JP7433750B2 - Video clip selector used for medical image creation and diagnosis - Google Patents

Description

The present invention relates to medical image creation and diagnosis, and more specifically to a video clip selector used for medical image creation and diagnosis.

Medical imaging refers to the process of producing visible images of internal parts of a mammal's body for the purpose of clinical analysis and therapeutic intervention. Medical imaging is the visualization of internal structures hidden by the external parts of a mammal's body to facilitate the diagnosis and treatment of diseases. Medical image production involves combining a plurality of different imaging means for acquiring images. Commonly used imaging methods include radiographic devices such as X-ray irradiation including computed tomography (CT), magnetic resonance imaging (MRI), medical ultrasound or diagnostic ultrasound, endoscopy, These include hardness measurements, tactile imaging, thermography, medical photography, and nuclear medicine imaging tests such as positron emission tomography (PET) and single photon emission tomography (SPECT). Depending on how the images are used for medical diagnostic purposes, or which tissue or organ or part of the organ is targeted, different imaging means are preferred for acquiring different groups of images.

BACKGROUND OF THE INVENTION The production of medical images of targeted areas of a mammal's body is made from a number of different views. Strictly speaking, since the group of medical images are actually two-dimensional images, the perspective view of the target area will differ depending on the angle and proximity of the imaging device. As in the case of imaging methods for producing medical images, clinical diagnostics vary depending on how the images are used for medical diagnostic purposes, or which tissue or organ or part of that organ is targeted. A specific visible image of the targeted area in the image is preferred.

Finally, the quality of clinical imaging of targeted areas of the mammalian body varies. That is, the clarity and focus of a medical image varies depending on the operator, ie, the technician, rather than the surgeon who ultimately makes the diagnosis based on the images produced. In some cases, the image taken of the target organ may not be complete because important features of the target organ are missing because the imaging sensor was not placed properly. There is. In other examples, even if the imaging sensor is properly positioned, external factors, such as structural features of the mammal's body, may reduce the clarity of key features of the targeted organ.

A conventional workflow for producing medical images begins with a technician pointing an imaging device at a patient to produce a predetermined set of images. The predetermined group of images to be produced depends on the imaging means used and may be still images or video clip images. Typically, technicians are aware of the ultimate purpose of the images they produce, which is to diagnose a specific disease or dysfunction in the target organ. Once a given image has been created, the collection of created images is stored in a central repository, typically called a "PACS" or "Picture Archival Communication System", and is stored in either digital or analog formats. A report is generated for the surgeon's review. The surgeon then retrieves the set of generated images and the report to perform an analysis of the generated image group. To perform this analysis, the surgeon needs to select the most suitable image from a set of images that have the correct appearance and the correct quality.

Efficient placement of the selected correct image on the display screen to facilitate diagnosis by simultaneously examining multiple different images of interest as well as selecting the correct image with the correct visible image and correct quality This process can be extremely time consuming as it requires the surgeon to do the following: If an image of the required quality is not in the produced image set but is needed, or if an image with the required visible image is not in the produced image set, the surgeon instructs the patient to The technician will have to come back to take the missing images. This is a huge waste of patient, health care facility, and surgeon resources.

The features of embodiments of the present invention summarized in this section overcome the shortcomings of the prior art described above and provide a novel and non-obvious method, data processing system, and computer for clip selection for clinical image production. Provide program products. The method of the present invention includes receiving a medical procedure and targeted portion of a selected mammalian organ via an interface to a medical imaging device. After this step, a medical imaging device is used to acquire a group of various video clip images of the targeted portion, and the acquired group of video clip images is stored in an image storage device. Each of the video clips of the group of acquired video clip images is further image-processed to determine the visible image and quality of the video clip, and to apply the procedure and target portion of the selected mammalian organ from a rule base. Read the corresponding rule. The rules retrieved in this step specify the visible image and quality of the video clip required for the specific measurements needed to perform the diagnostic procedure. Finally, the retrieved rules are applied as a filter to the obtained group of video clip images to create and create a subset of the group of video clip images that meets the specified required visual image and quality. A subset of the group of acquired video clip images is stored in an image storage device.

According to one feature of the present embodiment, applying the retrieved rules, if it is determined that there is no video clip in the group of video clip images that meets the specified required visible image and quality; , an alert is issued via the interface of the medical imaging device. According to another feature of the present embodiment, image processing of each video clip includes sending each video clip to a neural network, the neural network detecting images with a predetermined confidence in the sent video clips. including being trained to produce output indicative of an image. Instead, each video clip is sent to a content-based image retrieval system that compares the incoming video clip to a data store of known images having predetermined visible images. Shows the visible images in the video clips sent to you.

