High resolution synthetic medical imaging

Abstract

Example medical systems and techniques are disclosed. An example medical system includes a memory configured to store a plurality of images, each image of the plurality of images having a respective resolution. The plurality of images are generated during at least two separate medical procedures, the at least two separate medical procedures occurring during at least one of different times or different dates. The medical system includes processing circuitry communicatively coupled to the memory. The processing circuitry is configured to obtain the plurality of images. The processing circuitry is configured to generate a composite image based on averaging the plurality of images, the composite image having a higher resolution than at least a majority of the plurality of images. The processing circuit is configured to output the composite image.

Description

[0001] This application claims the benefit of U.S. Provisional Application No. 63 / 508,982, filed June 19, 2023, the entirety of which is incorporated by reference herein. TECHNICAL FIELD

[0002] The present disclosure relates to imaging, such as imaging used during medical procedures. BACKGROUND

[0003] During medical procedures, clinicians can use imaging systems to be able to visualize internal anatomy of a patient. Such imaging systems can display anatomy, medical instruments, etc., and can be used to diagnose a patient condition or help guide a clinician to move a medical instrument to an intended location within the patient. Imaging systems can use sensors to capture image data that can be displayed during a medical procedure. Imaging systems include computed tomography (CT) scanning systems, fluoroscopy systems (e.g., isocentric C-arm fluoroscopy systems), intravascular ultrasound (IVUS) systems, other ultrasound imaging systems, optical coherence tomography (OCT) fractional flow reserve (FFR) systems, magnetic resonance imaging (MRI) systems, positron emission tomography (PET) systems, and other imaging systems. Clinicians can determine treatment strategies based on images captured by such imaging systems. SUMMARY

[0004] Imagers have limited resolution. For example, CT imagers are typically limited to a resolution of 0.3 mm to 0.4 mm. In practice, fluoroscopy imagers are also typically limited to similar resolutions in order to minimize x-ray exposure to the patient and the patient’s attending clinician. Such 0.3 mm to 0.4 mm resolution has limited accuracy that can not be optimal for applications such as CT fractional flow reserve (FFR), FFR angiography, or for guidance regarding clinician decisions such as appropriate rotational atherectomy, burr size, balloon, stent, etc.

[0005] It can be beneficial to take multiple images, such as CT images, over time (weeks, months, years) as this can better enable a clinician to monitor disease progression and allow for a more comprehensive assessment of the patient and development of appropriate treatment strategies and / or changes to treatment strategies. However, CT imaging requires administration of contrast to the patient and exposes the patient (and the attending clinician) to radiation. Additionally, patient radiation exposure can be considered cumulative over a lifetime, so requiring only a minimum period of time between CT imaging sessions can not resolve the radiation exposure issue.

[0006] Generally, the present disclosure relates to various techniques and medical systems for improving resolution of image data from a medical imager while minimizing additional radiation exposure. An example technique includes generating a synthetic or composite higher resolution image from an average of a plurality of lower resolution images. The plurality of lower resolution images can be taken over time within a plurality of separate medical procedures. From a patient’s perspective, separate medical procedures should be understood to be medical procedures scheduled to occur at different times and / or on different days. From a patient’s perspective, a medical procedure scheduled to begin at one time on one day at one facility should not be considered a separate medical procedure even if such medical procedure can include different steps, different clinicians, different equipment, different locations within the facility, etc.

[0007] In one example, the present disclosure describes a medical system comprising: a memory configured to store a plurality of images of a patient, each image of the plurality of images having a respective resolution, wherein the plurality of images are generated during at least two separate medical procedures, the at least two separate medical procedures occurring during at least one of different times or different days; and processing circuitry communicatively coupled to the memory, the processing circuitry configured to: obtain the plurality of images; generate a synthetic image based on averaging the plurality of images, the synthetic image having a higher resolution than at least a majority of the plurality of images; and output the synthetic image.

[0008] In another example, the present disclosure describes a method comprising: obtaining, by processing circuitry of a medical system, a plurality of images, each image of the plurality of images having a respective resolution, wherein the plurality of images are generated during at least two separate medical procedures, the at least two separate medical procedures occurring during at least one of different times or different days; generating, by the processing circuitry, a synthetic image based on averaging the plurality of images, the synthetic image having a higher resolution than at least a majority of the plurality of images; and outputting, by the processing circuitry, the synthetic image.

[0009] In yet another example, the present disclosure describes a non-transitory computer-readable medium comprising instructions that, when executed, cause processing circuitry to: obtain a plurality of images, each image of the plurality of images having a respective resolution, wherein the plurality of images are generated during at least two separate medical procedures, the at least two separate medical procedures occurring during at least one of different times or different days. Generate a synthetic image based on averaging the plurality of images, the synthetic image having a higher resolution than at least a majority of the plurality of images; and output the synthetic image.

[0010] These and other aspects of the disclosure will become apparent from the following detailed description, taken in conjunction with the accompanying drawings. However, the disclosure is not to be limited to the specific embodiments described herein, but is to be given broadest scope consistent with the principles and novel features disclosed herein.

[0011] This summary is intended to provide an overview of the subject matter described in this disclosure. It is not intended to provide an exclusive or exhaustive explanation of the apparatus and methods described in detail within the accompanying drawings and description below. Further details of one or more examples are set forth in the accompanying drawings and description below. BRIEF DESCRIPTION OF DRAWINGS

[0012] Figure 1 is a schematic perspective view of one example of a system for generating a synthetic high resolution image in accordance with one or more aspects of the present disclosure.

[0013] Figure 2 is a schematic view of one example of a computing system of the system of Figure 1

[0014] Figure 3 is a conceptual diagram illustrating an example averaging processing technique in accordance with one or more aspects of the present disclosure.

[0015] Figure 4 is a conceptual diagram illustrating an example alignment and averaging processing technique in accordance with one or more aspects of the present disclosure.

[0016] Figure 5 is a conceptual diagram illustrating example contents of a Digital Imaging and Communications in Medicine (DICOM) file of a synthetic image in accordance with one or more aspects of the present disclosure.

[0017] Figure 6 is a flowchart of an example technique for generating a synthetic image in accordance with one or more aspects of the present disclosure.

[0018] Figure 7 is a conceptual diagram illustrating an example machine learning model in accordance with one or more aspects of the present disclosure.

