An endoscope image analysis method and system

Abstract

This invention provides an endoscopic image analysis method and system. The method includes: acquiring historical image information and constructing a human body simulation model of the patient based on the historical image information; acquiring endoscopic detection information including endoscopic image information, endoscopic trajectory and posture information, and physiological characteristic data of the patient during endoscopic detection, and performing simulation calculations based on the physiological characteristic data and the human body simulation model using a preset simulation model; acquiring target simulation image information during the simulation calculation based on the endoscopic trajectory and posture information, and analyzing the endoscopic image information based on the target simulation image information to determine calibration image information; optimizing the endoscopic image information based on the calibration image information and the target simulation image information to determine a target endoscopic image set, and analyzing the endoscopic image set to determine the patient's condition. This invention solves the problem of the lack of a highly accurate and individually adaptable endoscopic image analysis method and system in the prior art.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and in particular to an endoscopic image analysis method and system. Background Technology

[0002] In modern clinical medicine, endoscopy has become one of the core methods for diagnosing and assessing diseases of hollow organs such as the digestive and respiratory tracts. With the popularization of minimally invasive medical concepts, endoscopic examination, due to its advantages of minimal trauma, ease of operation, and real-time observation of tissue morphology, is widely used in the screening and diagnosis of diseases such as gastritis, ulcers, polyps, and even early-stage tumors. The clarity and accuracy of endoscopic images directly determine the accuracy of doctors' judgments on the location, size, and shape of lesions, and are a key prerequisite for avoiding missed diagnoses and misdiagnoses, and improving diagnostic and treatment efficiency. Therefore, how to effectively enhance and optimize endoscopic images through technological means has always been an important research direction in the field of medical imaging technology.

[0003] Currently, the enhancement and optimization of endoscopic images mainly relies on image processing algorithms. Existing techniques typically involve statistical analysis of a large number of endoscopic images or image fusion utilizing the spatiotemporal correlation between adjacent frames. For example, methods such as multi-frame registration and super-resolution reconstruction integrate information from a sequence of images to suppress noise and improve resolution. These methods, based on pixel-level processing of the images themselves, improve visual effects to a certain extent and have become common techniques in endoscopic image post-processing.

[0004] However, the aforementioned existing technologies have significant shortcomings. The processing logic of existing technologies is mostly based on static image data, which does not fully adapt to the dynamic changes in the internal organs and ignores the real-time impact of the patient's physiological activities on the shape and position of the organs during the detection process. This leads to data matching deviations when integrating multiple frames of images, which in turn affects the accuracy of the images. In addition, the optimization algorithms of existing technologies mostly adopt a generalized design and do not consider the differences between different patients. They only process image data through standardized processes, resulting in insufficient individual adaptability of the enhancement effect, affecting the accuracy of the images, and making it difficult to accurately analyze the patient's condition based on the images. Summary of the Invention

[0005] Based on this, the purpose of this invention is to provide an endoscopic image analysis method and system, aiming to solve the problem that there is a lack of an endoscopic image analysis method and system with high accuracy and strong individual adaptability in the prior art.

[0006] An endoscopic image analysis method according to an embodiment of the present invention includes:

[0007] Acquire historical image information and construct a human body simulation digital model of the patient based on the historical image information. The historical image information includes at least the image data collected by external detection measures on the patient before endoscopic examination.

[0008] Acquire endoscopic detection information, which includes at least endoscopic image information, endoscopic trajectory and posture information, and physiological characteristic data of the patient during endoscopic detection. Simulation calculations are performed based on the physiological characteristic data and the human body simulation model using a preset simulation model.

[0009] The target simulation image information is obtained during the simulation operation based on the endoscope trajectory and posture information, and the endoscope image information is analyzed based on the target simulation image information to determine the calibration image information. The calibration image information is the endoscope image information that has the same features as the target simulation image information and has a clarity greater than a preset clarity threshold.

[0010] The endoscopic image information is optimized based on the calibration image information and the target simulation image information to determine the target endoscopic image set, and the endoscopic image set is analyzed to determine the patient's condition.

