Internal secretion related anorectal traditional Chinese medicine pharmacological image sequence analysis method

Abstract

The invention relates to the technical field of image processing, in particular to an endocrine-related anorectal traditional Chinese medicine pharmacological image sequence analysis method, which comprises the following steps: acquiring continuous frame images, extracting annular boundaries, comparing contour trends to identify repeated areas, tracking a one-way evolution image layer to remove redundant frames, and positioning a traditional Chinese medicine path tracking and pushing area. And screening a rhythm image according to a hormone cycle, screening a rhythm section image, tracking a boundary drift analysis extension direction, screening a continuous path region, and obtaining a structure disturbance graph. According to the method, redundant regions are eliminated through contour trend comparison, the recognition continuity of structural changes between image frames is enhanced, a continuous response region is extracted through boundary consistency in the path direction, the time advancing trend of tissue morphology is analyzed in combination with contrast of a hormone cycle, and a structural disturbance trajectory is constructed according to the boundary advancing direction. And a response path is extracted by combining the offset distribution of the central axis structure in the sequence, so that the tissue change in the image has directivity, rhythmicity and sequencing.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image processing, in particular to an endocrine-related anorectal traditional Chinese medicine pharmacological image sequence analysis method. BACKGROUND

[0002] Image processing technology involves the acquisition, analysis, understanding and conversion of original image data, and the core includes image acquisition, preprocessing, feature extraction, classification recognition and result output, aiming to convert image information into structured data that can be used for diagnosis, recognition or decision making, which is commonly used in the medical field to assist in diagnosis, pathology recognition and treatment evaluation, especially in medical image sequence analysis, which requires continuous tracking and structural recognition of image changes in the time dimension to reveal the dynamic characteristics of tissue or pharmacological response. Among them, the traditional endocrine-related anorectal traditional Chinese medicine pharmacological image sequence analysis method is to record and classify the changes of tissues under the morphological changes of anorectal tissues in time sequence images under the regulation of endocrine and the intervention of traditional Chinese medicine, by means of frame extraction, tissue region demarcation and response recognition. This method relies on manual comparison and static image feature identification, and combines low-level features such as brightness distribution, color change and edge shape to visually identify key frames.

[0003] The prior art takes low-level image features as the basis for identification, lacks overall perception of image structure evolution, and the morphological changes between image frames do not form a directional analysis path. The response area often produces identification ambiguity due to the overlapping of frame contours, lacks path coherence support in the identification of traditional Chinese medicine action area, lacks a unified tracking mechanism for the changes of tissue boundaries, and the morphological transfer relationship before and after intervention is not systematically revealed. The spatial displacement of the layer structure cannot be sequentially demarcated in the image sequence, and the tissue response process therefore presents a state of dispersion, strong discontinuity and unclear direction, resulting in difficulty in constructing a pharmacological change trajectory with logical clues from the image sequence. SUMMARY

[0004] In order to solve the technical problems existing in the prior art, the endocrine-related anorectal traditional Chinese medicine pharmacological image sequence analysis method provided by the embodiments of the present application comprises the following steps:

[0005] In order to achieve the above-mentioned purpose, the present application adopts the following technical scheme: an endocrine-related anorectal traditional Chinese medicine pharmacological image sequence analysis method comprises the following steps:

[0006] S1: acquiring continuous frame images in an anorectal image sequence, extracting the boundary of the annular tissue in the intestinal tract, comparing the contour trend of adjacent frames, analyzing whether there is repeated backfolding and overlapping area, continuously tracking the unidirectional evolution layer, and removing the repeated response frame to obtain an image frame set with repeated morphological areas removed;

[0007] S2: Based on the position of the frame set of the image frame set of the removed morphological repeated region, the tissue region through which the traditional Chinese medicine path passes is located, the mucosa boundary line of the adjacent region is extracted, the change direction and path consistency are analyzed, the continuously moving region is tracked, and a traditional Chinese medicine intervention response structure region image is obtained;

[0008] S3: Based on the traditional Chinese medicine intervention response structure region image time period, the hormone release cycle is compared, the morphological change continuity and direction trend of the frame before and after the reaction region intervention are compared, the synchronous promotion relationship between the structure and the hormone rhythm is analyzed, the continuously covered rhythm section image is screened, and a response region image segment is obtained.

[0009] S4: Based on the response region image segment, the tissue boundary drift direction is tracked, whether the boundary continuously moves along the traditional Chinese medicine direction is analyzed, whether the boundary extension of the image layer before and after is consistent is judged, the direction consistent and continuously extended region is screened, and a structure disturbance path pattern is obtained.

[0010] As a further scheme of the present application, the image frame set of the removed morphological repeated region includes a ring-shaped tissue boundary trend graph, a repeated backfolding and overlapping region graph, a one-way evolution layer region graph, and a repeated response frame graph, the traditional Chinese medicine intervention response structure region image includes a traditional Chinese medicine action path graph, a mucosa boundary line graph, a path consistency analysis graph, and a morphological moving region graph, the response region image segment includes a continuous morphological change graph before and after intervention, a time sequence movement trend graph, and a hormone rhythm section coverage graph, and the structure disturbance path pattern includes a layer boundary drift graph, a boundary extension consistency graph, and a path direction extension graph.

[0011] As a further scheme of the present application, the one-way evolution layer refers to, in image sequence tracking, obtaining the position distribution of each frame of mucosa structure line, analyzing the spatial trend in the frame sequence direction, comparing the extension direction of the boundary line in the adjacent frames, and judging whether it continuously offsets to the same direction.

[0012] The tissue region refers to selecting a coverage range around the image center axis on both sides in the image frame through which the traditional Chinese medicine transmission path passes, and screening a structure region with continuous boundary signal response and complete contour.

[0013] As a further scheme of the present application, the region mucosa boundary line refers to identifying the edge with a prominent gray scale change in the tissue region, locating the mucosa contour in each frame of image, extracting the position coordinates frame by frame, continuously comparing the boundary trend and displacement, and analyzing the motion trend under the time axis.

