component 218's AI algorithms and generates outputs that are sent for review and stored for later access.
[0089] Display component 222 includes a graphical display device that may be a monitor, computer screen, projection device, or other hardware device for displaying graphical user interfaces containing images and other data from health studies, as well as resulting findings. of the application of automated image analysis algorithms to images in health studies. In one embodiment, display component 222 displays the GUI with the list of unread health studies or health studies with new images along with priority information highlighting those studies with findings that are more urgent or critical. In one embodiment, the list of unread health studies is ordered based on priority. In another embodiment, the list of health studies is not ordered, but the priority information is clearly displayed so that a physician can discern priority levels from the display status of unread health studies (e.g., they have higher priority and/or or lower priority).
[0091] In one embodiment, a history component 224 displays a history of different studies and clinical images associated with more than one healthcare image. History component 224 further allows a selection of one or more history images to be displayed in the viewer by display component 222. For example, selection component 212 may have received a selection from the clinician of a particular study. However, once the display component 222 has displayed the images that make up that selected study, the history component 224 may display other clinical studies and images that are of particular interest to the clinician. The clinician can then select additional history items to launch on the viewer.
[0092] In one embodiment, information component 226 displays additional information associated with more than one health image, history, or a combination thereof. The additional information comprises patient identification information, image-related information, study-related information, or a combination thereof. Such additional information may also include time-related information.
[0094] In one embodiment, a manipulation component 228 allows a clinician to manipulate a display of a healthcare image. For example, a clinician may determine that the image as it is displayed in the viewfinder is not large enough to see the desired level of detail. The clinician can zoom in or out and the manipulation component 228 manipulates the image display accordingly. Similarly, the clinician may wish to pan an image and the manipulation component 228 manipulates the image display accordingly.
[0096] Figure 3 is a data flow diagram of one embodiment of a process for automatically analyzing images from healthcare studies using automated image analysis algorithms, (eg, AI analysis algorithm, etc.) that have been applied to one or more of the images of the health studies.
[0098] Referring to Figure 3, medical image management system 310 receives one or more health studies 301. Health studies 301 may include unread health studies or health studies with one or more new images. In one embodiment, medical imaging management system 310 receives health studies 301 in response to monitoring logic 312 that monitors data sources for new health studies and/or health studies with new images for evaluation. In one embodiment, one or more unread health studies 301 are sent from one or more medical imaging modalities that perform medical imaging (eg, cardiovascular (CV), x-ray radiography, magnetic resonance imaging, ultrasound, endoscopy, imaging). tactile, thermography, nuclear medicine functional imaging techniques such as positron emission tomography (PET) and single photon emission computed tomography (SPECT), etc.). In another embodiment, medical image management system 310 receives one or more healthcare studies 301 from a remote location. In one embodiment, the remote location may comprise one or more modalities that create the studios or a remotely located image repository (for example, a picture archiving and communication system (PACS), VNA, etc.).
[0100] In one embodiment, monitoring logic 312 monitors for a first indication of a content change in one or more data sources (for example, arrival of new medical images that are part of a new healthcare study or an existing study created before the generation of the new medical images) and obtains those studies in response to the determination that a content change has occurred. In one embodiment, monitoring logic 312 determines that a content change has occurred in response to receiving notifications from the data sources that store the health care studies or the modalities/facilities that generate the health care studies that indicate that such studies are available for review and evaluation. In response to these notifications, the monitoring logic 312 accesses the data sources that store the studies to retrieve them. Access may include a request to the data source for a copy of the study. The request can be sent directly or through a network connection to the data source. In one embodiment, the medical image management system 310 comprises a network communication interface(s) (not shown) for sending the request to the data source for a copy of the study, receiving healthcare studies, and receiving studies. In another embodiment, medical image management system 310 requests and receives studies through a direct connection to individual data sources.
[0102] In one embodiment, the medical image management system 310 comprises a memory, such as, for example, memory 323, for storing received health studies that it has received.
[0104] In one embodiment, after obtaining the health studies with medical images to be analyzed, the controller 311 (eg, one or more processors) determines which of a plurality of image analysis engines 340 (eg, one or more processors) plus artificial intelligence (AI) engines) is going to analyze at least one of the new medical images. This is referred to herein as image analysis orchestration (eg, AI analysis). To the image analysis engines of the image analysis engines 340 identified to evaluate the medical images, the controller 311 sends a notification to cause those image analysis engines to start image analysis on one or more of the image analysis engines. new medical images. In one embodiment, the determination of which image analysis engines is based on information accompanying one or more images and/or the results of applying body part detection to at least one of the new images.
[0106] In one embodiment, each of the automated analysis engines (eg, AI image analysis engines) performs analysis of images and/or non-image data from healthcare studies. Examples of non-image data include, but are not limited to, text, waveforms, time series, structured/template-based reports. In one embodiment, these engines may be integrated into the medical image management system, as shown with automated image analysis engines (eg, AI analysis engines, etc.) 340. In another embodiment, one or more of these motors 302 are located remotely from the medical image management system. In one embodiment, image analysis engines 340 produce findings or results 320. In one embodiment, individual analysis engines produce findings based solely on image analysis, while other analysis engines produce findings based on a combination of analysis. of images and data. In one embodiment, the results of the automated image analysis algorithms application of the image analysis engines 340 include an anomaly score where the higher the anomaly score number, the greater probability of an anomaly will be identified. on one or more images in a study performed by AI or other image analysis. In one embodiment, the findings of the automated image analysis algorithm application of the image analysis engines 340 may include an indication that nothing was found in the one or more images and/or in the non-image data. The results (findings) of applying these engines to the images are sent, via wired or wireless communications, to the medical image management system 310.
[0108] After analyzing the images and/or non-images and producing the findings, the image analysis engines 340 send indications that their results have been generated and send them themselves to the platform. In response, the platform sends notifications to subscribers indicating the availability of image analysis results for access and viewing.
[0109] After health studies 301 and automated image analysis (eg, image analysis engines 340 have performed AI analysis for findings, output logic 313 obtains AI and other image analysis results 320 from memory 323 and uses that information to display the AI result 331i-331n in a GUI (or viewer) 330 on display device 314. In one embodiment, the AI results are displayed along with all or part of the study it contains. the image that was evaluated by one of the AI engines 340 to create the AI result That is, the display 314 allows a user to display, within a graphical user interface 330, one or more healthcare studies, or portions of along with IA or other image analysis results, such as, for example, IA 331i-331n results.This allows a physician or other medical professional to easily view the studies that have been received and the results of the scan. the image analysis that has been performed.
[0111] In one embodiment, controller 311 also controls other operations of medical image management system 310. In one embodiment, controller 311 comprises one or more processors, microcontrollers, and/or a combination of hardware, software, and/or firmware.
[0113] Figures 4A and 4B illustrate one embodiment of an architecture for an artificial intelligence platform that is part of a medical image management system, such as the medical image management system 310 of Figure 3. In one embodiment, the architecture of the AI platform of Figures 4A and 4B includes a number of components to perform the operations described in this document, including the following features: open APIs 403, monitoring logic 402a, orchestration engines 430 to perform the orchestration of the analysis of imaging (eg AI analytics), integration of 403A AI engines into the platform, 402B study notification logic, exploration of image analytics (eg AI), use of one or more AI servers (eg , server 403B), one or more embedded artificial intelligence servers (eg, server 403A), and the result of image analysis notification logic 460 (eg, example, AI result). Each of these will be described in greater detail below. In one embodiment, AI scanning refers to one or more AI engines or algorithms that process images and produce result data. In another embodiment, exploring a study also means the entire process of dividing a study into parts and sending it to individual AI algorithms (orchestration) plus individual AI processing.