[0019] Figure 8 is a conceptual diagram illustrating an example training process of a machine learning model in accordance with one or more aspects of the present disclosure. DETAILED DESCRIPTION

[0020] ​As discussed above, some imagers, such as CT images, are typically limited to a lower resolution (0.3 mm to 0.4 mm) that is lower than a resolution that can be desired, particularly when some medical instrument dimensions or differences between some medical instrument dimensions are smaller than the imager resolution. Other imagers, such as fluoroscopy imagers, are typically limited to such lower resolutions in practice due to concerns about x-ray exposure to patients and patients’ attending clinicians. Thus, it can be desirable to improve the resolution of images from lower resolution images by generating a synthetic higher resolution image from multiple lower resolution images. By generating such a synthetic higher resolution image, clinicians can better be able to visualize patients’ anatomical structures, medical instruments, implanted medical devices, etc., which can improve clinicians’ ability to correctly diagnose patients’ medical conditions, determine appropriate medical treatments, and / or guide clinicians during such medical treatments.

[0021] The present disclosure relates to imagers that can capture image data or images of a patient’s anatomy. There are many types of imagers to which the techniques of the present disclosure can be applicable. Fluoroscopy imagers can be imagers that capture real-time video data of movement inside a portion of a body by passing x-rays through the body over a period of time. CT imagers can be imagers that combine a series of x-ray images taken from different angles around a patient’s body and use computer processing to create cross-sectional images (slices) of bone, blood vessels, and soft tissue inside the body. Angiography imagers can use x-ray imaging to visualize blood vessels and often involve the injection of contrast material to make the blood vessels visible, which can be somewhat toxic to the human body. FFRangio procedures can involve non-invasively obtaining fractional flow reserve (FFR) measurements from angiography imagers. FFR can be a measurement that determines the ratio between the maximum achievable blood flow in a diseased coronary artery and the theoretical maximum flow in a normal coronary artery. FFR can allow clinicians to quantify the severity of a coronary stenosis. The techniques of the present disclosure can be applicable to these and other imagers and imaging techniques.

[0022] Many imaging techniques require the administration of a contrast agent to the patient, and many imaging techniques expose the patient (and the attending clinician) to radiation. Patient radiation exposure can be considered cumulative over a lifetime, so simply requiring a minimum period of time between CT imaging sessions can not solve the radiation exposure problem. Over time (weeks, months, years), taking multiple images (such as CT images) can be beneficial as it can better enable the clinician to monitor disease progression and allow for a more comprehensive assessment of the patient and development of appropriate treatment strategies and changes to treatment strategies. However, due to radiation exposure, it can be desirable to reduce the radiation exposure to the patient in each medical procedure during which imaging is performed. Radiation exposure can be reduced by shortening the length of time of x-ray exposure to the patient, which can result in either or both of the captured imaging data covering less anatomy or capturing lower resolution images. For the purposes of this disclosure, a medical procedure can include any procedure during which medical imaging occurs.

[0023] More detailed information can be obtained in higher resolution images than in lower resolution images. However, similar information contained in higher resolution images can be reconstructed from multiple lower resolution images, each of which exposes the patient to significantly less radiation than a high resolution scan. The techniques of this disclosure can include taking multiple lower resolution scans and using data analysis to generate a synthetic higher resolution image. In some examples, the multiple lower resolution scans can be taken over time (e.g., during multiple separate medical procedures), which can not only support the generation of a synthetic higher resolution image, but also can facilitate the clinician identifying and / or tracking disease progression over time.

[0024] The techniques of this disclosure can provide improved CT image resolution, fluoroscopy image resolution, and / or other image resolution. The techniques of this disclosure can provide more accurate CT-FFR, FFR angiography output, and / or plaque analysis than conventional techniques. For example, the techniques of this disclosure can help a clinician control a balloon or stent expansion with high accuracy. Additionally, the techniques of this disclosure can help a clinician select an accurate device size for use in a medical procedure, where the step size between device sizes is smaller than conventional CT image resolution. For example, the difference between some balloon, stent, and / or burr sizes can be approximately 0.25 mm, which is smaller than conventional CT image resolution. For example, the techniques of this disclosure can improve the effective resolution of a CT image from 0.3 mm to 0.15 mm, which can be smaller than burr size. Additionally, by using the techniques of this disclosure, a patient can be exposed to less radiation and contrast agent while the clinician can still obtain an image history of the patient's heart disease progression.

[0025] The techniques of the present disclosure take advantage of the possibility that there will be some relative movement between different image frames taken during one medical procedure (e.g., due to heartbeats, breathing, movement of the patient, etc.) and between different image frames taken over the course of multiple separate medical procedures. By averaging over more images, the techniques of the present disclosure can even further improve CT image resolution.

[0026] Figure 1 is a perspective view of one example of a system for generating a synthetic high resolution image in accordance with one or more aspects of the present disclosure. The system 100 includes a display device 110, a table 120, an imager 140, and a computing device 150. The system 100 can be an example of a system for use in an emergency room or a catheterization lab. In some examples, the system 100 can include other devices, which are not shown for simplicity. In some examples, the system 100 can also include a server 160, which can be located in the same location as the other devices of the system 100, or which can be located elsewhere. The system 100 can be used during a medical procedure, such as a diagnostic medical procedure and / or an interventional medical procedure. During such a medical procedure, and over the course of multiple separate medical procedures, the system 100 can obtain multiple images of a patient’s anatomy via the imager 140. At least some of the images of the patient’s anatomy can be at a relatively low resolution. The system 100 can generate a synthetic image based on averaging over the multiple low resolution images. The synthetic image can have a higher resolution than the multiple low resolution images.

[0027] The system 100 can include one or more machine learning models. The machine learning models can be trained to determine fiducial markers, identify fiducial markers, and / or align low resolution images so that the system 100 can average over the multiple low resolution images to generate a synthetic higher resolution image. For example, if the images are averaged over without being properly aligned, the resulting image can be incorrect. For example, the machine learning models can be used to determine what to use as a fiducial marker in the multiple images, identify the fiducial marker in at least some of the multiple images, and / or align at least some of the multiple images based on the fiducial marker. After the images are properly aligned, the system 100 can generate a synthetic image. In some examples, the machine learning models can be trained to quantify any anatomical diameter changes between the systolic and diastolic phases in the images 214 and / or in the synthetic image 216 and / or in multiple synthetic images, such as the synthetic image 216. The quantification of the diameter changes between the systolic and diastolic phases of the images can be useful to a clinician in monitoring disease progression.