[0014] The rhythm section image refers to aligning the image frame sequence with the hormone release cycle based on the hormone rhythm data, screening the image segment with synchronized boundary response and rhythm change, and judging whether the image falls into the rhythm cycle response section.

[0015] As a further scheme of the present application, the specific steps of S1 are:

[0016] S101: Obtain continuous frames in the anorectal image sequence, collect the contour coordinate point set of the intestinal ring tissue boundary in each frame, continuously track the boundary point position in adjacent frames, determine whether the contour structure repeats the same trend on the time axis, and number the frame sequence of the repeated structure in continuous frames to obtain the set of boundary contour repeated frame sequence numbers.

[0017] S102: Based on the set of repeated frame numbers of the boundary contour, filter the direction path of the boundary contour in the corresponding frame image, analyze whether the contour trend in the frame sequence continues in the same direction in the spatial position, remove the frame sequence of the region where the direction has not changed but the position overlaps, and obtain a set of image frames with non-overlapping directions.

[0018] S103: Based on the set of image frames with non-overlapping directions, extract the motion direction information of the boundary contours in consecutive frames in chronological order, filter the layers whose directions have not reversed and continue to advance in the frame sequence, extract the frames with continuous changes during the process, and obtain the set of image frames with repeated morphological regions removed.

[0019] As a further aspect of the present invention, the specific steps of S2 are as follows:

[0020] S201: Based on the position of the traditional Chinese medicine action frame in the image frame set after removing the morphological repetition region, extract the tissue region through which the traditional Chinese medicine path passes in the frame, analyze the mucosal contour line of the same region in adjacent frames, filter the continuously appearing boundary line segments along the path direction, and obtain a set of continuous boundary lines of the path region.

[0021] S202: Based on the morphological distribution of the mucosal contour in each frame of the continuous boundary line set of the path region, calculate the boundary morphological projection value, compare the direction of change of the projection value in adjacent frames, filter the boundary layer whose projection value continues to change along the path direction, and obtain the contour shift sequence.

[0022] S203: Based on the contour shift sequence, analyze the morphological shift region along the path direction in consecutive frames, extract the layer segments whose boundaries maintain the continuity of the path direction in the frame sequence, identify the regions in the layer whose morphological trends are coherent and consistent with the path direction, and obtain the image of the structural region of the Chinese medicine intervention response.

[0023] As a further aspect of the present invention, the specific steps of S3 are as follows:

[0024] S301: Based on the time period of the image of the structural region of the Chinese medicine intervention response, extract the boundary structure of the target region in the frames before and after the intervention, determine whether there is a break or direction deflection in the contour trend in the continuous frames, filter the region with continuous change in shape contour, and obtain the continuous change structure segment.

[0025] S302: Based on the frame number and hormone release cycle time point in the continuously changing structural segment, calculate the corresponding value between the frame number and hormone release time point, determine whether the regional continuity is synchronously arranged in the hormone rhythm, and obtain the segment sequence corresponding to the hormone rhythm.

[0026] S303: Based on the frame index in the segment sequence corresponding to the hormone rhythm, filter the image frames in the image sequence that respond continuously within the corresponding rhythm segment, analyze whether the boundary trend is continuously preserved in the layer, and obtain the response area image segment.

[0027] As a further aspect of the present invention, the specific steps of S4 are as follows:

[0028] S401: Based on the continuous image frames in the response region image segment, extract the tissue boundary line of each frame, track the position change of the boundary between adjacent frames in the layer sequence, analyze whether the boundary line shows directional drift under the time axis, and obtain the tissue boundary drift trajectory.

[0029] S402: Based on the tissue boundary drift trajectory, compare the transmission path of the Chinese medicine intervention area, locate the boundary line segment whose drift direction is consistent with the path direction, continuously track whether the direction continues in the frame sequence, and obtain the path direction extended boundary area.

[0030] S403: Based on the path direction, extend the boundary region, identify the extension trend of the boundary shape in the transmission direction in the previous and next frames, track the position changes of the boundary line in the layer sequence, analyze the continuous movement behavior of the shape along the path direction, and obtain the structural disturbance path graphic.

[0031] As a further aspect of the present invention, the method further includes:

[0032] S5: Based on the structural perturbation path graph, extract the position of the central axis structure on both sides of the path, compare the extension trend of the structural layer in the frame before and after hormone cycle release, identify whether the tissue layers continuously shift along the path direction, extract the sequential distribution segment of each frame's shift segment, and obtain the traditional Chinese medicine pharmacology linkage image response trajectory structure.

[0033] The image response trajectory structure linked to the pharmacology of traditional Chinese medicine includes a central axis structure position map, a structure extension direction map, a tissue layer continuous offset map, and a structure offset segment distribution map.

[0034] As a further aspect of the present invention, the specific steps of S5 are as follows:

[0035] S501: Based on the structural perturbation path graph, extract continuous image frames of the central axial tissue region on both sides of the path in the intestinal image sequence, track the extension direction of the structural line in the layer in adjacent frames, determine whether it continues to change along the path direction, and obtain the extension direction of the central axial structure.

[0036] S502: Based on the extension direction of the central axis structure, compare the positional changes of the corresponding structural regions in the frames before and after hormone release, analyze whether the consistent offset trend in the path direction is maintained in the continuous image frames, identify the tissue layer that is continuously shifting, and obtain the continuously offset tissue region within the path.

[0037] S503: Based on the continuous offset tissue regions within the path, follow the start and end positions of the offset of each frame structure in the image sequence, determine whether the region changes move forward according to the time sequence, track the continuous transfer positions of the response region in the sequence, and obtain the image response trajectory structure of traditional Chinese medicine pharmacology linkage.