[0011] In addition, the endoscopic image analysis method according to the above embodiments of the present invention may also have the following additional technical features:

[0012] Furthermore, the step of obtaining the target simulation image information during simulation calculation based on the endoscope trajectory and posture information includes:

[0013] Based on the endoscope trajectory and posture information, determine the node position and node shooting posture of the endoscope at different time points in the simulation operation;

[0014] Based on the node position, the real-time simulation model within a preset time range at the corresponding time node is frozen and the target area simulation model is extracted.

[0015] The position of the light projection is determined based on the position of the node, and the projection angle is determined based on the shooting posture of the node and the endoscope parameters.

[0016] Illumination projection is performed within the target region simulation model based on the illumination projection position and the projection angle. The overlapping area between the illumination projection and the target region simulation model is determined, and the target simulation image information is determined based on the overlapping area and the projection angle.

[0017] Furthermore, the step of analyzing the endoscopic image information based on the target simulation image information to determine the calibration image information includes:

[0018] Feature extraction is performed on each image within the target simulation image information to determine the first feature information;

[0019] The first feature information is matched with the preset region feature information to determine the portion of the first feature information whose similarity is greater than the first preset similarity threshold as the second feature information;

[0020] Based on the second feature information, the target simulation image information and the endoscope image information are filtered to determine the target simulation image and endoscope image corresponding to the second feature information, so as to determine the calibration image information according to the target simulation image and the endoscope image.

[0021] Furthermore, the step of determining the calibration image information based on the target simulation image and the endoscopic image includes:

[0022] The target endoscopic image is divided into segments, and the target endoscopic images with a clarity greater than the preset clarity threshold are identified as the first image units.

[0023] Feature extraction is performed on the first image unit to determine the third feature information, and the second feature information and the third feature information are matched for similarity to determine the part of the third feature information with a similarity greater than a preset similarity threshold as the fourth feature information;

[0024] The first image unit is divided according to the fourth feature information to determine the corresponding second image unit, and the information of multiple second image units constitutes the calibration image information.

[0025] Further, after the steps of extracting features from the first image unit to determine third feature information and performing similarity matching between the second feature information and the third feature information to determine the portion of the third feature information with a similarity greater than a preset similarity threshold as fourth feature information, the process includes:

[0026] The first image unit is divided according to the fourth feature information to determine the second image unit and the third image unit corresponding to the fourth feature information;

[0027] Determine whether there is a fifth feature information in the feature information of the third image unit that matches the features of the preset pathological feature library;

[0028] If they exist, the third image unit is divided according to the fifth feature information to determine the fourth image unit and the fifth image unit corresponding to the fifth feature information, and multiple second image units and the fourth image unit constitute the calibration image information.

[0029] Furthermore, the target endoscope image is composed of a second image unit, a fourth image unit, and a fifth image unit. The step of optimizing the endoscope image information to determine the target endoscope image set based on the calibration image information and the target simulation image information includes:

[0030] The fifth image unit is optimized and the optimized image unit is determined based on the calibration image information and the target simulation image information.

[0031] Based on the preset region feature information and the second feature information, the target endoscope image is classified to determine multiple different first region image sets;

[0032] The difference degree is judged for the second image unit, the fourth image unit and the optimized image unit corresponding to a single first region image set, and the optimized image sub-unit of the optimized image unit with a difference degree less than a preset difference threshold is determined.

[0033] Based on the optimized image subunit, the second image unit, the fourth image unit, and the target simulation image information, image fusion enhancement is performed to determine the second region image set, and multiple second region images constitute the target endoscope image set.

[0034] Furthermore, the step of determining the second region image set by performing image fusion enhancement based on the optimized image sub-unit, the second image unit, and the target simulation image information includes:

[0035] When performing image fusion enhancement based on the optimized image subunit, the second image unit, the fourth image unit, and the target simulation image information, the change in physiological feature data at the corresponding time node is obtained. The change includes at least respiratory rate fluctuation value and heart rate fluctuation value, and it is determined whether the change is greater than a preset change threshold.