[0114] The AI platform provides open APIs 403 to applications (for example, applications 401, applications running on AI servers 403A, 403B, etc.) and AI engines (for example, 403A, 403B) to effectively perform AI evaluation within a medical workflow (for example, a radiology workflow). In one embodiment, open APIs 403 allow the use of a DICOM gateway (eg, 480, 473) to access DICOM information. In one embodiment, the DICOM gateway (eg, 480, 473) is used to access DICOM information via QIDO-RS, WADO-RS, and STOW-RS.
[0116] In one embodiment, the open APIs 403 provide a mechanism and data models for working with DICOM GSPS, DICOM Segmentation, Secondary Capture, Basic Structured Display, Key Object Selection, Structured Reporting, etc., without having to implement DICOM conformance by applications (eg 401) or AI engines (eg 403A, 403B). As described in more detail below, in one embodiment, open APIs 403 include APIs for querying, retrieving, and instantiating idea registry to help manage and access the results of image analysis (e.g., AI analysis). ) in an organized manner across multiple systems. For more information on recording ideas, see Patent Application Serial No. 14/820,144, entitled "METHODS AND APPARATUS FOR RECORDING INFORMATION USING A MEDICAL IMAGING DISPLAY SYSTEM," filed August 6, 2015.
[0118] In one embodiment, open API 403 also provides access to AI engines (eg, 403A, 403B) and scanning workflow such that applications (eg, 401) can trigger an AI scan and track the status of an AI scan or AI analysis.
[0120] In one embodiment, the AI platform is integrated with a PACS system (400A, 400B) and/or a VNA. In one embodiment, the AI platform configures the PACS and VNA systems as data sources. Therefore, when the PACS systems (400A, 400B) and the VNA receive studies (eg 425, 426), through, for example, a wired or wireless network connection, the AI platform can obtain them.
[0122] Once connected to the PACS 400A and 400B systems, the AI platform uses the 402A monitoring logic to continuously monitor content changes in the data source, including the arrival of new studies, arrival of images DICOM or non-DICOM images to an existing study, or other new image arrivals. In one embodiment, monitoring logic 402A accomplishes studio monitoring through custom integration and through standard interfaces (eg, HL7/FHIR). In one embodiment, custom integration involves monitoring logic 402A being notified that new images have arrived via a notification (eg, image arrival 427) identifies the study with the new images via, for example , a study identifier (ID), and identifies an indication of what change occurred (eg, a new study has arrived, new images added to a study along with identifying information (eg, metadata) to specify the new images between all images in the study and the type of information (e.g. one or more body parts) are displayed on the new images In one embodiment, notification of image arrival is via HTTP communication (e.g. , POST) Rather, the monitoring logic 402A also uses standard interfaces to receive an indication that there is a content change (for example, a new study or new images). are provided as images arrive 428) at a data source, such as the PACS systems 400A and 400B. In one embodiment, this indication does not include information identifying the change that had occurred. In such cases, the study is accessed and analyzed to determine the new content and which of the image analysis engines is appropriate for reviewing the new content.
[0124] Upon identifying the content change in the PACS/VNA, the AI platform 400 acquires the study and performs AI orchestration using orchestration engines 430 to decide which of the AI algorithms should be assigned to evaluate the study. In one embodiment, the orchestration engines 430 perform the orchestration using various header-based filters (eg, DICOM header-based filters) to obtain and evaluate the information stored in the image file headers. For example, header information in an image, image series, or study may specify the contents (eg, body part, exam type, etc.) that could be used to identify appropriate image analysis engines to evaluate the study. In one embodiment, the orchestration engines 430 use built-in body part detection algorithms to determine the body parts represented in the images to identify the appropriate AI engines (403A, 403B) to analyze images with those types of body parts. .
[0125] Orchestration engines 430 may identify multiple AI engines that a study must evaluate. In one embodiment, when multiple AI engines require evaluation of a study, the orchestration engines 430 of the AI platform 400 decide the priority for each AI engine (or each AI algorithm to be performed) and assign the task of evaluating the study based on priority. In one embodiment, the orchestration engines 430 determine priority based on the characteristics of the study, its previous studies (eg, a study acquired for the same patient before the current study), and/or the nature of the algorithms. For example, with respect to pre-study, when the AI platform (eg AI platform 400) scans a study with AI algorithms, some of the AI scans require your pre-study because they will perform AI analysis (ie , an AI scan) in the above study as well. Regarding priority, for example, if the AI algorithm is to detect cerebral hemorrhage, it should be run with a high priority, since such hemorrhage must be treated immediately if it exists. Prioritization is also used when multiple images from different studies are available for evaluation at the same time by an AI engine that only handles one image analysis at a time.
[0127] Based on the orchestration decisions made by the orchestration engines 430, the AI platform 400 notifies AI servers (403A, 403B) or applications (401) that have been integrated into the AI platform 400. These AI engines they may be hosted within the AI platform 400 (eg, AI engines 403A) or may be hosted on a separate AI server (eg, AI server 403B) that is external to the platform. In different embodiments, the AI server 403B may live in a local environment or in a cloud environment. The AI platform 400 provides open APIs 403 that are used by AI applications or engines (401,403A, 403B) to access data (eg, studies, images) or store findings (eg, AI results).
[0129] In one embodiment, the notification is sent via notification logic 402B and includes information about the study and the images to be analyzed. In one embodiment, the AI applications or AI engines (401, 403A, 403B) may receive notifications from the AI platform 400, via notification logic 402B, to initiate the evaluation of AI in the studio. In one embodiment, the AI applications or engines (401, 403A, 403B) are subscribers, because they subscribe to receive notification if it meets particular characteristics. For example, in one embodiment, the motors of orchestration 430 use predefined filters and algorithm properties to determine if images for review have particular characteristics that meet the requirements of an artificial intelligence server to which they might be submitted for evaluation.
[0131] Upon notification, the AI application or engine (401, 403A, 403B) retrieves the images to be reviewed (for example, DICOM images) from the AI 400 platform via the DICOM 480 gateway APIs and analyzes one or more new images. Analysis results include AI findings or results.
[0133] In one embodiment, the AI platform has an embedded AI server 403A on which multiple enterprise-specific AI engines are integrated and deployed. In one embodiment, notifications to the IA server 403A are queued and processed based on the priority assigned by the orchestration engines 430 and other scalability considerations. In one embodiment, AI platform 400 acquires and caches DICOM instances of the data source specified in the notification message, using DICOM gateway APIs 480. AI platform 400 parses the data and prepares input required for AI engines. Input preparation includes preparation of DICOM headers and pixel data required for the engine. In one embodiment, some engines require a reformatted 3D volume to process the images, and the AI platform 400 performs the reconstruction, prepares the volumes in a suitable plane, and feeds the processed data to the AI engine. In one embodiment, the AI engines of AI server 403A use the input to evaluate DICOM images for various findings (abnormalities and diseases) and organs.
[0135] Each AI engine or application generates various forms of results that represent the corresponding findings and stores them using the corresponding open APIs. In one embodiment, the results include one or more of, but are not limited to, DICOM Grayscale Softcopy Presentation Status (GSPS), DICOM Structure Report (SR), a snapshot or a DICOM Basic Structured Display, original data (for example, DICOM or non-DICOM), a DICOM Key Object Selection Document (KOS), etc. For more information on snapshots, see U.S. Patent Application Serial No. 14/736,550, entitled "METHODS AND APPARATUS FOR TAKING A SNAPSHOT OF A MEDICAL IMAGING SCREEN," filed 11/11. June 2015. In one embodiment, the AI platform 400 stores the AI results in one or more storage devices, such as, for example, snapshot storage 410, raw data storage 420, and thumbnail storage 421.