[0038] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0039] In this invention, redundant regions are eliminated by comparing contour trends, enhancing the recognition continuity of structural changes between image frames. Continuous response regions are extracted by the boundary consistency along the path direction. The temporal progression trend of tissue morphology is analyzed in conjunction with the hormone cycle. Structural perturbation trajectories are constructed based on the direction of boundary shift. Response paths are extracted by combining the offset distribution of the central axis structure in the sequence, enabling tissue changes in the image to possess directionality, rhythm, and sequence. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 This is a schematic diagram of the steps of the present invention;

[0042] Figure 2 This is a detailed schematic diagram of S1 of the present invention;

[0043] Figure 3 This is a detailed schematic diagram of S2 of the present invention;

[0044] Figure 4 This is a detailed schematic diagram of S3 of the present invention;

[0045] Figure 5 This is a detailed schematic diagram of S4 of the present invention;

[0046] Figure 6 This is a detailed schematic diagram of S5 of the present invention. Detailed Implementation

[0047] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0048] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0049] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, their intended meanings are consistent. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, their intended meanings are consistent.

[0050] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0051] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0052] Please see Figure 1 This invention provides a method for analyzing pharmacological image sequences of endocrine-related anorectal traditional Chinese medicines, including the following steps:

[0053] S1: Obtain continuous frames from the anorectal image sequence, extract the boundary of the intestinal ring tissue, compare the contour trend of the boundary in adjacent frames, analyze whether there are repeated folds and overlapping areas, continuously track the layer areas in the image that show unidirectional evolution, remove duplicate response frames, and obtain an image frame set with morphological repetition areas removed.

[0054] S2: Based on the position of the Chinese medicine action frame in the image frame set after removing the morphological repetition region, locate the tissue region through which the Chinese medicine path passes, extract the mucosal boundary line in the adjacent region, analyze the consistency of the change direction and path in the continuous frame, track the morphological shift region where the contour continuously appears in the path direction, and obtain the image of the structural region of the Chinese medicine intervention response.

[0055] S3: Based on the time period of the image of the structural region of the intervention of traditional Chinese medicine, compare the hormone release cycle, and compare whether the morphological changes of the response region are continuous in the frames before and after the intervention. Analyze whether the structure moves in the same direction along the time sequence and advances synchronously with the hormone period. Select the image segment in the frame where the morphological changes continuously cover the hormone rhythm to obtain the response region image segment.

[0056] S4: Based on continuous image frames in the response region image segment, track the drift direction of the tissue boundary in the layer sequence, analyze whether the boundary morphology continues to move along the direction of the Chinese medicine action, determine whether the extension behavior of the boundary between the front and rear segments in the layer is consistent, filter the regions where the drift direction is consistent with the transmission path and the boundary extension is continuous, and obtain the structural disturbance path graphic.

[0057] S5: Based on the structural perturbation path graph, extract the position of the central structure on both sides of the path in the intestinal image sequence, compare the extension direction and position change trend of the structure in the frame before and after hormone cycle release, identify whether the tissue layers in the sequence show a continuous offset behavior along the path direction, extract the sequential distribution segment of the tissue structure offset segment in each frame under the time axis, and obtain the image response trajectory structure of traditional Chinese medicine pharmacology linkage.

[0058] The image frame set for removing morphological repetition regions includes a circular tissue boundary trend map, a map of repeated folds and overlapping areas, a map of unidirectional evolution layer areas, and a repetitive response removal frame map. The images of the structural regions responding to traditional Chinese medicine intervention include a map of the path of action of traditional Chinese medicine, a map of the mucosal boundary line, a map of path consistency analysis, and a map of morphological shift areas. The image segments of the response regions include a map of continuous morphological changes before and after intervention, a map of temporal sequential movement trends, and a map of hormone rhythm segment coverage. The structural disturbance path graphics include a map of layer boundary drift, a map of boundary extension consistency, and a map of path direction extension. The image response trajectory structure of traditional Chinese medicine pharmacological linkage includes a map of the central axis structure position, a map of structural extension direction, a map of continuous interlayer offset, and a map of structural offset segment distribution.

[0059] Please see Figure 2 The specific steps of S1 are as follows:

[0060] S101: Obtain continuous frames in the anorectal image sequence, collect the contour coordinate point set of the intestinal ring tissue boundary in each frame, continuously track the boundary point position in adjacent frames, determine whether the contour structure repeats the same trend on the time axis, and number the frame sequence of the repeated structure in continuous frames to obtain the set of boundary contour repeated frame sequence numbers.

[0061] First, using the circular intestinal tissue in each frame as the target region, the coordinates of the circular boundary points with dense edge pixels are sequentially collected in the image. The boundary points extracted from each frame are grouped into coordinate sets, forming the corresponding boundary point set in the frame sequence. Then, matching processing is performed based on the coordinate sets between adjacent frames. For each boundary point in an adjacent frame, spatially close coordinate points are identified in the next frame and matched accordingly. During image parsing, a spatial proximity value of 3.5 or less can be set, and point pairs within this range are considered nearest neighbors. When the number of consecutive nearest neighbor point pairs identified in each pair of frames reaches more than 70% of the total number of points, the frame is considered a true pair. The boundary trends in the images are correlated, and their corresponding frame numbers are then categorized into the boundary corresponding sequence. After obtaining the matching results of all image frames, the search continues to the next interval frame to detect whether there is a boundary contour distribution in a non-adjacent frame that is similar to the previous frame. If multiple frames show similar trends and the positional differences are within a set range, such as the angle difference between the boundary line segments of two frames being less than 15 degrees, these frame numbers are recorded in the same repeated contour frame group. Finally, in a set of image sequences, such as frames 5, 9, 14, and 18, which are identified as consecutively appearing contour structure frames, their numbers are set and labeled to obtain the set of boundary contour repeated frame sequence numbers.