[0036] If so, the weight parameters of the optimized image subunit, the second image unit, the fourth image unit, and the target simulated image information during the fusion enhancement process are adjusted according to the ratio of the change amount to the preset change threshold, wherein the larger the change amount, the smaller the weight ratio of the optimized image subunit.

[0037] Another object of the present invention is an endoscopic image analysis system, the system comprising:

[0038] The data acquisition module is used to acquire historical image information and construct a human body simulation digital model of the patient based on the historical image information. The historical image information includes at least the image data acquired by external detection measures on the patient before endoscopic examination.

[0039] The simulation module is used to acquire endoscopic detection information, which includes at least endoscopic image information, endoscopic trajectory and posture information, and physiological characteristic data of the patient during endoscopic detection. Based on the physiological characteristic data and the human body simulation model, simulation calculations are performed through a preset simulation model.

[0040] The image information determination module is used to obtain target simulation image information during simulation calculation based on the endoscope trajectory and posture information, and to analyze and determine calibration image information based on the target simulation image information. The calibration image information is the endoscope image information that has the same features as the target simulation image information and has a clarity greater than a preset clarity threshold.

[0041] The image processing module is used to optimize the endoscopic image information based on the calibration image information and the target simulation image information to determine the target endoscopic image set, and to analyze the endoscopic image set to determine the patient's condition.

[0042] Another objective of this invention is to provide a storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the endoscopic image analysis method described above.

[0043] Another object of the present invention is to provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the above-described endoscopic image analysis method.

[0044] This invention constructs a human body simulation model of the patient using image information obtained from external probing before endoscopic image acquisition. This allows for accurate acquisition of relatively macroscopic structural images of the patient's body. The simulation model is then used to perform calculations based on physiological characteristic data during endoscopic examination, thereby determining the patient's internal activity and dynamic changes during the procedure. Furthermore, based on the movement trajectory, posture, and time points during endoscopic examination, simulated images are acquired to determine the corresponding simulated images for each endoscopic image. These simulated images are then calibrated to identify consistent and clear image information within the endoscopic images. This calibration information is then used to optimize other endoscopic image information. This image optimization is based on the individual differences of the patient and the dynamic changing environment, effectively ensuring the accuracy and individual adaptability of the image optimization. Therefore, this invention solves the problem of the lack of a highly accurate and individually adaptable endoscopic image analysis method and system in the prior art. Attached Figure Description

[0045] Figure 1This is a flowchart of the endoscopic image analysis method in the first embodiment of the present invention;

[0046] Figure 2 This is a schematic diagram of the results of the endoscopic image analysis system in the second embodiment of the present invention;

[0047] Figure 3 This is a schematic diagram of the structure of the electronic device in the third embodiment of the present invention;

[0048] The following detailed description, in conjunction with the accompanying drawings, will further illustrate the present invention. Detailed Implementation

[0049] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Several embodiments of the invention are illustrated in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.

[0050] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0051] Example 1

[0052] Please see Figure 1 The figure shows an endoscopic image analysis method in the first embodiment of the present invention, which specifically includes steps S01-S04.

[0053] S01, acquire historical image information and construct a human body simulation digital model of the patient based on the historical image information, wherein the historical image information includes at least the image data collected by external detection measures on the patient before endoscopic examination.

[0054] Specifically, the historical image information can be image information acquired through external probing methods such as computed tomography, magnetic resonance imaging, or 3D ultrasound. This data can provide high-resolution 3D structural information of the patient's internal organs. As an example, and not a limitation, in some optional embodiments, when constructing a human body simulation model, the acquired image data can first be segmented and 3D reconstructed to generate an initial static geometric model of the corresponding organ. Subsequently, based on medical atlases and biomechanical knowledge, tissue material properties, such as elastic modulus and density, as well as physiological constraints, such as the connection relationships and range of motion between organs, are assigned to this geometric model, thereby forming a personalized human body simulation model that can respond to physical simulation. This model can accurately reflect the patient's unique anatomical structure, including the shape, size, and relative position of organs. Furthermore, the human body simulation model constructed in the above manner fundamentally solves the problem that existing generalized models cannot adapt to individual differences. It provides a dynamic carrier that realistically reflects the anatomical structure of a specific patient for subsequent simulation calculations, laying a solid foundation for achieving high-precision personalized image analysis.