[0028] The computing device 150 can include, for example, an off-the-shelf device such as a laptop computer, a desktop computer, a tablet computer, a smartphone, or other similar device, or can include a special-purpose device. The computing device 150 can perform various control functions with respect to the imager 140. In some examples, the computing device 150 can include a lead workstation. The computing device 150 can control operation of the imager 140 and receive output of the imager 140. The computing device 150 can execute machine learning algorithms and generate composite images.

[0029] The display device 110 can be configured to output instructions, images, and messages related to a medical procedure. For example, the display device 110 can display any of a plurality of images and / or a composite image obtained by the imager 140. The platform 120 can be, for example, an operating table or other platform suitable for use in a medical procedure.

[0030] In Figure 1 In examples, the imager 140, such as a CT imager, a fluoroscopy imager, an angiography imager, or other imaging device, can be used to image relevant portions of a patient’s anatomy during a medical procedure to visualize anatomy, characteristics, and locations of a lesion or other issue within a patient by generating imaging data. While primarily described herein as a CT imager or a fluoroscopy imager, the imager 140 can be any type of imaging device, such as a fluoroscopy device, a CT device, an angiography device, an intravascular ultrasound (IVUS) device, an OCT-FFR device, an MRI device, a PET device, an ultrasound device, etc. In some examples, the imager 140 can represent more than one imaging device, such as a plurality of any of the foregoing devices.

[0031] The imager 140 can image a region of interest in a patient’s body. The particular region of interest can depend on the anatomy, the medical procedure, the patient’s symptoms, etc. For example, when performing a cardiac medical procedure, a portion of the vasculature and / or heart can be within the region of interest.

[0032] The computing device 150 can be communicatively coupled to the imager 140, the display device 110, and / or the server 160, for example, by wired, optical, or wireless communication. The server 160 can or can not be located at a hospital server of an emergency room or catheter lab of a hospital, a cloud-based server, etc. The server 160 can be configured to store patient imaging data, electronic health records or medical histories, etc. In some examples, the server 160 can be configured to generate composite images.

[0033] Any or any combination of computing device 150, imager 140, and / or server 160 may include one or more machine learning models. For example, computing device 150, imager 140, and / or server 160 may acquire multiple images. Computing device 150, imager 140, and / or server 160 may execute one or more machine learning models to determine reference markers in the images, identify reference markers in other images, and / or align the images based on the reference markers. Once the images are aligned, computing device 150, imager 140, and / or server 160 may generate a synthetic image, for example, by averaging information within the aligned images. This synthetic image may have a higher resolution than some or all of the images used to generate the synthetic image. In some examples, computing device 150, imager 140, and / or server 160 may execute one or more machine learning models to quantify any diameter changes between contraction and relaxation phases in image 214 and / or synthetic image 216 and / or multiple synthetic images (such as synthetic image 216).

[0034] By generating synthetic images with higher resolution than other images, the techniques disclosed herein can influence specific treatments or prevention of diseases or medical conditions because they can reveal information that would otherwise be unavailable in lower-resolution images, while minimizing or reducing radiation and contrast agent exposure to the patient. These techniques can improve patient outcomes by revealing additional information absent in lower-resolution images via synthetic images, which can be used to diagnose and / or treat the patient's medical condition.

[0035] Figure 2 yes Figure 1 This is a schematic diagram of an example of a computing device 150 of system 10. The computing device 150 may include a workstation, desktop computer, laptop computer, smartphone, tablet, dedicated computing device, or any other computing device capable of performing the techniques of this disclosure.

[0036] The computing device 150 can be configured to perform processing, control, and other functions associated with the imager 140. For example... Figure 2 As shown, computing device 150 may represent multiple instances of computing devices, each of which may be associated with imager 140. Computing device 150 may include, for example, memory 202, processing circuitry 204, display 206, network interface 208, input device 210 and / or output device 212, each of which may represent any of multiple instances of such devices within a computing system for ease of description.

[0037] Although processing circuit 204 appears Figure 2In computing device 150, however, in some examples, features attributable to processing circuitry 204 may be executed by processing circuitry of any of computing device 150, imager 140, or server 160, or combinations thereof. In some examples, one or more processors associated with processing circuitry 204 in the computing system may be distributed and shared across any combination of computing device 150, imager 140, and server 160. Computing device 150 may be used to perform any of the techniques described in this disclosure and may, individually or in combination with other components such as computing device 150, imager 140, server 160, or components of a system including any or all such systems, form all or part of an apparatus or system configured to perform such techniques.

[0038] The memory 202 of the computing device 150 includes any non-transitory computer-readable storage medium for storing data or software that can be executed by the processing circuitry 204 and, where applicable, control the operation of the computing device 150 and / or the imager 140. In one or more examples, the memory 202 may include one or more solid-state storage devices, such as flash memory chips. In one or more examples, the memory 202 may include one or more mass storage devices connected to the processing circuitry 204 via a mass storage controller (not shown) and a communication bus (not shown).

[0039] While the description of computer-readable media herein refers to solid-state storage devices, those skilled in the art will understand that computer-readable storage media can be any available medium accessible by processing circuitry 204. That is, computer-readable storage media includes non-transitory, volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information such as computer-readable instructions, data structures, program modules, or other data. For example, computer-readable storage media includes RAM, ROM, EPROM, EEPROM, flash memory or other solid-state storage technologies, CD-ROM, DVD, Blu-ray or other optical storage devices, magnetic tape cassettes, magnetic tape, disk storage devices or other magnetic storage devices, or any other medium that can be used to store desired information and can be accessed by computing device 150. In one or more examples, computer-readable storage media may be stored in a cloud or remote storage device and accessed via at least one wired or wireless connection using any suitable one or more technologies.

[0040] Memory 202 may store Medical Digital Imaging and Communication (DICOM) files 220 and DICOM files 230. DICOM file 220 may include multiple DICOM files, each including one or more associated images 214. In some examples, image 214 may not be part of DICOM file 220, but may be stored in memory 202. DICOM file 230 may include composite image 216. In some examples, composite image 216 may not be part of DICOM file 230, but may be stored in memory 202.

[0041] DICOM files typically include a header and image data. The header usually includes information about the patient and information about the image data, such as the date the image data was acquired, the equipment used to acquire the image data, and the dimensions of the image data. In some examples, processing circuitry 204 may combine or average metadata from those DICOM files 220 (whose images are used to generate composite image 216) and store such metadata in DICOM file 230. DICOM file 230 will be discussed later in this disclosure. Figure 5 Further discussion.