[0062] S102: Based on the set of repeated frame numbers of boundary contours, filter the direction path of the boundary contours in the corresponding frame images, analyze whether the contour trend in the frame sequence continues in the same direction in space, remove the frame sequence of regions where the direction has not changed but the positions overlap, and obtain a set of image frames with non-overlapping directions.

[0063] First, the contour line composed of intestinal edge points is extracted from each numbered frame. Then, the contour direction vector of each frame is extracted in frame order. The boundary direction path is obtained based on the boundary point coordinate transformation between the previous and next frames, constructing a direction sequence composed of the direction vectors of each frame. Next, the boundary direction differences between adjacent frames are compared sequentially. If the angle between adjacent frame vectors is within 10 degrees, their contour continuation direction is determined to be the same. Then, the spatial position of the boundary point center coordinates is compared by horizontal and vertical axis displacement. If the position change is within 2 pixels... Frames marked as overlapping in the frame pair are identified as such. In the example image, the contour lines of frames 8 and 9 have the same direction and the center point displacements are 1.2 pixels and 0.9 pixels respectively, which are considered to be overlapping frames. The same process is continued for all frames in the sequence to identify all frames with continuous direction but spatial overlap. Finally, these frames are numbered and removed, and only frames with contour paths that do not spatially overlap are retained. For example, frames 3, 6, 10, 15, and 19 are retained. They are sorted by frame index and output as a set to obtain a set of image frames with non-overlapping directions.

[0064] S103: Based on the set of image frames with non-overlapping directions, extract the motion direction information of the boundary contours in consecutive frames in chronological order, filter the layers whose directions have not reversed and continue to advance in the frame sequence, extract the frames with continuous changes during the process, and obtain the set of image frames with repeated morphological regions removed.

[0065] First, the direction vectors between the start and end points of the contour lines in each frame are extracted by frame index and arranged in chronological order to form a direction sequence. Then, it is determined whether the contour direction between adjacent frames has reversed by more than 180 degrees. For example, if the contour direction vector in frame 11 is to the upper right and in frame 12 is to the right, a direction change of less than 45 degrees is considered as no reversal. Simultaneously, the positional changes of each point on the contour line are collected, and the trajectory is verified by whether the displacement of boundary points between consecutive frames shows a unidirectional advancement. If the point displacement direction is consistent and continuously increases along the same axis, it is determined that the advancement direction is consistent. For example, from frame 13 to frame 14, the average x-axis coordinate of the contour points increases from 125.8 to 132.6, advancing to the right without reversal. This process is repeated to select consecutive frames that meet the advancement condition. All layer data that meet this condition are then retained according to their frame index order. In the images from frame sequence 11 to 22, 8 frames without reversal and maintaining directional advancement are identified, and their sequence numbers are recorded as valid frame set indices, resulting in an image frame set with duplicate morphological regions removed.

[0066] Please see Figure 3 The specific steps of S2 are as follows:

[0067] S201: Based on the position of the Chinese medicine action frame in the image frame set after removing the morphological repetition region, extract the tissue region through which the Chinese medicine path passes in the frame, analyze the mucosal contour line of the same region in adjacent frames, filter the continuously appearing boundary line segments along the path direction, and obtain the set of continuous boundary lines of the path region.

[0068] First, the frame index number is obtained, extending forward and backward by four frames within the range above and below the current frame. The image frame numbers within this range are then determined as the operation frame sequence. Next, each frame in the operation frame sequence is partitioned, with approximately 40%–60% of the width of the frame's center defined as the herbal medicine pathway area. The mucosal contour line within this area is extracted as the analysis object. Subsequently, the coordinate point set of the mucosal contour line in each frame is numbered. By comparing the position and shape trends of the point sets in different frames within the same region, it is determined whether there is a continuous direction of change. Finally, the boundary is determined using the extreme value variation range of the contour points. Whether the line continues to advance in the path direction, for example, the lowest point on the right side of the mucosal boundary in frame 28 is y=74, and the corresponding point moves to y=78 in frame 29. The continuously increasing y value indicates that the boundary has a path-direction extension behavior. If a similar frame-by-frame advancement phenomenon is observed in all 5 frames, it can be determined that this outline segment is a continuous boundary segment. Then, a filtering operation is performed to retain only the boundary segments that appear in at least 4 frames, and their coordinate set numbers are recorded as the candidate boundary segment set. Finally, all boundary segments with continuous advancement performance are filtered out to obtain the continuous boundary line set of the path region.

[0069] S202: Based on the morphological distribution of the mucosal contour in each frame of the continuous boundary line set of the path region, calculate the boundary morphological projection value, compare the direction of change of the projection value in adjacent frames, filter the boundary layer whose projection value continues to change along the path direction, and obtain the contour migration sequence.

[0070] The specific formula for calculating the boundary shape projection value is as follows:

[0071]

[0072] Among them, P j L represents the boundary morphology projection value in the j-th frame image, m represents the number of structural points used in the calculation in the frame image, and L represents the boundary morphology projection value in the j-th frame image. i,j This represents the projection length of the i-th structure point in the j-th frame along the path direction. A represents the mean of the projected lengths of the m structural points in the j-th frame. i,j This represents the area weight value of the i-th structural point in the j-th frame;

[0073] Formula calculation logic: For each structural point i in path segment j, first calculate the actual boundary length L. i,j Average boundary length of the entire segment The difference between them is expressed using absolute values. Preserve the deviation in shape, eliminate the influence of positive and negative directions, and divide by Normalize the result to make it a dimensionless proportional change, and use the area A of the structure points. i,j With total area ratio This represents the area contribution of each structural point to the overall contour region. A normalized area weight term is constructed, and the two quantities mentioned above are multiplied to form the normalized morphological offset value × area weight value for each structural point i. This represents the response intensity of the structural point on path segment j. The response intensities of all structural points i = 1 to m on path segment j are accumulated to obtain the total response index P. j Essentially, it is an average morphological offset index weighted by area.