[0137] In one embodiment, an open API provides simplified APIs and data models for generating AI result representations of DICOM complaints without having to implement DICOM conformance. For example, a heat map can be stored as a DICOM segmentation instance using an open segmentation API, without having to implement DICOM segmentation by application or AI engine (401, 403A, 403B). In addition to the individual result instances, in one embodiment, the AI applications or engine (401, 403A, 403B) create an idea registry instance using a result summary in which the details of the findings, the result instances of related AI and key images representative of the findings and/or instances of AI results will be referenced. The idea registry instance manages AI findings in an organized way. In one embodiment, idea log instances are displayed using the idea log panel user interface 441 under control of the idea log service 440. In one embodiment, idea logs are stored in log storage. of ideas 411.
[0139] Once the result of the AI analysis is ready and an idea registration instance is generated, if any, the AI platform 400 sends a notification about the arrival of the AI results to a certain predefined set of people or locations . In one embodiment, notification about the arrival of the AI results is sent to all subscribed applications. In one embodiment, AI platform 400 supports different notification mechanisms, including Weblnvoke, SignaIR, HL7, and FHIR to allow various types of systems or applications to use the AI results. Systems that consume AI result reporting include, for example, but are not limited to, Radiology Workflow Manager, Image Viewing workstation, IT Billing systems, etc. In response to the notification, the applications (401) can use the results of the AI analysis by accessing the idea registry and each of the results generated by the AI engines (403A, 403B).
[0140] In one embodiment, the AI platform 400 includes an administration user interface 450 through which a user can configure the AI platform 400 or view the configuration information thereof, under the control of the configuration service 451, with the use of information stored in configuration storage 452. In one embodiment, an administration user interface 450 allows a user to review an audit log for, or view audit information for, the AI platform 400, under the control of the audit trail service 453, with the use of information stored in the audit trail 454.
[0142] Figure 5 is a flow chart of one embodiment of a process for processing medical images. In one embodiment, the processes are performed by processing logic that may comprise hardware (circuitry, dedicated logic, etc.), software (eg, software running on a chip), firmware, or a combination of all three. In one embodiment, the process is performed by a medical image management platform, such as, for example, but not limited to, the medical image management platforms described above in conjunction with Figures 3, 4A, and 4B.
[0144] Referring to Figure 5, the process begins by processing the monitoring logic for an indication of a content change in one or more data sources (processing block 501). In one embodiment, this indication specifies the arrival of one or more new medical images at one or more data sources. In one embodiment, these new medical images are part of a new health study or an existing study created prior to the generation of the new medical images.
[0146] In response to the prompt, in one embodiment, the processing logic obtains one or more new medical images associated with the content change (processing block 502) and determines which of a plurality of AI/image analysis engines (e.g. , applications) will analyze at least one image of the new medical images based on one or both of the information accompanying the one or more images and/or the results of applying body part detection to the at least one image (processing block 503).
[0148] In another embodiment, the medical management system's processing logic does not necessarily fetch the new images for the AI engines. In one embodiment, this depends on the level of integration between the AI platform (for example, AI platform 400) and the AI engines. For example, some AI algorithms are tightly integrated in such a way that the system provides actual image pixel data or reconstructed 3D volume. This is an example of a PUSH model where the algorithm/engine gets data from the AI platform (eg AI Platform 400) and does not have to fetch images from a PACS or VNA. In another embodiment, another level of integration that is used is a PULL model where the AI engine has to pull data from PACS/VNA. In one embodiment, this is based on UID information that the AI platform (eg, AI platform 400) provides to the AI engine. To the VNA or PACS PULL data, the IA platform (eg IA platform 400) provides the DICOM open link APIs that are used. In one embodiment, at least one of the AI/image analysis engines comprises an artificial intelligence (AI) engine.
[0150] If there are multiple image analysis engines (eg, AI engines or applications) to evaluate sets of new medical images, in one embodiment, the processing logic determines a priority for each image analysis engine/AI that indicates when that image/AI analysis engine must analyze the at least one image when there are multiple image/AI analysis engines (processing block 504). In one embodiment, the processing logic determines the priority based on the characteristics in a health study of the at least one image, priorities, and the nature of the multiple AI/image analysis engines.
[0152] After determining the image analysis engines to analyze the images, the processing logic sends a notification to the AI/image analysis engine to start AI/image analysis on at least one image of the new medical images ( processing block 505). In one embodiment, this notification is sent to each AI engine/image analysis engine that has been assigned to analyze at least one image of the new medical images.
[0154] In response to receiving the notification, the AI/image analysis engine processing logic determines whether one or more features in the at least one image meet predefined criteria indicative of a medical condition (processing block 506). In one embodiment, the one or more features include one or more anatomical features and abnormalities displayed on the medical image. In a embodiment, the AI/image analysis engine may operate to perform AI/image analysis to determine whether one or more features in the at least one image meet predefined criteria without user intervention.
[0156] After performing the AI/image analysis, the AI/image analysis engine processing logic generates an output of information related to or including the AI result (for example, a DICOM-compliant representation of AI results without implementing DICOM conformance) (processing block 507) and sends one or more outputs (e.g., DICOM-compliant AI result representation as a DICOM object) to the medical image management system and/or data sources (e.g., example, PACS, VNA, etc.) for storage therein (processing block 508). In one embodiment, the one or more outputs generated by some AI engines are non-DICOM AI output or non-DICOM compliant output. In one embodiment, this occurs when there are no findings. The processing logic also receives an indication that the AI/image analysis results are available from the AI/image analysis engine suite (processing block 509) and sends a notification to subscribers to indicate the availability of the AI/image analysis engines. AI/image analysis results to access and display them (processing block 510).
[0158] Figure 6 illustrates a more detailed data flow diagram of a process for processing health studies. In one embodiment, the processes are performed by processing logic that may comprise hardware (circuitry, dedicated logic, etc.), software (eg, software running on a chip), firmware, or a combination of all three. In one embodiment, the process is performed by a medical image management platform, such as, for example, but not limited to, the medical image management platforms described above in conjunction with Figures 3, 4A, and 4B.
[0160] The process begins when the applications 401 receive a notification of the arrival of new medical images (processing block 601). In one embodiment, applications 401 are part of an AI server hosting one or more AI engines or algorithms. In another embodiment, the application is an AI engine itself. In one embodiment, the AI algorithm or AI engine is part of a PACS, VNA, or other similar medical information management system. The system may have a viewer to display the results of image analysis (for example, AI results).
[0161] In one embodiment, the new medical images are part of a new health study. In another embodiment, new medical images comprise new images added to an existing healthcare study (eg, a healthcare study that was previously created and already contains medical images (eg, a series of medical images) and has had images added to it additional medical). Images can include DICOM or non-DICOM images.
[0163] In one embodiment, applications 401 are notified of the arrival of new medical images from monitoring and notification logic 402. In one embodiment, monitoring and notification logic 402 continuously monitors content changes in one or more data sources. . The data sources comprise one or more PACS 400A (for example, DICOM-based PACS, a vendor-specific database, etc.), VNA 400B (for example, DICOM-based VNAs, a non-DICOM VNA, etc.) or other medical image repositories. The PACS can be a publicly available PACS or a private PACS that is only available to predetermined groups of individuals or institutions (for example, subscribers). In one embodiment, when a new study arrives or new medical images have been added to an existing study in PACS 400A or VNA 400B, its DICOM server sends a notification to monitoring and notification logic 402.
[0165] In one embodiment, monitoring of the study performed by monitoring and notification logic 402 is accomplished through custom integration or through standard interfaces. For example, as a subscriber, the monitoring and notification logic 402 receives a notification using standard HL7/FHIR interfaces indicating a content change in one of the data sources such as PACS 400A and VNA 400B. One drawback of using such a standard interface is that the information provided is simply an identification of the study that has been changed without an indication of what change has occurred. Therefore, in response to this notification, the study must be retrieved and analyzed to determine the new content that constituted the change before any analysis can be performed. In one embodiment, the notification received by the monitoring and notification logic 402 includes information indicating the specific change in the study. For example, this information may specify new medical images added to the study. In this case, the analysis can start earlier.