According to yet another feature of the present embodiment, a generalized quality level of the transmitted image is computed based on the identified confidence level created during the output of the neural network. Image processing of each video clip further includes: Finally, according to yet another feature of this embodiment, the retrieved rules identify landmark features that are required to be in the visible image. In this way, the image processing of each video clip is performed by sending each video clip to a neural network trained to produce an output indicative of the landmark features identified at a particular confidence level in the fed video clip. , when applying the retrieved rule, the absence of landmark features identified at the particular confidence level corresponds to the video clip being of low quality, and the absence of landmark features identified at the particular confidence level. The presence of landmark features includes corresponding to the video clip being of high quality.

According to another embodiment of the present invention, a medical image data processing system has settings for clip selection. The medical image data processing system includes a host computing system that includes one or more computers, each computer having memory and at least one processor. A medical imaging computer program executes within the memory of the host computing system and provides instructions to a communicatively coupled medical imaging device. The medical imaging computer program further provides an interface to a medical imaging device.

Importantly, the program includes computer program instructions that operate during execution of the program to perform a method of selecting clips for medical image production. The method includes receiving a medical procedure and a selection of a targeted portion of a mammal's body via an interface to a medical imaging device. Once received, the medical imaging device acquires a plurality of video clip images of the targeted portion and stores the video clip images in an image storage device. Each video clip in the group of stored video clip images is image processed to determine the visible image and quality of each video clip, and to read rules matching selected clinical procedures and target areas from the rule base. The retrieved rules then specify the required visual image and quality of the group of stored video clip images. Finally, the retrieved rule is applied as a filter to the stored group of video clip images to create a subset of the group of video clip images that meets the specified required visible image and quality; A subset of is stored in the image storage device.

Some of the additional features of the claimed invention will be explained in the detailed description of the invention described below, and other parts will be obvious from the detailed description of the invention, and some of them will be apparent by carrying out the claimed invention. be understood. The features of the present invention will be realized and achieved by means of a plurality of elements and combinations thereof particularly pointed out in the claims. It is to be understood that the above general description of the invention and the following detailed description are representative and for illustration only, and do not limit the invention as claimed.

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and, together with the detailed description, serve to explain the principles of the invention. The embodiments illustrated in the accompanying drawings are presently preferred. However, the present invention is not limited to the configuration and means shown in these drawings.

FIG. 1 illustrates a process for performing clip selection for clinical image creation. FIG. 2 is a schematic diagram of a data processing system that performs clip selection for medical image production. FIG. 3 is a flowchart illustrating a process for performing clip selection for medical image creation.

Embodiments of the present invention enable medical imaging and video clip selection for diagnosis. According to one embodiment of the present invention, a set of multiple contemporaneous video clips of mammalian organs is selected for processing, and each video clip in the selected set is a visible image of the mammalian organ. and a neural network trained to classify each video clip according to the filming method used to capture that video clip. The confidence ratings produced by the neural network serve as input to a function that determines the quality of a video clip. A subsequent use of the obtained video clip is identified. The intended use of the video clip is, for example, to calculate predetermined dimensions of an organ in order to proceed with the calculation of dimensions relevant to a predetermined diagnostic procedure. Additionally, rules are identified that are read from the rule base and indicate the visible image, the means of capture, and the required quality characteristics for the intended use of the video clip. Optionally, the retrieved rules indicate how the video clips should be presented in the viewer. Based on the visible image, photographing means, and quality required characteristics suitable for the intended use of the video clip, which are indicated by the retrieved rules, the obtained video clip is filtered to obtain the specific visible image, photographing means, and quality required characteristics. Create a subset of video clips that have Ultimately, this created subset of video clips is sent as input to a clinical viewer that provides the subset of video clips for viewing by a health care professional. Optionally, the viewer determines how the created subset of video clips is presented according to the retrieved rules. In particular, the manner in which the subset of video clips is provided includes the provision of the subset of video clips such that the healthcare professional is first presented with the most appropriate video clip of the subset. Including reordering video clips.