[0074] In the image sequence, path segment j includes 4 structural points (m=4);

[0075] Boundary length L i,j They are 18.2px, 21.5px, 20.3px, and 20.0px respectively;

[0076] average length pass It can be concluded that;

[0077] Area data A i,j The corresponding sizes are 3.2, 4.1, 4.8, and 3.5 pixels. 2 ;

[0078] The total area is ∑A k,j =15.6px 2 ;

[0079] Substitute into the calculation process:

[0080] Break down each item and calculate them separately:

[0081] Item 1: |18.2-20.0|=1.8, normalized value Area weight

[0082] The result is 0.09 × 0.20513 = 0.01846;

[0083] Item 2: |21.5-20.0|=1.5, normalized value Area weight

[0084] The result is 0.075 × 0.26282 = 0.01971;

[0085] Item 3: |20.3-20.0|=0.3, normalized value Area weight

[0086] The result is 0.015 × 0.30769 = 0.00462;

[0087] Item 4: |20.0-20.0|=0, the result is 0;

[0088] Finally, sum the results of each item:

[0089] P j =0.01846+0.01971+0.00462+0=0.04279;

[0090] Interpretation of Results and Numerical Significance: The results show that there is a certain degree of morphological shift in the structural points of path segment j. A weighted response to this shift using area size is used to derive the morphological response index P. j =0.04279. When the value is within the preset range (0.03, 0.07), it can be determined that the shape change in the path segment is within an acceptable fluctuation range.

[0091] Explanation of the innovative aspects of the formula:

[0092] The advantage of this formula lies in its ability to circumvent the limitations of relying on a single variable to guide the judgment by introducing a weighted multiplication mechanism of length deviation and local area. In practical image boundary fluctuation detection, it can stably reflect the response characteristics of local structure as it changes with frame. (Parameters) Controlling the degree of deviation in shape, The contribution ratio of the control area, after fusion, can enhance the index's response to high-area, high-deformation points.

[0093] S203: Based on the contour shift sequence, analyze the morphological shift region along the path direction in consecutive frames, extract the layer segments whose boundaries maintain the continuity of the path direction in the frame sequence, identify the regions in the layer whose morphological trends are consistent with the path direction, and obtain the image of the structural region of the intervention response of traditional Chinese medicine.

[0094] First, the acquired set of continuous boundary lines of the path region is numbered. Image frames with consecutive numbers in the frame sequence are selected as the analysis objects. Boundary line point sets are extracted one by one according to the frame sequence. It is confirmed whether the horizontal or vertical offset direction of the contour lines in each frame is consistent. The position difference of the boundary points relative to the previous frame is represented by a vector. Then, the directions of the vectors in adjacent frames are compared. If the vector angle remains within 30 degrees for three or more consecutive frames, it can be considered that the path direction is consistent. For example, the coordinates of the left endpoints of the boundary lines in frames 32 to 35 are (61, 120), (63, 122), (66, 125), and (70, 128) respectively. This type of continuous advancement pattern constitutes a morphological advancement. The region is moved, and then the region segments with continuous changes of more than 3 pixels in the boundary line movement trajectory are extracted. It is determined whether they exist in both the preceding and following frames and whether the position change is uninterrupted. If the trajectory continuously spans 4 frames and the boundary point set with changes of 4, 5, 6 and 6 pixels respectively, it can be confirmed that the layer has stable morphological continuity. Such layer segments are then used as candidate layer segments, and further filtering is performed to retain only layer segments with an angle of less than 45 degrees to the path direction. For example, if the path direction is upward and the layer extension direction is mainly changed by the y-axis, then the region with continuously decreasing y-axis is selected as the target segment. Finally, the layer content that meets the above conditions is extracted to obtain the image of the TCM intervention response structure region.

[0095] Please see Figure 4 The specific steps of S3 are as follows:

[0096] S301: Based on the time period of the structural region image of the intervention response of traditional Chinese medicine, extract the boundary structure of the target region in the frames before and after the intervention, determine whether there is a break or direction deflection in the contour trend in the continuous frames, filter the region with continuous change in shape contour, and obtain the continuous change structure segment.

[0097] First, select the continuous layer content of the 5 frames before and 5 frames after the intervention in the image frame sequence. Extract the coordinate point set corresponding to the direction of the mucosal boundary in the target area frame by frame, and perform linear fitting on the boundary line to obtain the principal axis direction of the contour direction. For example, after extracting the contour of the central area in frames 18 to 28, record the boundary point set and convert it into a two-dimensional polyline to unify the projection direction. Then, determine whether there is a continuity break in the contour line within the frame sequence. Detect whether there is a pixel interruption in a specific segment by checking the connectivity of the point set. If the break length in a frame exceeds 4 pixels, it is considered that there is a frame segment with discontinuous direction. Then compare the contour line in... The extension trend of line segments in consecutive frames is analyzed to identify whether there is a directional change with a deflection angle greater than 45 degrees. For example, if the boundary line of frame 20 deviates from the direction of frame 19 by 50 degrees, it is judged as a directional deflection behavior. During the screening process, the above-mentioned frames with obvious breaks and deflections are excluded, and only the layer area with continuous contour extension within the frame sequence and a main axis direction difference of less than 20 degrees is retained. Combined with the condition that the total number of continuous frame segments is greater than 5 frames, the shape preservation condition is used. For example, the contour line direction within frames 21 to 27 is kept to be translated in a clockwise direction, and the direction angle change is less than 15 degrees. Finally, the image content corresponding to the segment is extracted to obtain the continuously changing structure segment.

[0098] S302: Based on the frame number and hormone release cycle time point in the continuously changing structural segment, calculate the corresponding value between the frame number and hormone release time point, determine whether the regional continuity is synchronously arranged in the hormone rhythm, and obtain the segment sequence corresponding to the hormone rhythm.