[0166] In response to receiving notification of new medical images, the processing logic obtains the new medical images (processing block 602). In one embodiment, requests 401 obtain the new medical images by obtaining the health study in which they are contained. In one embodiment, applications 401 obtain the health study through DICOM gateway 403 via open API 404.
[0168] After obtaining the new medical images, the processing logic determines which AI algorithms will be applied to the new medical images (processing block 603). In one embodiment, the information associated with each AI engine when the AI engine was added to the platform indicates its application as it relates to image analysis. That is, each AI engine is only applicable for analyzing medical images that contain a specific type of medical condition. Therefore, when new medical images are obtained, the new medical images are sent only to artificial intelligence engines designed to analyze their specific content. For example, if an AI engine is designed to analyze chest x-rays, only chest x-ray images are sent to the AI engine for analysis.
[0170] In one embodiment, the orchestration logic determines the content of the new medical images in order to identify the appropriate AI engines to be assigned to review the images. In one embodiment, the orchestration logic uses information accompanying the new medical images to determine their content. In one embodiment, the information accompanying the new medical images is included in the header information. For example, in one embodiment, the orchestration logic uses various filters based on DICOM headers to determine the content of new medical images. In another embodiment, the orchestration logic uses built-in body part detection algorithms to decide which AI engines should be assigned to evaluate the study. These algorithms analyze the images to determine which body parts are shown in the images, and based on this body part detection, the orchestration logic is able to determine the appropriate AI engines to analyze the images. In one embodiment, the orchestration logic only sends a relevant subset of images of the new images to the AI engine. That is, the orchestration logic determines which of the new images has a body part that can be evaluated by a particular AI engine, and then sends only those images to the AI engine for analysis. Note that in one embodiment, the Orchestration logic determines which AI engines to allocate new medical images for analysis using the information that accompanies the new medical images (for example, header-based information) and body part detection.
[0172] In one embodiment, if the orchestration logic determines that multiple AI engines should analyze new medical images (or portions thereof), then the orchestration logic determines a priority among the identified AI engines to control the order in which AI engines analyze certain images. In one embodiment, the priority is based on the critical nature of what the AI engine is analyzing. Therefore, if there are two or more AI engines analyzing the new medical images, the orchestration logic assigns a higher priority to the AI engine that evaluates the image for the most critical medical condition and causes that AI engine to run first. . Note that in one embodiment, if processing power is available, multiple AI engines run at the same time on the same set of images. However, the prioritization performed by the orchestration logic may be between multiple sets of new images from multiple studies that have been received at the same time. In this way, the analysis of the images for more critical medical conditions occurs before the analysis of less critical medical conditions. In this way, the techniques described herein better facilitate an early detection system in that new medical images are analyzed for more critical conditions before analysis of less critical conditions.
[0174] In one embodiment, the orchestration logic notifies the artificial intelligence engine to analyze the new medical images. In one embodiment, the orchestration logic notifies the AI engine by sending a notification to the AI engine that includes a study identifier (ID) and a list of one or more images to be analyzed. This is beneficial because the AI engine can focus its analysis on the relevant images from the set of new images instead of having to analyze all new images, including those that do not contain a theme relevant to the AI engine. In one embodiment, a list of image or frame identifiers, which could come from multiple studies, is sent even though one set is considered the current image or frame and is the target of the AI (as opposed to data from the previous study). The Study ID indicates which study is current. In general, however, it can be more granular than that (for example, at the series level, at the image level, at the frame level, etc.).
[0175] Once the AI algorithms to be applied to the new medical images have been identified and notifications have been sent to the AI engines to analyze the new medical images, the processing logic in the AI engines obtains the images needed to analyze, executes the AI analysis and generates AI results (processing block 604). In one embodiment, the AI engine obtains the images needed for analysis using DICOM gateway 403 and open API 404.
[0177] In one embodiment, the AI engines processing logic creates findings indicative of the AI results. In one embodiment, the findings comprise one or more DICOM objects that contain or are indicative of at least some aspect of the IA results. In one embodiment, DICOM objects are created without the AI engine having to be familiar with the DICOM standard. In one embodiment, the result object generation module 406 is used by the AI engines 405 to create DICOM objects or other output containing the findings. Findings may comprise a mask image that is only a relevant part of one of the new images showing the finding that will be reviewed by a physician or clinician. The finding may be a screenshot of the relevant part of an image that has been captured by the AI engine. In one embodiment, the result object generation module 406 captures the portion of the image. The finding may be a heatmap related to an image that was analyzed. In one embodiment, the find may be a snapshot of the raw data. In one embodiment, DICOM objects that can be generated include DICOM GSPS, DICOM SR, DICOM Basic Structured Display, original data (DICOM or non-DICOM), DICOM KOS.
[0179] The processing logic sends the one or more DICOM objects to PACS 400A for storage therein (processing block 605). In one embodiment, the processing logic sends the one or more DICOM objects to PACS 400A through DICOM gateway 403. In another embodiment, the processing logic sends the DICOM objects to the VNA 400B. Results can also be sent to other locations or destinations.
[0181] In one embodiment, the processing logic optionally generates and sends a request to create GSPS of an AI analysis result display state (processing block 606) and/or an idea log (processing block 607 ^).
[0183] In one embodiment, the processing logic sends one or more DICOM objects in messages to alert those persons (eg, doctors, technicians, etc.) or facilities (eg, hospitals, clinics, etc.) that have requested or are have subscribed to receive the results (processing block 608). These messages act as notifications and are sent by the notification logic. In one embodiment, notifications are sent to all subscribed applications. This is particularly important in situations where an AI engine has identified that a critical medical condition exists or is highly likely to exist. In one embodiment, the medical image management system supports different notification mechanisms including, for example, but not limited to Weblnvoke, SignaIR, HL7, and FHIR to allow various types of systems or applications to use the results of the AI. Systems that consume AI result reporting include, for example, but are not limited to, Radiology Workflow Manager, Image Viewing workstation, IT Billing systems, etc.
[0185] In one embodiment, the processing logic in the applications 401 uses the results of the AI analysis by accessing the idea log and then each of the results generated by the AI engines (processing block 609). In this case, the processing logic retrieves and displays the idea log, thus allowing a person to review the AI output. The idea log may include a snapshot, such as the snapshot mentioned above. Using the idea log to review AI findings is advantageous because AI findings are stored in the PACS in separate locations, and the idea log puts all AI findings in one place for access in an organized way. .
[0187] While the above process discusses the automatic analysis of new medical images that occurs in response to notification of the arrival of new images, in one embodiment, the individual AI engines can be started manually. In one embodiment, a set of APIs are used to start one of the AI engines. This is necessary in cases where the nature of the AI algorithms is such that preliminary analysis is necessary before the AI engine is activated. The following are some examples of this. However, it should be noted that other examples are possible.
[0188] 1. A user wants to analyze a specific part of the 3D image/volume by marking an area on the 3D image/volume using an image viewer application. In this case, the application can call the API to trigger AI exploration by specifying the user-marked image points.
[0190] 2. A historical study exists and is already stored in VNA or PACS before the AI workflow is introduced. If the user wishes to perform an AI scan on that study, the user can do so by selecting the study from the list of studies.
[0192] 3. If the user wishes to perform an AI scan through each previous patient study to determine if there is a trend regarding a particular abnormality (for example, a cure rate), etc., if the facility does not configure the system to perform this step automatically for some reason (eg billing concerns, etc.), a user can trigger a full scan for the patient by including all their previous studies manually.