In a more detailed description of the invention, FIG. 1 illustrates the process of clip selection for medical image production. As shown in FIG. 1, a medical imaging device 100 captures a group of images of a target organ of a mammal into an image collection 130. Each of the images in image set 130 depicts a certain visible image of the target organ and has a particular quality. Clip selection circuit 150 loads each image in image set 130 and determines both the visible image that image projects and the determined quality 160. Thereafter, the procedure 110 to be performed is determined and sent to the medical imaging device 100, and rules regarding the required quality and visible image of the group of images are selected from the rule table 120 based on the procedure 110 to be performed. Additionally, optionally, predetermined dimensions are determined that are to be calculated in order to proceed with the procedure 110 to be performed. The selected rule 140 is sent to clip selection logic 150.

Clip selection logic 150 has a filter function that applies a selected rule 140 to each image in image set 130 , the filter function is such that rule 140 filters image set 130 or sorts image set 130 . (reordering) to create an image subset 190 that has sufficient visible image and quality to support a given procedure 110. Optionally, this image subset 190 consists only of images in order with the most desirable viewing and quality images at the top. Thereafter, if clip selection logic 150 determines that one or more images are present in image subset 190, image subset 190 is stored in data storage 175 for use in the medical diagnosis comprising identified procedure 110. It will be done. Alternatively, if clip selection logic 150 detects that image subset 190 remains empty set 185, clip selection logic 150 detects the visible images and necessary information needed to proceed with identified procedure 110. A prompt 195 is provided to the medical imaging device 100 indicating that a new set of images meeting one or both of the above qualities should be acquired again.

The processes described in connection with FIG. 1 may be implemented in a data processing system. In the following description, FIG. 2 schematically shows a data processing system with settings for clip selection for medical image production. The data processing system includes a host computing system 200 that includes memory 220, at least one processor 210, and a display 230. The host computing system 200 is also coupled to a medical imaging device 250 for acquiring medical images of the target organ, and an image storage device 240 for storing the acquired images. An operating system 260 executes in memory of host computing system 200. Operating system 260 supports execution of the clip selection module 300 program code.

When the clip selection module 300 program code is executed in memory by the processor 210, the clip selection module 300 program code operates to receive procedure instructions and directed procedures related to the target organ, and perform clinical imaging. A collection of images acquired by the medical imaging device 250 is received and placed into the image storage 240 in accordance with a directed procedure. While the program code is being executed, its operation analyzes each image in the collection of images and provides each image with both its appearance and quality. Regarding this process, the program code of clip selection module 300 may send each image in the collection of images to neural network 280. This neural network 280 is trained to generate a probabilistic measure of visible image and quality corresponding to the transmitted image. Alternatively, the program code of clip selection module 300 may send each image to content-based image retrieval system 270. The content-based image retrieval system 270 is configured to classify each image of a given visible image and a predetermined quality based on a group of images having a known visible image and a known quality. The group can be compared to a set of groups of known images.

Optionally, the content-based image retrieval system 270 can indicate quality based on the appearance in the image of mammalian body landmark features that are expected to be displayed in a given view of the image. Invisibility of such landmark features indicates poor quality. Varying degrees of visibility of landmark features indicate varying degrees of quality. For example, having landmark features clearly visible in an image when you expect them to be visible indicates high quality. Conversely, partially visible landmark features indicate mediocre quality.

Another option is to calculate for each image a feature (Echo) distance related to the difference between the orientation of the image acquisition device when the image was acquired and the ideal orientation. More specifically, a set of training images in which each training image is annotated with a known pose, and the known pose is used to select the A set of training images to obtain training images corresponding to that known pose, and optionally a deviation from the ideal pose that is known in advance to obtain the best quality form of training images. Since there is a correlation, by comparing the images obtained later with the training images, it is possible to estimate the difference in posture, which is called a feature distance. The above is done by a neural network trained with training images to show content-based image retrieval or feature distance. The quality of the subsequently obtained image is based on the feature distance that correlates with the quality of the image, such that as the feature distance of the subsequently obtained image increases and approaches a predetermined threshold feature distance, the quality of the image decreases. It can be decided.

As another option, neural network 280 can indicate that the visible images in the sent video clips have been identified with a predetermined confidence level.