[0099] The specific formula for calculating the correspondence between frame number and hormone release time point is as follows:

[0100]

[0101] Among them, C ft This represents the correspondence between the frame number and the hormone release time point. μ represents the frame number of the i-th frame in a continuously changing structural segment. f represent The average value, σ f represent standard deviation Represents the time point of hormone release of the i-th hormone. The matched reference frame position, μ m Represents a matching frame The average value, σ m represent standard deviation The arithmetic mean of the frame numbers in a continuous segment is represented by n, which represents the number of frames involved in the calculation in a continuously changing structural segment.

[0102] Formula Logic: This formula quantifies the degree of synchronization between continuously changing structural segments in an image frame sequence and the hormonal cycle rhythm. The formula consists of three multiplied parts, averaged over all frame numbers, to form the overall synchronization deviation value C. ft ;

[0103] First item This represents the normalized result of the deviation of the i-th frame from the mean of the frame number of the consecutive segments, and is used to measure the relative dispersion of local frame changes.

[0104] Second item It represents the normalized offset of the hormone release matching frame position corresponding to the i-th frame relative to all hormone matching frames, and is used to measure the dispersion of hormone frame points in the image space.

[0105] Third item This is the offset root value of the current frame relative to the global center frame of the structural segment, representing the offset magnitude of the structural center position on the time axis, which can reflect the spatial dynamic weight of the frame;

[0106] Taking actual collected data as an example, the number of frames in the analysis segment is n=5;

[0107] The frame number of the continuous structural change segment is f (c) =102, 104, 107, 110, 112;

[0108] The corresponding hormone release frame sequence is f (m) =100, 103, 108, 109, 113;

[0109] The average value μ of consecutive frames can be obtained through calculation. f =107.0, standard deviation σ f =3.688;

[0110] The average value μ of hormone-matched frames m =106.6, standard deviation σ m =4.587; at the same time ;

[0111] Substituting the above data into the formula one by one, the calculation is as follows:

[0112] Item 1:

[0113] Item 2:

[0114] Item 3:

[0115] The product of the three terms is 1.356 × 1.438 × 2.236 = 4.365;

[0116] The product in the second frame is

[0117] The product in the third frame is Because its center frame has no offset;

[0118] The product in frame 4 is

[0119] The product in frame 5 is

[0120] Sum the terms and divide by the number of frames:

[0121]

[0122] Interpretation of results and numerical significance: The average synchronization offset between the current structural segment and the hormone rhythm frame is 2.087 frames. The synchronization matching threshold is set to ±3 frames. Based on the image frame rate of 30fps and the physiological delay tolerance interval of 0.1s, the structural segment is judged to be the hormone rhythm synchronization response segment.

[0123] Explanation of the innovative aspects of the formula:

[0124] The advantage of the formula is that by introducing a standardized ratio between continuous structural frames and hormone frames, and combining it with the structural center offset root value adjustment, it measures the temporal stability and rhythmic alignment of frame positions, and avoids mismatches caused by local structural fluctuations or periodic misalignments.

[0125] S303: Based on the frame index in the segment sequence corresponding to the hormone rhythm, filter the image frames in the image sequence that have continuous response in the corresponding rhythm segment, analyze whether the boundary trend is continuously preserved in the layer, and obtain the response area image segment.

[0126] First, the frame index range marked in the hormone release cycle table is used to extract the layer image of the corresponding frame number in the image sequence. The target tissue boundary contours in the layer images are arranged sequentially to construct an inter-frame contour sequence. Then, the connection method of the boundary structure between each layer and the contour trend are examined to see if there is a continuous extension behavior. In this process, the boundary contours extracted in each frame are marked in the form of a set of pixel coordinate points, and the projection direction between the starting point and the ending point of the boundary in consecutive frames is compared to see if there is a trend deviation. If the extension direction of the contour line between two consecutive frames is less than 30 degrees, the trend is considered to be continuous. If the angle exceeds this range, the trend is considered to be continuous. If a frame is excluded from the subsequent extended segment statistics, the number of frames that meet the directional consistency standard in the continuous layers within the hormone rhythm frame index interval is further counted, and it is checked whether it covers more than 60% of the layers in the rhythm segment. For example, when the rhythm frame segment index is frames 12 to 24, and the frames with consistent boundary trends in the layers are frames 13 to 22, with a coverage rate of 83%, it is considered that there is a stable and continuous boundary response in the rhythm segment, and thus it is confirmed that there is a continuous boundary change in the corresponding area of ​​this image segment. Finally, all image layers in frames 13 to 22 are extracted as continuous response frame segments to obtain the response area image segment.

[0127] Please see Figure 5 The specific steps of S4 are as follows:

[0128] S401: Based on continuous image frames in the response region image segment, extract the tissue boundary line of each frame, track the position change of the boundary between adjacent frames in the layer sequence, analyze whether the boundary line shows directional drift under the time axis, and obtain the tissue boundary drift trajectory.

[0129] First, the boundary lines of the target tissue are extracted from the locations of abrupt changes in the grayscale distribution of each frame image and represented as a sequence of coordinate points. Then, the coordinate sets of the boundary lines in each frame are arranged sequentially according to the frame index, constructing a sequence of boundary line positions within the frame order. The boundary lines of adjacent frames in this sequence are compared point-by-point, and the relative position changes of corresponding points on each pair of boundary lines are statistically analyzed. Finally, the displacement direction and magnitude of the boundary lines between consecutive frames are summarized. Based on this, it is determined whether the boundary forms a continuous drift trajectory along a certain main direction. For example, if the image frame order is f1 to f10, in an image with a layer resolution of 512×512 pixels, the starting point of the boundary line in frame f1 is (240, 260). If the coordinates of the corresponding points in each frame from f2 to f10 move along the positive Y-axis in steps of 1 to 3 pixels, it is determined that there is directional drift behavior in the boundary structure of the image frame. All the boundary point positions that move continuously along this direction are integrated into a path line that changes with the frame order. The construction of this path line needs to exclude boundary segments with discontinuous jumps of more than 10 pixels to ensure the continuity of the trajectory. At the same time, after the trajectory is extracted, the frame order start and end indexes and the total displacement of the boundary points are marked. For example, if the frame index is f1 to f10 and the starting point displacement is 24 pixels, it is recorded as "f1-f10: displacement 24". Finally, the tissue boundary drift trajectory is obtained based on the displacement change of the tissue boundary line under the continuous frame order.