[0042] Image 214 may include, for example, multiple images obtained from imager 140. Image 214 may include at least some images with relatively low (e.g., at least 0.3 mm) resolution. Composite image 216 may be an image generated by processing circuitry 204 based on image 214. For example, processing circuitry 204 may average image 214 to generate higher resolution composite image 216.

[0043] Image 214 may be generated by imager 140 of a patient's anatomy and obtained by computing device 150 via a network interface 208 communicatively coupled to imager 140. In some examples, image 214 may include images generated at different times, such as during individual medical procedures. In some examples, image 214 may include images generated by imagers of different modalities (e.g., CT imagers, fluoroscopy imagers, etc.) such as imager 140 and / or generated at different medical facilities.

[0044] For example, image 214 can be generated by imager 140 ( Figure 1 Image 214 is captured during multiple individual medical procedures for the patient. Processing circuitry 204 can acquire image 214 from imager 140 and store image 214 in memory 202. Processing circuitry 204 can execute user interface 218 to enable display 206 (and / or Figure 1 The display device 110 presents a user interface 218 to one or more clinicians performing medical procedures.

[0045] The memory 202 may also store one or more machine learning models 222 and a user interface 218. The machine learning model 222 may be configured, when executed by the processing circuitry 204, to determine a reference marker in a first image of image 214, identify reference markers in the remaining set of images 214, and / or align each image of the remaining set of images 214 with the first image based on the reference markers. Proper alignment of image 214 (or at least those images of image 214 that include reference markers) facilitates the processing circuitry 204 in generating a higher-resolution synthetic image 216 based on image 214. Additionally or alternatively, the machine learning model 222 may be configured, when executed by the processing circuitry 204, to quantify any diameter variations between contraction and relaxation phases in image 214 and / or in synthetic image 216 and / or in multiple synthetic images (such as synthetic image 216).

[0046] For example, processing circuitry 204 may use reference markers when generating composite image 216. Reference markers may be anatomical (e.g., based on cardiac or vascular structures) or non-anatomical (e.g., medical implants, etc.). For example, for a patient who has previously had a stent implanted, the stent itself may be used as a reference marker or the basis of a coordinate system. For example, processing circuitry 204 may use a vessel wall, a specific vessel, a valve, a sternal ligature, a pacemaker, etc., as reference markers for aligning the image in image 214.

[0047] In some examples, processing circuitry 204 may determine a specific individual image frame in image 214 that will be used to generate composite image 216. For example, processing circuitry 204 may generate composite image 216 based on multiple low-resolution images of the same phase of the cardiac cycle.

[0048] When generating the composite image 216, the processing circuitry 204 can average information from the image 214, such as pixel information. For example, the processing circuitry can determine the median pixel value associated with a region of the patient's anatomy to determine the pixel values ​​for the same region of the patient's anatomy in the composite image 216. For example, the processing circuitry 204 can create a composite or compounded higher-resolution image (e.g., composite image 216) from the average of multiple lower-resolution images (e.g., image 214). In some examples, multiple lower-resolution images may be taken during a single medical procedure. In some examples, multiple lower-resolution images may be taken over multiple separate medical procedures over time. By taking images over time, such as image 214, clinicians can track disease progression over time.

[0049] In some examples, multiple lower-resolution images may be part of a DICOM file 220. In this case, the processing circuit 204 may use the image data within the DICOM file to generate a composite image 216.

[0050] In some examples, sensor 214 may include one or more relatively high-resolution images. For example, imager 140 may capture one or more high-resolution images of a region of particular interest or the entire region of interest (e.g., the left main coronary artery or the entire heart). In some examples, imager 140 may capture multiple high-resolution images of different portions of the region of interest over time, and processing circuitry 204 may use the multiple high-resolution images to generate a composite image 216 to produce a composite ultra-high-resolution image, or it may use multiple high-resolution images and multiple low-resolution images to generate a composite image 216.

[0051] For example, imager 140 may capture high-resolution images during the initial patient visit (which can be considered a medical procedure) and only low-resolution images during subsequent visits. In such an example, processing circuitry 204 may generate a synthetic image 216 based solely on the low-resolution image in image 214. In this case, clinicians can compare the synthetic image 216 with the initial high-resolution image to determine disease progression.

[0052] In some examples, the low-resolution images include imaging data of portions of the heart that differ from those represented in the initial high-resolution images, and / or imaging data of portions of the heart that overlap with the imaging data of the initial high-resolution images. In some examples, each low-resolution image in the low-resolution image set includes imaging data of a different portion of the heart. In some examples, the segment of the heart to be included in a given low-resolution image is determined prior to the appropriate medical protocol and based on external data (e.g., electrocardiogram (ECG) data indicating a potential specific problem, such as data from stress testing). In some examples, the segment of the heart to be included in a given low-resolution image is the left main coronary artery (LM), because undiagnosed LM disease is more likely to be fatal.

[0053] Processing circuitry 204 may be implemented by one or more processors, which may include any number of fixed-function circuits, programmable circuits, or combinations thereof. In various examples, control of any function by processing circuitry 204 may be implemented directly or in conjunction with any suitable electronic circuitry appropriate for the specified function. Fixed-function circuitry refers to circuitry that provides specific functionality and is pre-configured for executable operations. Programmable circuitry refers to circuitry that can be programmed to perform various tasks and provide flexible functionality in executable operations. For example, programmable circuitry may execute software or firmware that causes the programmable circuitry to operate in a manner defined by the instructions of the software or firmware. Fixed-function circuitry may execute software instructions (e.g., receiving or outputting parameters), but the type of operation performed by fixed-function circuitry is generally immutable. In some examples, one or more units within the unit may be different circuit blocks (fixed-function or programmable), and in some examples, the one or more units may be integrated circuits.

[0054] Instructions may be executed by one or more processors, such as one or more digital signal processors (DSPs), general-purpose microprocessors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), graphics processing units (GPUs), or other equivalent integrated or discrete logic circuits. Therefore, the term "processing circuit 204" as used herein may refer to one or more processors having any of the foregoing processors or processing structures, or any other structure suitable for implementing the techniques described herein. Additionally, in some aspects, the functionality described herein may be provided within dedicated hardware or software modules configured for encoding and decoding, or incorporated into combined codecs. Furthermore, these techniques may be fully implemented in one or more circuit or logic elements.