Once the visible image and quality of each image in the image collection has been calculated by the program code of the clip selection module 300, the program code further operates to select a predetermined rule from the base of rules that matches the directed procedure. , apply the selected rule to each image in the set of images. In this step, the determined visible images and qualities of each image in the image set are selected to determine whether these visible images and qualities exceed those required by the selected rule. is sent as input to a predetermined rule. If the selected rule exceeds what is required, the submitted image is added to the subset of images in image storage 240. Otherwise, the sent image will be discarded. Once each image in the image collection is processed according to the predetermined rules described above, the program code of clip selection module 300 determines whether the image is present in the subset of images in image storage 240. If not, the clip selection module 300 generates and inputs an alert to the interface indicating that it is necessary to instruct the medical imaging device 250 to acquire additional groups of images.

FIG. 3, which further explains the operation of the clip selection module 300, is a flowchart illustrating the process of selecting clips for creating medical images. The process begins at block 310 with identifying and inputting a targeted procedure and targeted organ into an interface to a medical imaging device. Block 320 finds rules that apply to the identified targeted procedure and targeted organ. Thereafter, block 330 reads the first image in the image collection from the data storage device and loads the quality and visible image imparted to the first image into memory. Optionally, retrieving this first image from the data storage device occurs in real time while the medical imaging device is acquiring the image set. Block 340 applies the found rules to the given quality and visible image to make the decision of block 350. At decision block 350, if the first image has sufficient quality and visibility for the identified target procedure and target organ, then at block 360 the first image is added to the subset. Otherwise, the process continues to decision block 370.

Decision block 370 indicates that if additional images are to be further processed for processing, block 380 selects the next image in the collection of images and performs the process of block 340 of applying the found rules. repeat. Otherwise, block 390 determines whether there are any images in the subset. If there are images in the subset, the subset is returned at block 400 for use in the analysis of the targeted procedure. However, if this is not the case, then at block 410 a prompt is presented at the interface to the medical imaging device indicating that additional groups of images should be acquired having the required quality, the required visible image, or both. .

The present invention may be implemented as a system, method, computer program product, or any combination thereof. A computer program product includes a computer readable storage medium or media that stores computer readable program instructions that cause a processor to implement the features of the present invention. A computer-readable storage medium is a tangible device that retains and stores instructions for use by a device to execute the instructions. The computer readable storage medium may be, for example and without limitation, an electronic storage device, a magnetic storage medium, an optical storage medium, an electromagnetic storage medium, a semiconductor storage medium, or any combination of the above media.

The computer readable program instructions herein can be downloaded from a computer readable storage medium to a separate computing/processing device or via a network to an external computer or external storage device. The computer-readable program instructions may be executed as a stand-alone software package, entirely on a user's computer, partially on a user's computer and partially on a remote computer, or entirely on a remote computer or computer. Runs on the server. Features of the present invention are described herein with reference to flowchart illustrations and/or block schematic diagrams of methods, apparatus (systems), and computer program products in accordance with embodiments of the present invention. Each block in the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, is implemented by computer-readable instructions.

Such computer-readable instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing device such that instructions for execution on the processor of the computer or other programmable data processing device may be , creating means for implementing the functions/acts identified in the blocks of the flowcharts and/or blocks of the block diagrams. Such computer-readable instructions may also be stored on a computer-readable storage medium, where the computer-readable storage medium may be a computer, programmable data processing apparatus, and/or other operative device in a given form. command. Accordingly, a computer-readable storage medium having instructions stored thereon includes an article of manufacture that includes instructions implementing the functions/acts identified in the blocks of the flowcharts and/or blocks of the block diagrams.

The computer readable instructions described above are loaded into a computer, other programmable data processing apparatus or other device and executed on the computer, other programmable data processing apparatus or other device, so that the computer readable instructions are executed by the computer. Create a process to As a result, instructions executed on a computer, other programmable data processing apparatus, or other device implement the functions/acts identified in the blocks of the flowchart diagrams and/or blocks of the block diagrams.

The flowcharts and block schematic diagrams in the accompanying drawings describe the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products in accordance with various embodiments of the present invention. In this regard, each block in the flowcharts or block diagrams represents a module, segment, or portion that includes one or more executable instructions for implementing particular logical functions. In alternative implementations, the functionality depicted in the blocks may result from the instructions depicted in the accompanying figures. For example, two blocks shown in succession may in fact be executed substantially concurrently, or the blocks may be executed in the reverse order depending on the functionality required. Each block in the block schematic diagram and/or flowchart illustration, and combinations of blocks in the block schematic diagram and/or flowchart illustration, may be implemented by a system based on special purpose hardware, such as a It should be understood that a combination of dedicated purpose hardware and computer instructions implements the functions or acts of a computer.