[0194] An example of medical image management system
[0196] Figure 7 illustrates an exemplary embodiment of a logical representation of a medical imaging and information management system 700 that generates and displays layouts with current and past values of the parameters described above. In one embodiment, system 700 is part of a medical imaging system as detailed above.
[0198] Medical imaging and information management system 700 includes one or more processors 701 that are coupled to communication interface logic 710 via a first transmission medium 720. Communication interface logic 710 enables communications with other devices. electronic devices, specifically enabling communication with remote users such as doctors, nurses and/or medical technicians, remote databases (for example, PACS) that store health studies, health care modalities that generate and send studies, and one or more remote locations (for example, cloud-based servers) that apply image analysis algorithms (for example, AI algorithms) on images of studies and generate findings based on the results. In accordance with one embodiment of the disclosure, communication interface logic 710 may be implemented as a physical interface that includes one or more ports for wired connectors. Additionally or alternatively, communication interface logic 710 may be implemented with one or more radio units to support wireless communications with other electronic devices.
[0200] The one or more processors 701 is further coupled to persistent storage 730 via 2nd transmission medium 725. According to one embodiment of the disclosure, persistent storage 730 may include data and code to implement: (a) logical interface 741 , (b) monitoring logic 742, (c) notification/alert logic 743, (d) image analysis logic (eg, AI analysis engine) 744, (e) orchestration logic 731, (f) an import logic 732, (g) a snapshot generation/idea logging/output logic 733, (h) a display control logic 734, (i) a notes database 736 and (j) a database data records 737.
[0202] In one embodiment, interface logic 741 includes logic to enable interaction between platform components, as well as between a user and display areas displayed on the display screen. In one embodiment, interface logic 741 includes implementations for managing open APIs. User interfaces include GUI generation with studies, or portions thereof, and their associated AI results.
[0204] Monitoring logic 742 includes logic for continuous monitoring to determine when changes have occurred to data sources such as PACS systems, VNAs, or databases to determine when new images are available for analysis by image analysis engines /IA.
[0206] Orchestration logic 731 includes logic to determine which AI/image analysis engines will be assigned to analyze new medical images. This logic includes image header and study analysis logic and body part detection logic to identify the content of medical images, as well as matching logic to compare header/study analysis results and the detection of body parts with the features of available AI/image analysis algorithms to perform mapping adequate. In one embodiment, orchestration logic 731 may receive input from the user to manually initiate a scan or review of an existing medical image.
[0208] Notification/alert logic 743 includes logic for issuing and sending notifications to AI/image analysis engines and applications for evaluating one or more new medical images. Notification/alert logic 743 also generates and sends notifications and/or alerts for study reviews, including AI/image analysis results, to one or more physicians and medical staff. In one embodiment, the notification/alert logic 743 sends an alert (eg, SMS, text, email, or other message, a chat prompt indicating that a chat session with the doctor is desired, etc.) in response. to a predetermined finding in the results of automated image analysis performed on one or more images from a health study. In one embodiment, the predetermined finding comprises an anomaly score above a threshold level. In another embodiment, the predetermined find comprises one or more keywords in the finds. In yet another embodiment, the predetermined finding comprises an anomaly score above a threshold level and one or more keywords in the findings. In one embodiment, the alert is sent to one or more predetermined health care providers responsible for managing a condition associated with the predetermined finding. For example, in one embodiment, if the findings indicate that the patient is likely to have experienced a stroke, an alert is automatically sent to a stroke team at a particular medical center to care for the patient. In one embodiment, the alert includes a link to an image related to the health study findings that was generated by the automated image analysis algorithm. In such a case, the alert may include a link that the user selects to open a study that contains the image associated with the finding, and the system displays the image.
[0210] Image analysis logic 744 performs one or more image analysis algorithms on healthcare imaging images. In one embodiment, the image analysis algorithms are AI analysis algorithms. The results of the application of image analysis algorithms can be displayed on a screen.
[0212] Import logic 732 may include logic to retrieve one or more pieces of information from a storage device and import each of the one or more pieces of information in a separate display area of a viewer or viewer template. For example, pieces of information may include, but are not limited or restricted to, (i) findings from automated image analysis algorithms (eg, AI algorithms); (ii) medical images, including x-rays, mammograms, computed tomography (CT) scans, magnetic resonance imaging (MRI), positron emission tomography (PET) scan, and/or ultrasound images, (iii) notes from the physician with respect to one or more of the medical images and/or (iv) medical records corresponding to one or more of the subjects of the one or more medical images.
[0214] The snapshot/idealog/output generation logic 733 includes logic to generate a snapshot (saving the state of the design template), an idealog, and/or image analysis/AI output as described above . Saving the state may include storing at least (i) the one or more pieces of information, and (ii) the display properties of each of the one or more pieces of information on a non-transient computer-readable medium. The design template can represent one or more images from a health study showing image data that is relevant to a finding of an automated image analysis algorithm. The snapshot/idea log/output logic 733 is capable of saving the snapshot, idea log, or AI/image analysis result to a medical record or report and/or send it to a predetermined location.
[0216] Display control logic 734 includes logic for displaying user interfaces, images, and AI/image analysis results that have been rendered locally as described above. In one embodiment, display control logic 734 includes logic to display a browser in which the images, user interfaces described above, are displayed.
[0218] Notes database 736 stores notes recorded by a doctor, nurse, clinical technician, etc., which a user can import into a display area of a layout template. Lastly, record database 737 stores medical records that a user can import into a display area of a layout template.
[0219] There are a number of exemplary embodiments described herein.
[0221] Example 1 is a method comprising: monitoring, by a medical image management system, a first indication of a content change in one or more data sources; determine, in response to the first prompt, which of a plurality of image analysis engines will analyze at least one of the one or more new medical images associated with the content change based on one or both of the information accompanying the or the images or the results of applying the body part detection on the at least one image; send a first notification to start image analysis on the at least one image of one or more medical images, the first notification being sent to each image analysis engine in a set of one or more determined image analysis engines to analyze the at least one image of the one or more new medical images; receiving a second indication that image analysis results are available from the set of one or more image analysis engines; and sending a second notification to subscribers to indicate the availability of the image analysis results for access and viewing.
[0223] Example 2 is the method of Example 1 which may optionally include one or more of the plurality of image analysis engines comprising an artificial intelligence (AI) engine.
[0225] Example 3 is the method of Example 1 which may optionally include the first prompt specifying the arrival of one or more new medical images in one or more data sources.
[0227] Example 4 is the method of Example 1 which may optionally include one or more new medical images being part of a new health study or an existing study created prior to the generation of the new medical images.
[0229] Example 5 is the method of Example 1 which may optionally include determining a priority for each image analysis engine determined to analyze the at least one image when there are multiple image analysis engines.
[0231] Example 6 is the method of Example 5 which may optionally include priority determination based on characteristics in a health study of the at least one image, priors, and the nature of multiple image analysis engines.
[0233] Example 7 is the method of Example 1 which may optionally include, in response to receiving the notification, determining, using the image analysis engine, whether one or more features in the at least one image meet predefined criteria indicative of a medical condition. ; generate, by the image analysis engine, a DICOM-compliant representation of AI results without implementing DICOM conformance; and sending the DICOM-compliant AI result representation as a DICOM object to the medical image management system.
[0235] Example 8 is the method of Example 7 which may optionally include one or more features including the one or more anatomical features and abnormalities displayed on the medical image.