[0130] S402: Based on the tissue boundary drift trajectory, compare the transmission path of the Chinese medicine intervention area, locate the boundary line segment whose drift direction is consistent with the path direction, continuously track whether the direction continues in the frame sequence, and obtain the path direction extended boundary area.

[0131] First, the boundary line point set sequence in the drift trajectory is called to extract the relative displacement direction of the boundary line segments within the frame sequence range. The main direction vector corresponding to the transmission path is obtained from the traditional Chinese medicine intervention area. The two are compared directionally in terms of angle representation. By setting an angle offset judgment threshold, such as ±10°, the part of the boundary line segment whose direction difference is within the threshold range is identified as the path direction alignment segment. At the same time, the alignment segment is labeled with a frame index, and the frame sequence range corresponding to each segment is summarized. For example, if the difference between the boundary line and the traditional Chinese medicine path direction in frames f11 to f20 is only 6°, it is recorded as "segment f11-f20 is a consistent direction segment". Then, these consistent direction segments are further arranged sequentially in the frame sequence, and the boundary line is checked frame by frame to see if it continues to extend in the same direction in the next frame. In each continuous segment, the boundary endpoint position of the first and last frames is tracked. By calculating the trend of its X or Y coordinate change in the frame, it is confirmed whether its direction continues. If the coordinate change maintains a unidirectional trend, it is considered that the direction continues. Finally, all boundary line segments that meet the consistent direction and extend continuously in the frame sequence are summarized and output to obtain the path direction extension boundary area.

[0132] S403: Based on the path direction extension boundary region, identify the extension trend of the boundary shape in the transmission direction in the previous and next frames, track the position change of the boundary line in the layer sequence, analyze the continuous movement behavior of the shape along the path direction, and obtain the structural disturbance path graphic.

[0133] First, the layer frame sequence covered by the region is extracted, and the contour coordinate point set of the target boundary line segment is extracted in each frame. The start and end directions of the boundary are identified by the coordinate arrangement order. Then, the difference between the horizontal and vertical coordinates of the boundary coordinate points in two consecutive frames is used to determine whether the boundary line segment is continuously extending along the direction vector defined by the transmission path. For example, if the path direction is from left to right in the layer coordinates, the horizontal coordinate of the boundary point in consecutive frames should increase frame by frame. When the horizontal coordinate increases from 132 to 165 in frames f10 to f16, it can be considered that the boundary line segment is in an extended state. Then, taking the starting frame of the extended segment as the reference, the corresponding positions of the boundary in each frame are extracted sequentially. The set of coordinate points is used to construct a boundary trajectory change map according to the frame sequence, and the displacement direction and step size range of the boundary in the layer position in each frame are marked. For example, in frames f11 to f15, the displacement distance in each frame is maintained between 5 and 8 pixels, and the direction is stable and consistent. Then, a comparison and screening is performed in the overall layer sequence to identify those boundary segments that extend continuously in the frame sequence, whose displacement direction is completely consistent with the path direction, and whose step size is not zero. By setting a frame number threshold, such as no less than 5 frames, scattered fragments are eliminated, and only the layer change segments that meet the requirements of continuity and direction are retained. The region contour obtained by the above screening process is mapped to the layer coordinate matrix to obtain the structural disturbance path graphic.

[0134] Please see Figure 6 The specific steps of S5 are as follows:

[0135] S501: Based on the structural perturbation path graph, extract continuous image frames of the central axial tissue region on both sides of the path in the intestinal image sequence, track the extension direction of the structural line in the layer in the adjacent frame, determine whether it continues to change along the path direction, and obtain the extension direction of the central axial structure.

[0136] First, the coordinates of the left and right sides of the layer coordinate system are extracted. Using the path centerline as the reference axis, buffer zones of 5 to 8 pixels wide are defined on both sides of the path as the central axis organization range. In each frame, the image is cropped according to the region mask, retaining the central axis organization region and extracting its boundary line segments. In the image frame sequence, the contour lines of the boundary line segments in adjacent frames are paired. By comparing the start and end coordinates between the boundary point sets, it is determined whether the boundary extension direction is offset along the horizontal or vertical axis. If the offset direction is consistent with the propagation direction defined by the path direction, the boundary contour coordinates corresponding to the central axis region in the current frame are further added to the boundary direction list. At the same time, the start and end positions of the boundary line segments in each frame in the layer matrix are recorded. In the example layer, when the path direction is defined as self... From left to right, if the average horizontal coordinate of the right central axis segment in frames f32 and f33 moves from 112 to 118, then the frame interval is recorded as an extended segment. Subsequently, contour direction comparison is performed in all frame pairs to define the extension direction vector angle change value of the boundary contour between each pair of frames. In consecutive frame pairs, when the angle change is less than 15° and the extension distance is greater than 5 pixels, it is considered to be an extension with consistent direction. Then, by accumulating each frame group sequence that meets the conditions, a continuous set of frame segments is selected. At the same time, the extension direction of the left and right central axis regions is compared synchronously to exclude frame segments with one-sided deflection or stagnation. Only frames with synchronous advancement of the boundary directions on both sides are retained, and a continuous extension image segment of the central axis region boundary is formed in the layer from the first frame to the last frame, finally obtaining the extension direction of the central axis structure.

[0137] S502: Based on the extension direction of the central axis structure, compare the positional changes of the corresponding structural regions in the frames before and after hormone release, analyze whether the consistent offset trend in the path direction is maintained in the continuous image frames, identify the tissue layers that show continuous shifting, and obtain the continuously offset tissue regions within the path.