[0055] S02, acquire endoscopic detection information, which includes at least endoscopic image information, endoscopic trajectory and posture information, and physiological characteristic data of the patient during endoscopic detection, and perform simulation calculations based on the physiological characteristic data and the human body simulation model through a preset simulation model.

[0056] Specifically, the physiological characteristic data in the endoscopic examination information includes at least real-time monitored respiratory waveforms, electrocardiograms, and body surface movements. During simulation, the real-time collected physiological characteristic data is used as input parameters to drive the human body simulation model in dynamic calculations. For example, respiratory waveform data can be converted into diaphragmatic motion boundary conditions, and electrocardiogram signals can be correlated with cardiovascular pulsation models, thereby simulating the real-time deformation, displacement, and motion states of the patient's internal organs during endoscopic examination due to physiological activities such as breathing and heartbeat. This process generates a dynamically changing real-time simulation model synchronized with the examination timeline. This then activates the static human body simulation model, enabling it to realistically simulate the dynamic changes of organs during the examination. This effectively overcomes the shortcomings of existing technologies that ignore the influence of physiological activities and treat dynamic scenes as static, allowing the subsequently acquired target simulation image information to highly match the dynamic environment of the real endoscopic image, significantly improving the accuracy of subsequent image matching and optimization.

[0057] S03, obtain the target simulation image information during simulation calculation based on the endoscope trajectory and posture information, and analyze the endoscope image information based on the target simulation image information to determine the calibration image information. The calibration image information is the endoscope image information that has the same features as the target simulation image information and has a clarity greater than a preset clarity threshold.

[0058] Specifically, the steps for obtaining the target simulation image information during simulation calculation based on the endoscope trajectory and posture information include:

[0059] Based on the endoscope trajectory and posture information, the node positions and node shooting postures of the endoscope at different time points in the simulation operation are determined. The real-time simulation model within a preset time range at the corresponding time point is frozen and cropped to determine the target region simulation model based on the node positions. The illumination projection position is determined based on the node positions, and the projection angle is determined based on the node shooting postures and endoscope parameters. Illumination projection is performed within the target region simulation model based on the illumination projection position and the projection angle, determining the overlapping area between the illumination projection and the target region simulation model, and determining the target simulation image information based on the overlapping area and the projection angle. In specific implementation, the simulation model at the corresponding time point is cropped based on the trajectory information during endoscope detection to ensure matching between the two. Then, using the node positions, posture information, and endoscope parameter information, illumination is applied in a preset direction at the corresponding node. The projection image of the simulation model is obtained along the illumination direction from the node positions, and the projection image is divided based on the range of the illumination projection to determine the target simulation image information corresponding to the endoscope image at the current time point.

[0060] Furthermore, the step of analyzing the endoscopic image information based on the target simulation image information to determine the calibration image information includes:

[0061] Feature extraction is performed on each image within the target simulation image information to determine first feature information; the first feature information is matched with the feature information of a preset region to determine the portion of the first feature information with a similarity greater than a first preset similarity threshold as second feature information; based on the second feature information, the target simulation image information and the endoscope image information are filtered to determine the target simulation image and endoscope image corresponding to the second feature information, so as to determine the calibration image information according to the target simulation image and the endoscope image.