Finally, the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used herein, the singular forms "a,""an," and "the" include plural references unless the context clearly dictates otherwise. When the terms "includes" and/or "including" are used herein, it does not imply that the described features, integers, steps, processes, elements and/or parts are present. This does not exclude the presence or addition of one or more other features, integers, steps, steps, elements, components and/or combinations thereof.

The same structure, materials and operation as, or equivalents of, the elements performing the function of all means or steps recited in the claims below may be claimed as: It is intended to include structures, materials, or acts for performing that function in combination with other elements within the scope of the invention. The detailed description of the invention is presented for purposes of illustration and description and is not intended to be exhaustive or to limit the invention to the form disclosed. It will be apparent to those skilled in the art that there are many variations and modifications that do not depart from the scope and spirit of the invention. In order to best explain the principles of the invention, and to enable those skilled in the art to understand the invention as having various embodiments with various modifications suitable for the particular uses contemplated, this specification has been detailed. The embodiments of are those selected and described.

Since the invention of this specification has been described in detail and with reference to the embodiments as described above, there are variations that do not depart from the technical scope of the invention of this application as defined in the following claims. Obviously, modifications and variations are possible.

Claims (18)

A method for selecting clips for creating medical images, the method comprising:
through the interface to the medical imaging device, the medical imaging device receives a medical procedure and a targeted portion of the selected mammalian organ;
the medical imaging device acquires a group of various video clip images of the targeted portion;
a host computing system coupled to an image storage device stores the group of video clip images in the image storage device;
the host computing system image- processes each video clip of the group of video clip images to determine a visible image and quality of each video clip;
the host computing system reads from a rule base a rule corresponding to the selected clinical procedure and targeted portion that specifies a required visual image and required quality of the group of video clip images;
the host computing system applies the retrieved rule as a filter to the group of video clip images to create a subset of the group of video clip images that meets the identified required visible image and required quality;
the host computing system storing a subset of the group of video clip images on the image storage device ;
If it is determined that there is no video clip in the group of video clip images that satisfies the identified required visible image and required quality by applying the retrieved rules, controlling the interface to the medical imaging device; How to raise an alert via . The image processing of each video clip includes sending each video clip to a neural network trained to create an output indicative of visible images discerned with a predetermined confidence in the sent video clip. The method described in Section 1. The image processing of each video clip involves comparing the incoming video clip with a data store of known images having specific visible images to determine the identified visible images present in the incoming video clip. 2. The method of claim 1, comprising sending each video clip to a content-based image retrieval system showing the content. The image processing of each video clip comprises computing a generalized quality level of the transmitted image based on the identified confidence level created in the output of the neural network. 3. The method of claim 2 , comprising: The loaded rules further specify landmark features that need to be present in the visible image, and the image processing of each video clip determines the landmark features that need to be present in the visible image, and the image processing of each video clip identifies landmark features that need to be present in the visible image. When applying the retrieved rules by feeding each video clip to a neural network trained to produce an output indicative of the marked features, the absence of landmark features identified at the particular confidence level indicates that the 2. The method of claim 1, wherein the video clip is of low quality and the presence of landmark features identified at the particular confidence level corresponds to the video clip being of high quality. While the clinical imaging device is used to acquire a group of various video clip images of the target area in real time, the group of acquired video clip images is stored in an image storage device. The method described in Section 1. a host computing system having one or more computers, each computer having memory and at least one processor;
A medical imaging computer program executed in the memory of the host computing system and providing control instructions to a communicatively coupled medical imaging device and providing an interface to the medical imaging device, the medical imaging computer program comprising: Working while the image preparation computer program is running,
receiving a medical procedure and targeted portion of a selected mammalian organ via an interface to the medical imaging device;
acquiring a group of various video clip images of the targeted area using the medical imaging device;
storing the group of video clip images in an image storage device;
image processing each video clip of the group of video clip images to determine the visual image and quality of each video clip;
reading from the rule base a rule corresponding to the selected clinical procedure and target portion, the rule specifying the required visible image and required quality of the group of video clip images;
applying the retrieved rule as a filter to the group of video clip images to create a subset of the group of video clip images that satisfies the identified required visible image and required quality;
storing a subset of the group of video clip images in the image storage device of the host computing system;