[0237] Example 9 is the method of Example 7 which may optionally include the image analysis engine being operable to perform image analysis to determine whether one or more features in the at least one image meet predefined criteria without user intervention.
[0239] Example 10 is a medical image management system comprising: a network communication interface to receive health studies; a memory coupled to the network communication interface to store the health studies received; a display screen coupled to the memory to display the health studies received; and one or more processors coupled to the network connection interface, the memory and the display screen and configured to monitor a first indication of a content change in the one or more data sources; determine, in response to the first prompt, which of a plurality of image analysis engines will analyze at least one of the one or more new medical images associated with the content change based on one or both of the information accompanying the or the images or the results of applying the body part detection on the at least one image; send a first notification to start image analysis on the at least one image of one or more medical images, the first notification being sent to each image analysis engine in a set of one or more image analysis engines determined to analyze the at least one image of the one or more new medical images; receiving a second indication that image analysis results are available from the set of one or more image analysis engines; and sending a second notification to subscribers to indicate the availability of the image analysis results for access and viewing.
[0241] Example 11 is the system of Example 10 which may optionally include one or more of the plurality of image analysis engines comprising an artificial intelligence (AI) engine.
[0243] Example 12 is the system of Example 10 which may optionally include the first prompt specifying the arrival of one or more new medical images in one or more data sources.
[0245] Example 13 is the system of Example 12 which can optionally include one or more new medical images being part of a new health study or an existing study created prior to the generation of the new medical images.
[0247] Example 14 is the system of Example 10 which may optionally include one or more processors operable to determine a priority for each image analysis engine determined to analyze the at least one image when there are multiple image analysis engines.
[0249] Example 15 is the system of Example 14 which may optionally include priority determination based on characteristics in a health study of the at least one image, priorities, and the nature of multiple image analysis engines.
[0251] Example 16 is a non-transient computer-readable storage medium having instructions stored therein that, when executed by a system having at least one processor, memory, and display screen thereon, cause the system perform a method comprising: monitoring a first indication of a content change in one or more data sources; determining, in response to the first prompt, which of a plurality of image analysis engines will analyze at least one of the one or more new medical images associated with the content change based on one or both of the information accompanying the image(s) or the results of applying body part detection to the at least one image; send a first notification to start image analysis on the at least one image of one or more medical images, the first notification being sent to each image analysis engine in a set of one or more determined image analysis engines to analyze the at least one image of the one or more new medical images; receiving a second indication that image analysis results are available from the set of one or more image analysis engines; and sending a second notification to subscribers to indicate the availability of the image analysis results for access and viewing.
[0253] Example 17 is the computer-readable storage medium of Example 16 which may optionally include one or more of the plurality of image analysis engines comprising an artificial intelligence (AI) engine.
[0255] Example 18 is the computer readable storage medium of Example 16 which may optionally include the first indication specifying the arrival of one or more new medical images in one or more data sources.
[0257] Example 19 is the computer readable storage medium of Example 18 which may optionally include one or more new medical images being part of a new health study or an existing study created prior to the generation of the new medical images.
[0259] Example 20 is the computer-readable storage medium of Example 16 which may optionally include that the method further comprises determining a priority for each given image analysis engine to analyze the at least one image when there are multiple image analysis engines.
[0261] Example 21 is the computer-readable storage medium of Example 20 which may optionally include priority determination based on characteristics in a health study of the at least one image, priors, and the nature of multiple image analysis engines. images.
[0262] Example 22 is the computer-readable storage medium of Example 16 which may optionally include that the method further comprises, in response to receiving the notification, determining, using the image analysis engine, whether one or more features in the at least an image meets predefined criteria indicative of a medical condition; generate, by the image analysis engine, a DICOM-compliant representation of AI results without implementing DICOM conformance; and sending the DICOM-compliant AI result representation as a DICOM object to the medical image management system.
[0264] Example 23 is the computer readable storage medium of Example 22 which may optionally include one or more features including the one or more anatomical features and abnormalities displayed in the medical image.
[0266] Example 24 is the computer-readable storage medium of Example 22 which may optionally include the image analysis engine operable to perform image analysis to determine if one or more features in the at least one image meet predefined criteria. without user intervention.
[0268] Portions of the foregoing detailed descriptions are presented in terms of algorithms and symbolic representations of operations on data bits within a computer's memory. These descriptions and algorithmic representations are the means used by those skilled in the art, in the field of data processing, to most effectively convey the essence of their work to other experts in the field. Here's an algorithm, and it's usually meant to be a self-consistent sequence of steps that lead to a desired result. The steps are those that require physical manipulations of physical quantities. Typically, but not necessarily, these quantities take the form of electrical or magnetic signals that can be stored, transferred, combined, compared, and otherwise manipulated. It has proven convenient at times, primarily for reasons of common usage, to refer to these signals as bits, values, elements, symbols, characters, terms, numbers, or the like.
[0270] It should be noted, however, that all of these and similar terms must be associated with the appropriate physical quantities and are merely convenient labels applied to these quantities. Unless specifically stated otherwise, as is evident from the following description, appreciates that throughout the description, the expressions of use that are used, such as "that processes", "that computes" or "that calculates", "that determines" or "that shows", or the like, refer to the action and processes of a computer system, or similar electronic computing devices, that manipulate and transform data represented as physical (electronic) quantities within the computer system's registers and memories into other data similarly represented as physical quantities within the computer system memories or registers or other similar information storage, transmission or display devices.
[0272] The present invention also relates to an apparatus for carrying out the operations herein. This apparatus may be specially built for the required purposes, or may comprise a general purpose computer selectively activated or reconfigured by a computer program stored in the computer. Such a computer program may be stored on a computer-readable storage medium, such as, but not limited to, any type of disk, including floppy disks, optical disks, CD-ROMs, and magnetic-optical disks, read-only memories ( ROM), random access memory (RAM), EPROM, EEPROM, magnetic or optical cards, or any type of support suitable for storing electronic instructions, and each of them coupled to a computer system bus.
[0274] The algorithms and screens presented in this document are not inherently related to any particular computer or other device. Various general purpose systems can be used with programs in accordance with the teachings herein, or it may be convenient to build more specialized apparatus to perform the required steps of the method. The structure required for a variety of these systems will appear in the description below. Furthermore, the present invention is not described with reference to any particular programming language. It will be appreciated that a variety of programming languages can be used to implement the teachings of the invention as described herein.
[0276] A machine-readable medium includes any mechanism for storing or transmitting information in a form readable by a machine (eg, a computer). For example, a machine-readable medium includes read-only memory ("ROM"); random access memory ("RAM"); disk storage media magnetic; optical storage media; flash memory devices; electrical, optical, acoustic or otherwise propagated signals (eg carrier waves, infrared signals, digital signals, etc.); etc.
[0278] While many alterations and modifications of the present invention will no doubt become apparent to one of ordinary skill in the art after reading the foregoing description, it should be understood that any particular embodiment shown and described by way of illustration is not intended in any way. that is considered limiting. Therefore, references to details of various embodiments are not intended to limit the scope of the claims, which, by themselves, list only those features considered essential to the invention.