[0062] In practice, during endoscopic inspection, images are not captured directly on a single area, but rather within a defined trajectory range. This involves acquiring images from multiple different regions, resulting in a large number of endoscopic images, including those from non-target areas. Therefore, image filtering is necessary to reduce subsequent processing volume. However, real-time acquired endoscopic images, before optimization, suffer from issues such as image clarity, blurriness, and occlusion, making it difficult to quickly and accurately extract corresponding feature information for filtering. Simulated images, on the other hand, are mapped to individual endoscopic images and are derived from digital simulation models, ensuring sufficient clarity and eliminating obstructions. Therefore, feature extraction and filtering of simulated images are performed, and the corresponding endoscopic images are determined based on these filtered images, thus achieving efficient endoscopic image filtering. It should be noted that although the simulation model is constructed from image information obtained from external probes, and external probes cannot accurately capture the subtle structures such as color, gloss, blood vessels, and texture of the mucosal surface inside the cavities of internal organs, they can accurately capture the structure, outline, and morphology of organs. These characteristics are sufficient to filter images of the observation area.

[0063] Furthermore, the step of determining the calibration image information based on the target simulation image and the endoscopic image includes:

[0064] The target endoscopic image is divided into segments, and images with a clarity greater than a preset clarity threshold are identified as first image units. Feature extraction is performed on the first image units to determine third feature information. The second and third feature information are then matched for similarity, and the portion of the third feature information with a similarity greater than a preset similarity threshold is identified as fourth feature information. Based on the fourth feature information, the first image units are further divided to determine corresponding second image units. Information from multiple second image units constitutes the calibration image information. In real-time, feature extraction is performed on the filtered endoscopic images, and the features are compared with those of the simulation image. Regions with consistent features and sufficient clarity on each endoscopic image are identified as calibration image information, serving as a benchmark and basis for subsequent image optimization and enhancement, ensuring the accuracy of image optimization and enhancement.

[0065] Additionally, after the steps of extracting features from the first image unit to determine third feature information and performing similarity matching between the second feature information and the third feature information to determine the portion of the third feature information with a similarity greater than a preset similarity threshold as fourth feature information, the process includes:

[0066] The first image unit is divided according to the fourth feature information to determine the second and third image units corresponding to the fourth feature information; it is then determined whether the feature information of the third image unit contains a fifth feature information that matches the features of a preset pathological feature library; if so, the third image unit is divided according to the fifth feature information to determine the fourth and fifth image units corresponding to the fifth feature information, and multiple second image units and the fourth image unit constitute the calibration image information. Furthermore, in specific implementations, images with sufficient clarity but inconsistent features with the simulation image need to be compared with the preset pathological feature library. This avoids situations where the time interval between the patient's external probing and endoscopic examination is too long, or the condition changes drastically, resulting in images of sudden lesions in the patient's body, thus preventing feature inconsistencies. This effectively prevents the risk of "optimizing" away real lesions, allowing the system to not only optimize image quality but also preserve and highlight potential pathological information. If there is a mismatch, it is judged as an artifact, and multiple second image units constitute the calibration image information.

[0067] S04, optimize the endoscopic image information based on the calibration image information and the target simulation image information to determine the target endoscopic image set, and analyze the endoscopic image set to determine the patient's condition.

[0068] Specifically, the step of optimizing the endoscope image information based on the calibration image information and the target simulation image information to determine the target endoscope image set includes:

[0069] The fifth image unit is optimized and an optimized image unit is determined based on the calibration image information and the target simulation image information; the target endoscope image is classified and multiple different first region image sets are determined based on the preset region feature information and the second feature information; the difference degree of the second image unit, the fourth image unit and the optimized image unit corresponding to a single first region image set is judged respectively, and the optimized image sub-unit of the optimized image unit with a difference degree less than a preset difference threshold is determined; the second region image set is determined by image fusion enhancement based on the optimized image sub-unit, the second image unit, the fourth image unit and the target simulation image information, and multiple second region images constitute the target endoscope image set.