[0055] The display 206 may be touch-sensitive or voice-activated, enabling it to function as both an input and output device. Alternatively, a keyboard (not shown), a mouse (not shown), or other data input devices (e.g., input device 210) may be used.

[0056] Network interface 208 may be adapted to connect to a network, such as a local area network (LAN), wide area network (WAN), wireless mobile network, Bluetooth network, or the Internet, including wired or wireless networks. For example, computing device 150 may acquire image 214 from imager 140 during a medical procedure. Computing device 150 may receive updates to its software (e.g., application 217) via network interface 208. Computing device 150 may also display a notification on display 206 that a software update is available.

[0057] Input device 210 may include any device that enables a user to interact with computing device 150, such as, for example, a mouse, keyboard, foot pedal, touch screen, augmented reality input device that receives input such as gestures or body movements, or voice interface.

[0058] Output device 212 may include any connection port or bus, such as, for example, a parallel port, a serial port, a universal serial bus (USB), or any other similar connection port known to those skilled in the art.

[0059] Application 217 may be one or more software programs stored in memory 202 and executed by processing circuitry 204 of computing device 150. Processing circuitry 204 may execute user interface 218, which may display images 214, composite images 216, DICOM files 220 and / or DICOM files 230 on display 206 and / or display device 110. Clinicians may use the displayed images or files for diagnosis, tracking the progression of medical conditions over time, determining treatment strategies, monitoring the progress of interventional medical procedures, etc.

[0060] Figure 3 This is a conceptual diagram illustrating an example averaging technique according to one or more aspects of this disclosure. Image 300 represents a first image of a portion of a captured blood vessel 310. Image 302 represents a second image of the same portion of the captured blood vessel 310 as image 300. Images 300 and 302 may be image 214 ( Figure 2 Examples of two images are provided. Each square within images 300 and 302 represents a pixel with a resolution of 0.3 mm. Images 300 and 302 use cross-hatching to depict those pixels indicating the vessel wall. Figure 3 For the purpose of alignment, images 300 and 302 are typically based, for example, on reference marks that can be used to align images 300 and 302. Figure 3 (Not shown in the image) for alignment. As can be seen, in image 300, the boundary of blood vessel 310 is detected at a location different from the boundary of blood vessel 310 in image 302.

[0061] Image 304 may represent a composite image based on images 300 and 302 (e.g., Figure 2 (Example of the composite image 216). Image 304 may have a higher resolution than images 300 and 302, at 0.15 mm per pixel. Therefore, the vessel wall of blood vessel 310 can be located more accurately, as shown in image 304.

[0062] Figure 4 This is a conceptual diagram illustrating example alignment and averaging techniques according to one or more aspects of this disclosure. Images 400, 402, and 404 can be multiple images 214 ( Figure 2Example images are shown in (). Each of images 400, 402, and 404 is depicted as having the same resolution, where each small square within each image represents a pixel. Image 402 is depicted as vertically offset from image 400. Image 404 is depicted as both vertically and horizontally offset from image 400. It should be noted that any of the plurality of images 214 may be vertically, horizontally, or both vertically and horizontally offset from each other. In some examples, one or more of the plurality of images 214 may be substantially offset from each other, such that a reference mark represented in one image may not be represented in another image.

[0063] The processing circuit 204 can determine a reference marker (marked with an X) in the image 400. Although shown as a single pixel for illustrative purposes, it should be understood that the reference marker can be larger than a single pixel; for example, the reference marker can occupy multiple pixels.

[0064] Processing circuitry 204 can identify the reference marker X in image 402. Image 402 may, for example, be offset from image 400 in the vertical direction as shown. Therefore, the bottom right pixel L of image 400 will not represent the same anatomical structure as the bottom right pixel L of image 402, which will represent an anatomical structure that does not even exist in image 400. Therefore, averaging pixel L will result in blurring or distortion of images 400 and 402, rather than a composite image with a higher resolution than images 400 and 402.

[0065] The processing circuit 204 can also identify a reference mark X in image 404. Image 404 can be offset from image 400, for example, in the horizontal and vertical directions as shown. The processing circuit 204 can identify the reference mark X in multiple images 214 ( Figure 2Pixel A is represented in more than one image (e.g., one pixel above and one pixel to the right of reference marker X). For example, pixel A is represented in each of images 400, 402, and 404. Processing circuit 204 may perform averaging on pixels from the three images 400, 402, and 404. For example, processing circuit 204 may determine the median of pixel A in images 400, 402, and 404, and may use this median as the value of the associated pixel A in a composite image (such as composite image 216). Processing circuit 204 may perform the same technique for each pixel relative to reference marker X. In some instances, due to the offset of the images from one another, a pixel may appear in only a subset of all images 400, 402, and 404. In such cases, processing circuit 204 may perform averaging on such pixels, excluding images in which that particular pixel does not appear. For example, processing circuit 204 may determine the median of corresponding pixels in multiple images 400, 402 and 404 (or multiple images 214) to generate the average pixel value of the corresponding pixels in order to generate each pixel in the composite image 216.

[0066] Figure 5 This is a conceptual diagram illustrating example content of a DICOM file representing a composite image according to one or more aspects of this disclosure. DICOM file 500 can be DICOM file 230 (…). Figure 2 Example of a DICOM file 500. The DICOM file 500 may include a file header 502 and image data 504. The file header 502 may include metadata 512. The metadata 512 may include information associated with the image data 504, such as patient identification information, modality-related information, settings, and the date associated with the image data 504. The metadata 512 may include a date 522, which may include date information about the time when each image used by the processing circuitry 204 to generate the composite image 216 was generated. For example, if the processing circuitry 204 uses images from multiple images 214 generated by an imager (such as imager 140) on January 31, 2020, and July 1, 2021, then the date 522 would include both January 31, 2020, and July 1, 2021. Therefore, the metadata 512 may retain and include information associated with each image used to generate the composite image 216, such as the date. The date 522 may also include the date on which the composite image 216 was generated.

[0067] Image data 504 may include composite image 216. In some examples, image data 504 may be stored in a normalized format such as JPEG, TIFF, GIF, PNG, etc. Image data 504 may include luminance data 514. For example, each pixel of composite image 216 may have an associated luminance, which may be stored in luminance data 514.