权利要求:
Claims (24)
[1]
1. A method comprising:
monitor, by means of a medical image management system, a first indication of a content change in one or more data sources;
determine, in response to the first prompt, which of a plurality of image analysis engines will analyze at least one of the one or more new medical images associated with the content change based on one or both of the information accompanying the or the images or the results of applying the body part detection on the at least one image;
send a first notification to start image analysis on the at least one image of one or more medical images, the first notification being sent to each image analysis engine in a set of one or more determined image analysis engines to analyze the at least one image of the one or more new medical images;
receiving a second indication that image analysis results are available from the set of one or more image analysis engines; and sending a second notification to subscribers to indicate the availability of the image analysis results for access and viewing.
[2]
The method defined in claim 1, wherein one or more of the plurality of image analysis engines comprise an artificial intelligence (AI) engine.
[3]
3. The method defined in claim 1, wherein the first indication specifies the arrival of one or more new medical images at one or more data sources.
[4]
4. The method defined in claim 3, wherein the one or more new medical images are part of a new health study or an existing study created prior to the generation of the new medical images.
[5]
The method defined in claim 1, further comprising determining a priority for each determined image analysis engine to analyze the at least one image when there are multiple image analysis engines.
[6]
6. The method defined in claim 5, wherein the priority determination is based on characteristics in a health study of the at least one image, priors, and the nature of the multiple image analysis engines.
[7]
7. The method defined in claim 1, further comprising, in response to receiving the notification,
determining, using the image analysis engine, whether one or more features in the at least one image meet predefined criteria indicative of a medical condition;
generate, by the image analysis engine, a DICOM-compliant representation of AI results without implementing DICOM conformance; Y
send the DICOM-compliant AI result representation as a DICOM object to the medical image management system.
[8]
The method defined in claim 7, wherein the one or more features include one or more anatomical features and abnormalities displayed in the medical image.
[9]
The method defined in claim 7, wherein the image analysis engine is operable to perform image analysis to determine whether one or more features in the at least one image meet predefined criteria without user intervention.
[10]
10. A medical image management system comprising:
a network communication interface to receive health studies;
a memory coupled to the network communication interface to store the health studies received;
a display screen coupled to the memory to display the health studies received; Y
one or more processors coupled to the network connection interface, memory, and display screen and configured to
monitor for a first indication of a content change in one or more data sources;
determining, in response to the first prompt, which of a plurality of image analysis engines will analyze at least one of the one or more new medical images associated with the content change based on one or both of the information accompanying the image(s) or the results of applying body part detection to the at least one image;
send a first notification to start image analysis on the at least one image of one or more medical images, the first notification being sent to each image analysis engine in a set of one or more determined image analysis engines to analyze the at least one image of the one or more new medical images;
receiving a second indication that image analysis results are available from the set of one or more image analysis engines; and sending a second notification to subscribers to indicate the availability of the image analysis results for access and viewing.
[11]
The system defined in claim 10, wherein one or more of the plurality of image analysis engines comprise an artificial intelligence (AI) engine.
[12]
12. The system defined in claim 10, wherein the first indication specifies the arrival of one or more new medical images at one or more data sources.
[13]
13. The system defined in claim 12, wherein the one or more new medical images are part of a new health study or an existing study created prior to the generation of the new medical images.
[14]
14. The system defined in claim 10, wherein one or more processors are operable to determine a priority for each image analysis engine determined to analyze the at least one image when there are multiple image analysis engines.
[15]
15. The system defined in claim 14, wherein the priority determination is based on characteristics in a health study of the at least one image, priors, and the nature of the multiple image analysis engines.
[16]
16. A non-transient computer-readable storage medium having instructions stored thereon which, when executed by a system having at least one processor, one memory, and one display screen thereon, causes the system to perform a method comprising:
monitor for a first indication of a content change in one or more data sources;
determine, in response to the first prompt, which of a plurality of image analysis engines will analyze at least one of the one or more new medical images associated with the content change based on one or both of the information accompanying the or the images or the results of applying the body part detection on the at least one image;
send a first notification to start image analysis on the at least one image of one or more medical images, the first notification being sent to each image analysis engine in a set of one or more determined image analysis engines to analyze the at least one image of the one or more new medical images;
receiving a second indication that image analysis results are available from the set of one or more image analysis engines; and sending a second notification to subscribers to indicate the availability of the image analysis results for access and viewing.
[17]
17. The computer-readable storage medium defined in claim 16, wherein one or more of the plurality of image analysis engines comprise an artificial intelligence (AI) engine.
[18]
18. The computer-readable storage medium defined in claim 16, wherein the first indication specifies the arrival of one or more new medical images at one or more data sources.
[19]
19. The computer-readable storage medium defined in claim 18, wherein the one or more new medical images are part of a new health study or an existing study created prior to the generation of the new medical images.
[20]
20. The computer-readable storage medium defined in claim 16, wherein the method further comprises determining a priority for each determined image analysis engine to analyze the at least one image when there are multiple image analysis engines.
[21]
21. The computer-readable storage medium defined in claim 20, wherein the priority determination is based on characteristics in a health study of the at least one image, priors, and the nature of multiple image analysis engines .
[22]
22. The computer-readable storage medium defined in claim 16, wherein the method further comprises, in response to receiving the notification, determining, using the image analysis engine, whether one or more features in the at least one image meet predefined criteria indicative of a medical condition;
generate, by the image analysis engine, a DICOM-compliant representation of AI results without implementing DICOM conformance; Y
send the DICOM-compliant AI result representation as a DICOM object to the medical image management system.
[23]
23. The computer-readable storage medium defined in claim 22, wherein the one or more features include one or more anatomical features and abnormalities exposed in the medical image.
[24]
24. The computer-readable storage medium defined in claim 22, wherein the image analysis engine is operable to perform image analysis to determine whether one or more features in the at least one image meet predefined criteria without user intervention.