[0070] In practical implementation, the remaining third image units are optimized using the calibration image information and target simulation image information of the second image unit with sufficient clarity and accuracy. Specifically, image features of the calibration image information, such as contrast, sharpness, and noise level, can be used as a benchmark to adaptively adjust the entire image or areas with uneven quality, so that the overall quality of the final output image tends to be consistent and optimal. Furthermore, when the third image unit has insufficient resolution, the corresponding known structural information in the target simulation image information can be used as prior knowledge to reconstruct the image system, making it clearer. In addition, for areas obscured by bubbles or dirt, the obscured parts can be intelligently inferred and repaired based on the structural information provided by the target simulation image information. After optimizing the third image unit of each image separately, the similarity of the images of the same area at different time points is compared and verified to obtain more accurate optimized image sub-units. Then, the optimized image sub-units, the target simulation image, and the second image unit are fused. Although it is impossible to make each endoscope image clear and accurate, the image fusion can obtain a clear and accurate endoscope image of the target area. The target endoscope image set is composed of multiple endoscope images of the same area displayed from multiple different angles.

[0071] Furthermore, the step of determining the second region image set by image fusion enhancement based on the optimized image sub-unit, the second image unit, and the target simulation image information includes:

[0072] When performing image fusion enhancement based on the optimized image subunit, the second image unit, the fourth unit image, and the target simulation image information, the change in physiological feature data at corresponding time points is obtained. This change includes at least respiratory rate fluctuations and heart rate fluctuations. It is then determined whether the change exceeds a preset change threshold. If so, the weight parameters of the optimized image subunit, the second image unit, the fourth unit image, and the target simulation image information during the fusion enhancement process are adjusted according to the ratio of the change to the preset change threshold. The larger the change, the smaller the weight percentage of the optimized image subunit. Furthermore, in specific implementations, it is also necessary to adjust the proportion of different dimensions during the image fusion enhancement process based on the change in physiological feature data. This is because, under conditions of rapid breathing, the quality, accuracy, and clarity of images acquired in real-time by the endoscope are relatively lower, thus requiring reasonable weight adjustment to ensure the accuracy of image fusion.

[0073] In summary, the endoscopic image analysis method in the above embodiments of the present invention constructs a human body simulation model of the patient using image information obtained from external probing before endoscopic image acquisition. This allows for accurate acquisition of relatively macroscopic structural images of the patient's body. Then, based on physiological characteristic data during endoscopic examination, simulation calculations are performed on the human body simulation model to determine the patient's internal activity and dynamic changes during endoscopic examination. Furthermore, based on the movement trajectory, posture, and time points during endoscopic examination, images from the simulation are acquired to determine the corresponding simulation images for each endoscopic image. Subsequently, image information with consistent features and high clarity in the endoscopic images can be calibrated based on the simulation images. This calibration information is then used to optimize other endoscopic image information. Therefore, image optimization is based on the individual differences of the patient and the dynamic changing environment, effectively ensuring the accuracy and individual adaptability of image optimization. Thus, the present invention solves the problem of the lack of a highly accurate and individually adaptable endoscopic image analysis method and system in the prior art.

[0074] Example 2

[0075] Please see Figure 2 The diagram shown is a structural block diagram of the endoscopic image analysis system proposed in the second embodiment of the present invention. The endoscopic image analysis system 200 includes: a data acquisition module 21, a simulation module 22, an image information determination module 23, and an image processing module 24, wherein:

[0076] The data acquisition module 21 is used to acquire historical image information and construct a human body simulation digital model of the patient based on the historical image information. The historical image information includes at least the image data acquired by external detection measures before endoscopic examination.

[0077] Simulation module 22 is used to acquire endoscopic detection information, which includes at least endoscopic image information, endoscopic trajectory and posture information, and physiological characteristic data of the patient during endoscopic detection. Simulation calculation is performed based on the physiological characteristic data and the human body simulation model through a preset simulation model.

[0078] The image information determination module 23 is used to obtain the target simulation image information during the simulation operation based on the endoscope trajectory and posture information, and to analyze and determine the calibration image information based on the target simulation image information. The calibration image information is the endoscope image information that has the same features as the target simulation image information and has a clarity greater than a preset clarity threshold.