[0068] Figure 6 This is a flowchart of an example technique for generating a synthetic image according to one or more aspects of this disclosure. The following describes processing circuitry 204. Figure 6 The technology, but such technology can be developed by Figure 1 The processing circuitry of the device depicted or capable of performing such technology may be used to execute it.

[0069] Processing circuitry 204 can acquire multiple images (600). For example, processing circuitry 204 can retrieve multiple images 214 from memory 202 and / or receive multiple images 214 from imager 140 or server 160 via network interface 208. Each of the multiple images has a corresponding resolution, and the multiple images are generated during at least two separate medical procedures that occur at at least one of different times or different dates. For example, separate medical procedures can be spaced out in time, such as by hours, days, weeks, months, years, etc. It should be understood that different steps of a procedure arranged for a patient to arrive at a specific time on a specific date for a medical procedure should not be considered separate medical procedures, even if the different steps involve different clinicians, equipment, locations, etc.

[0070] Processing circuit 204 can generate a composite image based on averaging multiple images, the composite image having a higher resolution than at least a majority of the multiple images (602). A majority may be greater than 50%. For example, processing circuit 204 can generate a composite image 216 by averaging the pixel values ​​of the pixels in the multiple images 214 (e.g., determining the median of the pixel values). In this way, processing circuit 204 can generate a composite image 216 that has a relatively higher resolution compared to the multiple images 214.

[0071] Processing circuit 204 can output a composite image (604). For example, processing circuit 204 can output composite image 216 to display 206 and / or display device 110 for display, or output composite image 216 to server 160 ( Figure 1 (For storage purposes.) Clinicians may view the synthetic images 216 for diagnostic and / or therapeutic purposes.

[0072] In some examples, processing circuitry 204 is configured to determine a reference marker (e.g., reference marker X) in a first image (e.g., image 400) among a plurality of images 214. In such examples, processing circuitry 204 may identify the reference marker in the remaining set of images among the plurality of images (e.g., images 402 and 404). In such examples, processing circuitry 204 may align each image in the remaining set of images among the plurality of images with the first image based on the reference marker. In some examples, processing circuitry 204 is configured to execute machine learning model 222 to perform at least one of the following: determine the reference marker, identify the reference marker, or align each image in the remaining set of images among the plurality of images with the first image. In some examples, machine learning model 222 is trained based on images of the current patient's anatomy, images of the anatomy of other patients, etc. Training data may include image data from past medical procedures performed on multiple patients with different patient conditions and different prior medical procedures, annotations or labels associated with the image data, etc.

[0073] In some examples, aligning each of the remaining images in the plurality of images with the first image makes the corresponding pixel (e.g., pixel A) of each of the images in the plurality of images aligned, and wherein averaging the plurality of images includes determining the median of the corresponding pixel in the plurality of images. In some examples, the corresponding pixel includes luminance information (e.g., luminance data 514).

[0074] In some examples, processing circuitry 204 is configured to execute machine learning model 222 to determine the change in anatomical diameter of a patient’s anatomical structures between the systolic and diastolic phases based on at least one of image 214 or synthetic image 216.

[0075] In some examples, at least one of the multiple images has a higher resolution than the remaining images. In some examples, at least one of the multiple images was generated during the earliest of at least two separate medical procedures. In some examples, at least one of the multiple images includes image data of a region of interest for the patient, which was identified as a region associated with a relatively higher risk compared to other regions of the patient.

[0076] In some examples, each of the multiple images is contained within a corresponding Medical Digital Imaging and Communications (DICOM) file (e.g., DICOM file 220) stored in memory, and wherein, as part of acquiring the multiple images, processing circuitry is configured to extract the multiple images 214 from the corresponding DICOM file. In some examples, processing circuitry 204 is also configured to generate a DICOM file (e.g., DICOM file 230) including a composite image 216, and to perform at least one of the following: storing the DICOM file in memory, or outputting the DICOM file to a device for storage. In some examples, processing circuitry 204 is also configured to generate metadata associated with the composite image (e.g., metadata 512), which is based on the metadata associated with the multiple images. In some examples, the metadata associated with the composite image includes multiple dates (e.g., date 522) corresponding to dates of at least two separate medical procedures. In some examples, processing circuitry 204 is also configured to insert the metadata associated with the composite image into the Medical Digital Imaging and Communications (DICOM) file.

[0077] In some examples, multiple images are based on multiple imaging modalities. In some examples, at least one of the multiple images is generated using at least one of fluoroscopy or computed tomography.

[0078] Figure 7 This is a conceptual diagram illustrating an example machine learning model according to one or more aspects of this disclosure. Machine learning model 700 may be an example of machine learning model 222. Machine learning model 700 may be an example of a deep learning model or deep learning algorithm trained to determine the identification of an implanted medical device and / or to determine a recommended treatment strategy. One or more of computing devices 150 and / or servers 160 may train, store, and / or utilize machine learning model 700, but in some examples, other devices of system 100 may apply input to machine learning model 700. In some examples, and in others, other types of machine learning and deep learning models or algorithms may be utilized. For example, a convolutional neural network model of ResNet-18 may be used. Some non-limiting examples of models that can be used for transfer learning include AlexNet, VGGNet, GoogleNet, ResNet50, or DenseNet. Some non-limiting examples of machine learning techniques include support vector machines, K-nearest neighbor algorithms, and multilayer perceptrons.

[0079] like Figure 7As shown in the example, the machine learning model 700 may include three types of layers. These three types of layers include an input layer 702, a hidden layer 704, and an output layer 706. The output layer 706 includes the output of the transfer function 705 from the output layer 706. The input layer 702 represents each of the input values ​​X1 to X4 provided to the machine learning model 700. In some examples, as described above, the input values ​​may include any of the values ​​input into the machine learning model. For example, as described above, the input values ​​may include image 214. Additionally, in some examples, the input values ​​of the machine learning model 700 may include additional data, such as other data that may be collected by system 100 or stored in the system.

[0080] Each input value from the input values ​​of each node in input layer 702 is provided to each node in the first layer of hidden layer 704. Figure 7 In the example, hidden layer 704 comprises two layers, one with four nodes and the other with three nodes, but fewer or more nodes may be used in other examples. Each input from input layer 702 is multiplied by a weight, and then summed at each node of hidden layer 704. During training of machine learning model 700, the weights of each input are adjusted to establish a relationship between image 214, latent benchmarks that can be found in one or more images within image 214, and / or alignments of image 214 based on one or more benchmarks. In some examples, one hidden layer may be combined into machine learning model 700, or three or more hidden layers may be combined into machine learning model 700, where each layer comprises the same or different numbers of nodes.