类似技术:

---

公开号 | 公开日 | 专利标题

---

US10607341B2|2020-03-31|Rules-based processing and presentation of medical images based on image plane

---

JP6081126B2|2017-02-15|Medical image processing apparatus, diagnostic imaging apparatus, computer system, medical image processing program, and medical image processing method

---

JP2021061042A|2021-04-15|Systems and methods for anonymizing health data and modifying and redacting health data across geographic regions for analysis

---

EP2169577A1|2010-03-31|Method and system for medical imaging reporting

---

US11195610B2|2021-12-07|Priority alerts based on medical information

---

US9895130B2|2018-02-20|Water equivalent diameter determination from scout images

---

US20190198158A1|2019-06-27|Automatic Creation of Imaging Story Boards from Medical Imaging Studies

---

ES2762302A1|2020-05-22|CONFIGURATION AND DISPLAY OF A USER INTERFACE WITH HEALTHCARE STUDIES |

---

US20150339457A1|2015-11-26|Method and apparatus for integrating clinical data with the review of medical images

---

US10176569B2|2019-01-08|Multiple algorithm lesion segmentation

---

US20130322710A1|2013-12-05|Systems and methods for computer aided detection using pixel intensity values

---

CN114051623A|2022-02-15|Image processing and routing using AI orchestration

---

ES2888523A2|2022-01-04|A platform for evaluating medical information and method for using the same

---

US11094062B2|2021-08-17|Auto comparison layout based on image similarity

---

US20210193296A1|2021-06-24|Enabling The Use Of Multiple Picture Archiving Communication Systems By One Or More Facilities On A Shared Domain

---

US20200118659A1|2020-04-16|Method and apparatus for displaying values of current and previous studies simultaneously

---

US20210157459A1|2021-05-27|Methods and systems for displaying associations and timelines of medical studies

---

JP6949909B2|2021-10-13|Image viewer

---

US20210158931A1|2021-05-27|Systems and Methods for Processing Medical Images For In-Progress Studies

---

US20210158930A1|2021-05-27|Systems and Methods for Processing Medical Images Using Relevancy Rules

---

US20200160996A1|2020-05-21|Multi-disciplinary decision support

---

US20180239876A1|2018-08-23|Optimal contrast injection protocol engine

---

US20180068071A1|2018-03-08|Active monitoring of clinical workflows

同族专利:

---

公开号 | 公开日

---

US10977796B2|2021-04-13|

---

EP3948885A1|2022-02-09|

---

US20200311938A1|2020-10-01|

---

CA3131877A1|2020-10-08|

---

WO2020205116A1|2020-10-08|

---

GB202114832D0|2021-12-01|

---

AU2020253890A1|2021-10-28|

---

DE112020001683T5|2022-01-13|

---

ES2888523R1|2022-01-12|

---

GB2596960A|2022-01-12|

引用文献:

---

公开号 | 申请日 | 公开日 | 申请人 | 专利标题

---

---

US8949403B1|2003-10-24|2015-02-03|Sap Se|Infrastructure for maintaining cognizance of available and unavailable software components|

---

DE102006008508A1|2006-02-23|2007-09-13|Siemens Ag|Medical visualization method for e.g. cardiological diagnosis of heart, involves storing recorded organ layer pictures with information about series of organ layer pictures and time period of recording in image data memory|

---

JP4545185B2|2007-11-21|2010-09-15|ザイオソフト株式会社|Image processing apparatus, image processing apparatus control method, and image processing apparatus control program|

---

US8813024B2|2008-09-22|2014-08-19|International Business Machines Corporation|System and a method for cross-platform porting of business application and making them contextually-aware on target platforms|

---

JP2013128725A|2011-12-22|2013-07-04|Hitachi Medical Corp|Examination information display device and system|

---

US8908947B2|2012-05-21|2014-12-09|Terarecon, Inc.|Integration of medical software and advanced image processing|

---

US20150331995A1|2014-05-14|2015-11-19|Tiecheng Zhao|Evolving contextual clinical data engine for medical data processing|

---

WO2016090326A1|2014-12-05|2016-06-09|Declara, Inc.|Intent based digital collaboration platform architecture and design|

---

US9704094B2|2015-02-19|2017-07-11|International Business Machines Corporation|Mapping of algorithms to neurosynaptic hardware|

---

US10332251B2|2015-06-12|2019-06-25|Merge Healthcare Incorporated|Methods and systems for automatically mapping biopsy locations to pathology results|

---

US10140100B2|2016-03-04|2018-11-27|Google Llc|Device common model interface|

---

US10452813B2|2016-11-17|2019-10-22|Terarecon, Inc.|Medical image identification and interpretation|

---

US10957442B2|2018-12-31|2021-03-23|GE Precision Healthcare, LLC|Facilitating artificial intelligence integration into systems using a distributed learning platform|

---

法律状态:
2022-01-04| BA2A| Patent application published|Ref document number: 2888523 Country of ref document: ES Kind code of ref document: A2 Effective date: 20220104 |

---

2022-01-12| EC2A| Search report published|Ref document number: 2888523 Country of ref document: ES Kind code of ref document: R1 Effective date: 20220104 |

优先权:

---

申请号 | 申请日 | 专利标题

---

US16/370,339|US10977796B2|2019-03-29|2019-03-29|Platform for evaluating medical information and method for using the same|

---

PCT/US2020/020516|WO2020205116A1|2019-03-29|2020-02-28|A platform for evaluating medical information and method for using the same|

---