[0079] The image processing module 24 is used to optimize the endoscopic image information based on the calibration image information and the target simulation image information to determine the target endoscopic image set, and to analyze the endoscopic image set to determine the patient's condition.

[0080] Example 3

[0081] In another aspect, the present invention also proposes an electronic device, please refer to [link to relevant documentation]. Figure 3 The diagram shows an electronic device according to the third embodiment of the present invention, including a memory 20, a processor 10, and a computer program 30 stored in the memory and executable on the processor. When the processor 10 executes the computer program 30, it implements the endoscopic image analysis method as described above.

[0082] In some embodiments, the processor 10 may be a central processing unit (CPU), controller, microcontroller, microprocessor or other data processing chip, used to run program code stored in memory 20 or process data, such as executing access restriction programs.

[0083] The memory 20 includes at least one type of readable storage medium, such as flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 20 can be an internal storage unit of an electronic device, such as the hard disk of the electronic device. In other embodiments, the memory 20 can also be an external storage device of the electronic device, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc. Furthermore, the memory 20 can include both internal and external storage units of the electronic device. The memory 20 can be used not only to store application software and various types of data of the electronic device, but also to temporarily store data that has been output or will be output.

[0084] It should be pointed out that, Figure 3 The structure shown does not constitute a limitation on the electronic device. In other embodiments, the electronic device may include fewer or more components than shown, or combine certain components, or have different component arrangements.

[0085] This invention also proposes a computer-readable storage medium storing a computer program that, when executed by a processor, implements the endoscopic image analysis method described above.

[0086] Those skilled in the art will understand that the logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can mean any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0087] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0088] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0089] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0090] The above embodiments merely illustrate several implementation methods of the present invention, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this patent should be determined by the appended claims.

Claims

1. An endoscopic image analysis method, characterized in that, The method includes: Acquire historical image information and construct a human body simulation digital model of the patient based on the historical image information. The historical image information includes at least the image data collected by external detection measures on the patient before endoscopic examination. Acquire endoscopic detection information, which includes at least endoscopic image information, endoscopic trajectory and posture information, and physiological characteristic data of the patient during endoscopic detection. Simulation calculations are performed based on the physiological characteristic data and the human body simulation model using a preset simulation model. The target simulation image information is obtained during the simulation operation based on the endoscope trajectory and posture information, and the endoscope image information is analyzed based on the target simulation image information to determine the calibration image information. The calibration image information is the endoscope image information that has the same features as the target simulation image information and has a clarity greater than a preset clarity threshold. The endoscopic image information is optimized based on the calibration image information and the target simulation image information to determine the target endoscopic image set, and the endoscopic image set is analyzed to determine the patient's condition.

2. The endoscopic image analysis method according to claim 1, characterized in that, The steps for obtaining the target simulation image information during simulation calculation based on the endoscope trajectory and posture information include: Based on the endoscope trajectory and posture information, determine the node position and node shooting posture of the endoscope at different time points in the simulation operation; Based on the node position, the real-time simulation model within a preset time range at the corresponding time node is frozen and the target area simulation model is extracted. The position of the light projection is determined based on the position of the node, and the projection angle is determined based on the shooting posture of the node and the endoscope parameters. Illumination projection is performed within the target region simulation model based on the illumination projection position and the projection angle. The overlapping area between the illumination projection and the target region simulation model is determined, and the target simulation image information is determined based on the overlapping area and the projection angle.

3. The endoscopic image analysis method according to claim 1, characterized in that, The steps for analyzing the endoscopic image information and determining the calibration image information based on the target simulation image information include: Feature extraction is performed on each image within the target simulation image information to determine the first feature information; The first feature information is matched with the preset region feature information to determine the portion of the first feature information whose similarity is greater than the first preset similarity threshold as the second feature information; Based on the second feature information, the target simulation image information and the endoscope image information are filtered to determine the target simulation image and endoscope image corresponding to the second feature information, so as to determine the calibration image information according to the target simulation image and the endoscope image.