[0081] The results of each node within hidden layer 704 are applied to the transfer function of output layer 706. The transfer function can be linear or nonlinear, depending on the number of layers within the machine learning model 700. Example nonlinear transfer functions could be sigmoid functions or rectified functions. The output 707 of the transfer function can be image 214 and / or 3D composite image 216 indicating a classification of a specific implanted medical device and / or a specific recommended treatment strategy.

[0082] As illustrated in the examples above, by applying machine learning model 700 to input data such as image 214, processing circuitry 204 is able to determine reference markers in one image, identify reference markers in other images, and / or align images based on reference markers. This improves the ability of processing circuitry 204 to generate synthetic image 216 as an image with a higher resolution than one or more of the images in image 214. Additionally or alternatively, by applying machine learning model 700 to input data (such as image 214) and / or synthetic images (such as synthetic image 216), processing circuitry 204 is able to quantify any diameter changes in patient anatomy represented in such images between systolic and diastolic phases.

[0083] Figure 8 This is a conceptual diagram illustrating an example training process for a machine learning model according to one or more aspects of this disclosure. Process 870 can be used to train machine learning model 222 or machine learning model 700. Machine learning model 874 (which can be an example of machine learning model 700 and / or machine learning model 222) can be implemented using any number of models for supervised and / or reinforcement learning, such as, but not limited to, artificial neural networks, decision trees, Naive Bayes networks, support vector machines or k-nearest neighbor models, convolutional neural networks (CNNs), region neural networks (RNNs), long short-term memory (LSTMs), ensemble networks, to name just a few examples.

[0084] In some examples, one or more of the computing device 150 and / or server 160 initially train the machine learning model 874 based on a corpus of training data 872. Training data 872 may include, for example, images of the anatomy of the current patient, images of the anatomy of other patients, etc. In some examples, training data 872 may include annotations identifying one or more benchmark markers and / or anatomy structures contained in the images of training data 872. Training data 872 may include data from past medical procedures performed on multiple patients with different patient conditions and different prior medical procedures, annotations or labels, other training data mentioned herein, etc. In some examples, training data 872 may include images of different resolutions, including lower-resolution images and higher-resolution images.

[0085] When training the machine learning model 874, the processing circuitry of system 100 can compare the prediction or classification with the target output 878 876. Processing circuitry 204 can use the error signal from this comparison to train (learn / train 880) the machine learning model 874. Processing circuitry 204 can generate machine learning model weights or other modifications, which can be used to modify the machine learning model 874. For example, processing circuitry 204 can modify the weights of the machine learning model 874 based on learning / training 880. For example, one or more of computing devices 150 and / or servers 160 can, for each training instance in training data 872, modify, based on training data 872, the way in which reference markers are determined, reference markers are identified, and / or images are aligned, and / or the way anatomical diameters between systolic and diastolic phases are quantified.

[0086] The techniques described in this disclosure can be implemented, at least in part, in hardware, software, firmware, or any combination thereof. For example, aspects of the techniques can be implemented within one or more processors or processing circuits, including one or more microprocessors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or any other equivalent integrated or discrete logic circuits, and any combination of such components. The terms “controller,” “processor,” or “processing circuit” generally refer to any of the aforementioned logic circuits individually or in combination with other logic circuits, or any circuit of any other equivalent nature. Control units including hardware can also perform one or more of the techniques disclosed herein. Such hardware, software, and firmware can be implemented within the same device or in separate devices to support the various operations and functions described in this disclosure. Furthermore, any described unit, circuit, or component can be implemented together or independently as discrete but interoperable logic devices. Describing different features as circuits or units is intended to highlight different functional aspects and does not necessarily imply that such circuits or units must be implemented by separate hardware or software components. Conversely, the functions associated with one or more circuits or units may be performed by independent hardware or software components, or integrated within common or independent hardware or software components.

[0087] The techniques described in this disclosure can also be embedded or encoded in a computer-readable medium (such as a computer-readable storage medium) containing instructions. Instructions embedded or encoded in a computer-readable storage medium can cause a programmable processor or other processor to perform the method, for example, when executing those instructions. The computer-readable storage medium may include random access memory (RAM), read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), or electrically erasable programmable read-only memory (EEPROM) or other computer-readable media.

[0088] This disclosure includes the following non-limiting embodiments.

[0089] Example 1. A medical system comprising: a memory configured to store a plurality of images of a patient, each of the plurality of images having a corresponding resolution, wherein the plurality of images are generated during at least two separate medical procedures, the at least two separate medical procedures occurring at at least one of different times or different dates; and processing circuitry communicatively coupled to the memory, the processing circuitry being configured to: acquire the plurality of images; generate a composite image based on averaging the plurality of images, the composite image having a higher resolution than at least a majority of the plurality of images; and output the composite image.

[0090] Example 2. The medical system according to Example 1, wherein the processing circuit is further configured to: determine a reference marker in a first image of the plurality of images; identify the reference marker in the remaining set of the plurality of images; and align each of the images in the remaining set of the plurality of images with the first image based on the reference marker.

[0091] Example 3. The medical system according to Example 2, wherein the processing circuit is configured to execute a machine learning model to perform at least one of the following: determining the reference marker, identifying the reference marker, or aligning each of the images in the remaining set of the plurality of images with the first image.

[0092] Example 4. The medical system according to Example 3, wherein the machine learning model is trained based on multiple past images.

[0093] Example 5. A medical system according to any one of Examples 2 to 4, wherein each of the images in the remaining set of the plurality of images is aligned with the first image such that the corresponding pixels of each of the images in the plurality of images are aligned, and wherein averaging the plurality of images includes determining the median of the corresponding pixels in the plurality of images.

[0094] Example 6. The medical system according to Example 5, wherein the corresponding pixel includes brightness information.

[0095] Example 7. A medical system according to any one of Examples 1 to 6, wherein the processing circuit is configured to execute a machine learning model to determine, based on at least one of the plurality of images or the synthetic image, the change in anatomical diameter of the patient's anatomical structures between the systolic and diastolic phases.

[0096] Example 8. A medical system according to any one of Examples 1 to 7, wherein at least one of the plurality of images has a higher resolution than the remaining images of the plurality of images.

[0097] Example 9. The medical system according to Example 8, wherein at least one of the plurality of images was generated during the earliest of the at least two separate medical procedures.