the computer program is further executed applying the retrieved rules and determining that there is no video clip in the group of video clip images that meets the identified required visual image and quality; a medical imaging computer program including computer program instructions for generating an alert via the interface of the medical imaging device; and a medical imaging data processing system suitable for selecting clips to include. Image processing of each video clip includes sending each video clip to a neural network, which generates an output indicative of the visible image in the sent video clip as determined with a predetermined confidence level. 8. The system of claim 7 , wherein the system is trained to: The image processing of each video clip includes sending each video clip to a content-based image retrieval system, the content-based image retrieval system converting the sent video clips into known images of a predetermined visible image. 8. The system of claim 7 , wherein the system compares the data storage device with a distinct visible image in the transmitted video clip. 5. The image processing of each video clip further comprises computing a generalized quality level of the transmitted image based on a predetermined confidence level created in the output of the neural network. 8. The system described in 8 . Identifying landmark features that are required to be in the visible image according to the retrieved rules, and the image processing of each video clip includes:
sending each video clip to a neural network trained to generate an output indicative of landmark features identified at a particular confidence level in the sent video clip;
When applying the retrieved rule, the absence of landmark features identified at the particular confidence level corresponds to the video clip being of low quality, and the absence of landmark features identified at the particular confidence level. 8. The system of claim 7 , wherein visibility of landmark features corresponds to the video clip being of high quality. A program product for selecting clips for medical image production, the program product comprising a computer readable storage medium having program instructions, the program instructions being executable by a predetermined device, and wherein the program instructions The device is
receiving a medical procedure and targeted portion of the selected mammalian body via an interface to a medical imaging device;
acquiring a group of various video clip images of the targeted area using the medical imaging device;
storing the group of video clip images in an image storage device;
image processing each video clip of the group of video clip images to determine the visual image and quality of each video clip;
retrieving from the rule base a rule corresponding to the selected clinical procedure and targeted portion that specifies the required visible image and required quality of the group of video clip images;
applying the retrieved rule as a filter to the group of video clip images to create a subset of the group of video clip images that satisfies the identified required visible image and required quality;
storing a subset of the group of video clip images in the image storage device;
The method performed by the device applies the retrieved rules to determine that there is no video clip in the group of video clip images that meets the identified required visual image and quality. a program product that generates an alert via the interface of the medical imaging device if The image processing of each video clip includes sending each video clip to a neural network that generates an output indicative of the visible image in the sent video clip as distinguished with a predetermined confidence level. 13. The program product of claim 12 . The image processing of each video clip involves sending each video clip to a content-based image retrieval system which stores the sent video clip in a data storage device comprising known images of a given visible image. 13. The program product of claim 12 , further comprising displaying a distinct visible image in the transmitted video clip. Image processing of each video clip as described above further includes computing a generalized quality level of the transmitted image based on a predetermined confidence level created in the output of the neural network. 14. The program product of claim 13 . further identifying landmark features that are required to be in the visible image according to the read rule;
Image processing of each video clip includes sending each video clip to a neural network that generates an output indicative of landmark features identified at a particular confidence level in the sent video clip. , when applying the retrieved rule, the absence of landmark features identified at the particular confidence level corresponds to the video clip being of low quality, and the absence of landmark features identified at the particular confidence level. 13. The program product of claim 12 , wherein the presence of landmark features corresponds to the video clip being of high quality. 13. The program product of claim 12 , wherein the storing of the group of video clip images is performed while acquiring the group of various video clip images of the target area using the medical imaging device in real time. A method for selecting clips for creating medical images, the method comprising:
through the interface to the medical imaging device, the medical imaging device receives a medical procedure and a targeted portion of the selected mammalian organ;
the medical imaging device acquires a group of various video clip images of the targeted portion;
a host computing system coupled to an image storage device, storing the group of video clip images in the image storage device;
the host computing system image- processes each video clip of the group of video clip images to determine a visible image and quality of each video clip;
The host computing system selects video clips in a rule and/or interface that corresponds to a selected clinical procedure and target portion and specifies the required visual image and required quality of the group of video clip images. Reads rules indicating how to arrange images to be presented from the rule base,
the host computing system applies the retrieved rules as a filter to the group of video clip images to meet the identified required visual image and required quality and/or created according to rules indicating the ordering; creating a subset of the group of video clip images;
the host computing system stores a subset of the group of video clip images on the image storage device;
Here, image processing of each video clip includes sending each video clip to a neural network trained to create an output indicative of visible images discerned with a predetermined confidence in the sent video clip. method including.

Images