4. The endoscopic image analysis method according to claim 3, characterized in that, The step of determining the calibration image information based on the target simulation image and the endoscope image includes: The target endoscopic image is divided into segments, and the target endoscopic images with a clarity greater than the preset clarity threshold are identified as the first image units. Feature extraction is performed on the first image unit to determine the third feature information, and the second feature information and the third feature information are matched for similarity to determine the part of the third feature information with a similarity greater than a preset similarity threshold as the fourth feature information; The first image unit is divided according to the fourth feature information to determine the corresponding second image unit, and the information of multiple second image units constitutes the calibration image information.

5. The endoscopic image analysis method according to claim 4, characterized in that, The steps of extracting features from the first image unit to determine the third feature information and performing similarity matching between the second feature information and the third feature information to determine the portion of the third feature information with a similarity greater than a preset similarity threshold as the fourth feature information are followed by: The first image unit is divided according to the fourth feature information to determine the second image unit and the third image unit corresponding to the fourth feature information; Determine whether there is a fifth feature information in the feature information of the third image unit that matches the features of the preset pathological feature library; If they exist, the third image unit is divided according to the fifth feature information to determine the fourth image unit and the fifth image unit corresponding to the fifth feature information, and multiple second image units and the fourth image unit constitute the calibration image information.

6. The endoscopic image analysis method according to claim 5, characterized in that, The target endoscope image is composed of a second image unit, a fourth image unit, and a fifth image unit. The step of optimizing the endoscope image information to determine the target endoscope image set based on the calibration image information and the target simulation image information includes: The fifth image unit is optimized and the optimized image unit is determined based on the calibration image information and the target simulation image information. Based on the preset region feature information and the second feature information, the target endoscope image is classified to determine multiple different first region image sets; The difference degree is judged for the second image unit, the fourth image unit and the optimized image unit corresponding to a single first region image set, and the optimized image sub-unit of the optimized image unit with a difference degree less than a preset difference threshold is determined. Based on the optimized image subunit, the second image unit, the fourth image unit, and the target simulation image information, image fusion enhancement is performed to determine the second region image set, and multiple second region images constitute the target endoscope image set.

7. The endoscopic image analysis method according to claim 6, characterized in that, The steps for determining the second region image set by image fusion enhancement based on the optimized image subunit, the second image unit, and the target simulation image information include: When performing image fusion enhancement based on the optimized image subunit, the second image unit, the fourth image unit, and the target simulation image information, the change in physiological feature data at the corresponding time node is obtained. The change includes at least respiratory rate fluctuation value and heart rate fluctuation value, and it is determined whether the change is greater than a preset change threshold. If so, the weight parameters of the optimized image subunit, the second image unit, the fourth image unit, and the target simulated image information during the fusion enhancement process are adjusted according to the ratio of the change amount to the preset change threshold, wherein the larger the change amount, the smaller the weight ratio of the optimized image subunit.

8. An endoscopic image analysis system, characterized in that, The system for implementing the endoscopic image analysis method as described in any one of claims 1 to 7 comprises: The data acquisition module is used to acquire historical image information and construct a human body simulation digital model of the patient based on the historical image information. The historical image information includes at least the image data acquired by external detection measures on the patient before endoscopic examination. The simulation module is used to acquire endoscopic detection information, which includes at least endoscopic image information, endoscopic trajectory and posture information, and physiological characteristic data of the patient during endoscopic detection. Based on the physiological characteristic data and the human body simulation model, simulation calculations are performed through a preset simulation model. The image information determination module is used to obtain target simulation image information during simulation calculation based on the endoscope trajectory and posture information, and to analyze and determine calibration image information based on the target simulation image information. The calibration image information is the endoscope image information that has the same features as the target simulation image information and has a clarity greater than a preset clarity threshold. The image processing module is used to optimize the endoscopic image information based on the calibration image information and the target simulation image information to determine the target endoscopic image set, and to analyze the endoscopic image set to determine the patient's condition.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the endoscopic image analysis method as described in any one of claims 1 to 7.

10. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the endoscopic image analysis method as described in any one of claims 1-7.

Images