[0098] Example 10. A medical system according to Example 8 or Example 9, wherein at least one of the plurality of images includes image data of a region of interest of the patient, the region of interest being identified as a region associated with a relatively higher risk compared to other regions of the patient.

[0099] Example 11. A medical system according to any one of Examples 1 to 10, wherein each of the plurality of images is contained in a corresponding Medical Digital Imaging and Communication (DICOM) file stored in the memory, and wherein, as part of acquiring the plurality of images, the processing circuitry is configured to extract the plurality of images from the corresponding DICOM file.

[0100] Example 12. The medical system according to Example 11, wherein the processing circuit is further configured to: generate a DICOM file including the composite image; and perform at least one of the following: storing the DICOM file in the memory, or outputting the DICOM file to a device for storage.

[0101] Example 13. The medical system according to Example 12, wherein the processing circuit is further configured to generate metadata associated with the synthetic image, the metadata being based on metadata associated with the plurality of images.

[0102] Example 14. The medical system according to Example 13, wherein the metadata associated with the synthetic image includes multiple dates corresponding to the dates of the at least two separate medical procedures.

[0103] Example 15. The medical system according to Example 14, wherein the processing circuitry is further configured to insert the metadata associated with the synthetic image into a Medical Digital Imaging and Communication (DICOM) file.

[0104] Example 16. A medical system according to any one of Examples 1 to 15, wherein the plurality of images are based on a plurality of imaging modalities.

[0105] Example 17. The medical system according to any one of Examples 1 to 16, wherein at least one of the plurality of images is generated using at least one of fluoroscopy or computed tomography.

[0106] Example 18. A medical system according to any one of Examples 1 to 17, the medical system further comprising: a display, wherein the processing circuitry is configured to output the composite image for display on the display.

[0107] Example 19. A method comprising: obtaining a plurality of images by processing circuitry of a medical system, each of the plurality of images having a corresponding resolution, wherein the plurality of images are generated during at least two separate medical procedures, the at least two separate medical procedures occurring at at least one of different times or different dates; generating a composite image by the processing circuitry based on averaging the plurality of images, the composite image having a higher resolution than at least a majority of the plurality of images; and outputting the composite image by the processing circuitry.

[0108] Example 20. A non-transitory computer-readable storage medium storing instructions that, when executed, cause a processing circuit to: acquire a plurality of images, each of the plurality of images having a corresponding resolution, wherein the plurality of images are generated during at least two separate medical procedures, the at least two separate medical procedures occurring during at least one of different times or different dates; generate a composite image based on averaging the plurality of images, the composite image having a higher resolution than at least most of the plurality of images; and output the composite image.

[0109] Various embodiments have been described. These and other embodiments are within the scope of the appended claims.

Claims

1. A medical system for generating synthetic images, the medical system comprising: A memory configured to store multiple images of a patient, each of the multiple images having a corresponding resolution, wherein the multiple images were generated during at least two separate medical procedures, the at least two separate medical procedures occurring at at least one of different times or different dates; as well as Processing circuitry, communicatively coupled to the memory, is configured to: Obtain the multiple images; A composite image is generated by averaging the plurality of images, and the composite image has a higher resolution than at least most of the plurality of images. as well as Output the synthesized image.

2. The medical system according to claim 1, wherein the processing circuit is further configured to: A reference marker is determined in the first image among the plurality of images; Identify the reference marker in the remaining set of the plurality of images; and Align each of the remaining images in the plurality of images with the first image based on the reference marker.

3. The medical system of claim 2, wherein the processing circuitry is configured to execute a machine learning model to perform at least one of the following: determining the reference marker, identifying the reference marker, or aligning each of the images in the remaining set of the plurality of images with the first image.

4. The medical system according to claim 3, wherein the machine learning model is trained based on multiple past images.

5. The medical system according to any one of claims 2 to 4, wherein aligning each of the images in the remaining set of the plurality of images with the first image such that corresponding pixels of each of the images in the plurality of images are aligned, and wherein averaging the plurality of images includes determining the median of the corresponding pixels in the plurality of images.

6. The medical system according to claim 5, wherein the corresponding pixel includes brightness information.

7. The medical system according to any one of claims 1 to 6, wherein the processing circuitry is configured to execute a machine learning model to determine, based on at least one of the plurality of images or the synthetic image, the change in anatomical diameter of the patient's anatomical structures between systolic and diastolic phases.

8. The medical system according to any one of claims 1 to 7, wherein at least one of the plurality of images has a higher resolution than the remaining images of the plurality of images.

9. The medical system of claim 8, wherein at least one of the plurality of images was generated during the earliest of the at least two separate medical procedures.

10. The medical system of claim 8 or claim 9, wherein at least one of the plurality of images includes image data of a region of interest of the patient, the region of interest being identified as a region associated with a relatively higher risk compared to other regions of the patient.

11. The medical system according to any one of claims 1 to 10, wherein each of the plurality of images is contained in a corresponding Medical Digital Imaging and Communication (DICOM) file stored in the memory, and wherein, as part of acquiring the plurality of images, the processing circuitry is configured to extract the plurality of images from the corresponding DICOM file.

12. The medical system of claim 11, wherein the processing circuit is further configured to: Generate a DICOM file including the synthesized image; and Perform at least one of the following: store the DICOM file in the memory, or output the DICOM file to a device for storage.

13. The medical system of claim 12, wherein the processing circuitry is further configured to generate metadata associated with the synthetic image, the metadata being based on metadata associated with the plurality of images.

14. A method for generating a synthetic image, the method comprising: Multiple images are obtained by the processing circuitry of a medical system, each of the multiple images having a corresponding resolution, wherein the multiple images are generated during at least two separate medical procedures, which occur at least one of different times or different dates; The processing circuit generates a composite image based on averaging the plurality of images, the composite image having a higher resolution than at least most of the plurality of images; as well as The processing circuit outputs the synthesized image.

15. A non-transitory computer-readable storage medium storing instructions that, when executed, cause processing circuitry to generate a composite image, wherein the processing circuitry: Multiple images are obtained, each of the multiple images having a corresponding resolution, wherein the multiple images are generated during at least two separate medical procedures, the at least two separate medical procedures occurring at at least one of different times or different dates; A composite image is generated by averaging the plurality of images, and the composite image has a higher resolution than at least most of the plurality of images. as well as Output the synthesized image.

Images