[0138] First, locate the corresponding frames before and after the traditional Chinese medicine intervention in the image sequence. Delineate the structural region covered by the extended path in each frame. Extract the boundary line segments of this region in each frame using layer coordinate indexing. Extract the structural boundary contour point set frame by frame in the image segment with frame index f20 to f35 and store its start and end boundary coordinate range. Use the outermost boundary point pairs in the structural point set to determine the horizontal and vertical bounding boxes. Then, obtain the position center coordinates of the structural bounding boxes in the frames before and after the hormone intervention node. For example, if the center coordinates of the structural region are (142, 98) in frame f25 and change to (149, 100) in frame f26, the horizontal offset is 7 pixels and the vertical offset is 2 pixels. Record this offset trend in the structural movement record table. Continue to extract the position center of the corresponding structure in the same way in subsequent frames. The coordinates are compared by the direction of change of the center coordinates between adjacent frames. If the direction is consistent with the extension direction of the central axis structure, the current frame is marked as a trend continuation frame, and a set of continuous trend frames is constructed. Then, the inter-frame offset distance is accumulated in all trend continuation frames. If the accumulated offset distance increases continuously and the direction does not reverse, the current frame segment is determined to be a continuous offset frame segment. Then, the layer in the frame segment is used to extract the boundary graphics of the structural region. The degree of difference between the continuous frames of the boundary contour point set is compared to determine whether the structural contour maintains continuity. When the change of the contour point position in the continuous frame is less than 10 pixels and the horizontal direction of the start and end boundaries is consistent, it is recorded as a structurally continuous frame. The final set of continuous shifting layers is filtered by the intersection of the structurally continuous frames and the trend continuation frames. The continuous offset organization region within the path is finally obtained in this set of layers.

[0139] S503: Based on the continuous offset tissue region within the path, follow the start and end positions of the offset structure of each frame in the image sequence, determine whether the region change moves forward according to the time sequence, track the continuous transfer position of the response region in the sequence, and obtain the image response trajectory structure of traditional Chinese medicine pharmacology linkage.

[0140] First, determine the start and end frames of the tissue region in the sequence frames. Define the path region boundary using the coordinate ranges on both sides of the central axis. Then, extract the start and end coordinates of the tissue region's boundary in each frame, forming a set of structural offset coordinate points ordered by frame index. Let the current frame be f20. In frame f20, the left boundary of the tissue region is located at x-coordinate 120, and the right boundary at x-coordinate 158. In frame f21, these are recorded as 125 and 163. At this point, the offset direction can be determined as an overall rightward shift. Subsequent frames f22, f23, etc., continuously extract the current frame's offset coordinates, recording the boundary change in each consecutive frame. If the offset difference between adjacent frames is positive, it indicates that the structural change has a consistent offset trend in the horizontal direction. Further, establish a one-to-one mapping table between frame indexes and boundary center point change coordinates. Verify this by reading the coordinate increment sequence of each frame in the mapping table. If the change in direction occurs, the recording is interrupted. If the change in direction is maintained in consecutive frames, it is determined to be a sequentially advancing frame segment. Then, the layers of all frames in the advancing frame segment are extracted. The pixel graphics of the tissue area within the path are extracted in each frame. The structural trend is matched by the frame sequence consistency check to determine whether there is abnormal deformation or missing edge of the boundary shape. If the contour trend is stable and there is no sudden change, it is recorded as a morphologically stable frame. The positions of the boundary center points of all morphologically stable frames on the layer are collected to form a spatial change trajectory. Then, the trajectory is compared with the direction line of the path area to determine whether the trajectory direction is consistent with the path extension direction. When the angle between the two is within 15 degrees and the cumulative continuous offset distance exceeds one-third of the width of 10 frames, the trajectory is considered to be the continuous transfer path of the response area in the layer. Finally, the response trajectory structure of the traditional Chinese medicine pharmacology linkage image is obtained.

[0141] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for analyzing pharmacological image sequences of endocrine-related anorectal traditional Chinese medicines, characterized in that, Includes the following steps: S1: Obtain continuous frame images from the anorectal image sequence, extract the boundary of the intestinal ring tissue, compare the contour trend of adjacent frames, analyze whether there are repeated folds and overlapping areas, continuously track the unidirectional evolution layer, remove duplicate response frames, and obtain an image frame set with morphological repetition areas removed. S2: Based on the position of the Chinese medicine action frame in the image frame after removing the morphological repetition region, locate the tissue area through which the Chinese medicine path passes, extract the mucosal boundary line of the adjacent area, analyze the consistency between the change direction and the path, track the continuously moving area, and obtain the image of the structural region of the Chinese medicine intervention response. S3: Based on the time period of the image of the response structure region of the Chinese medicine intervention, compare the morphological changes and directional trends of the frames before and after the intervention of the response region with the hormone release cycle, analyze the synchronous advancement relationship between the structure and the hormone rhythm, screen the image segments that continuously cover the rhythm, and obtain the response region image segments. S4: Based on the image segment of the response area, track the direction of tissue boundary drift, analyze whether the boundary continues to move along the direction of the Chinese medicine, determine whether the extension of the boundary of the front and back segments of the layer is consistent, filter the regions with consistent direction and continuous extension, and obtain the structural disturbance path graphic.

2. The method for analyzing pharmacological image sequences of endocrine-related anorectal traditional Chinese medicines according to claim 1, characterized in that, The image frame set for removing morphological repetition regions includes a circular tissue boundary trend map, a map of repeated folds and overlapping areas, a map of unidirectional evolution layer areas, and a repetitive response removal frame map. The images of the structural regions affected by traditional Chinese medicine intervention include a map of the path of action of traditional Chinese medicine, a map of mucosal boundary lines, a path consistency analysis map, and a map of morphological shift areas. The image segments of the response regions include a map of continuous morphological changes before and after intervention, a time-series movement trend map, and a hormone rhythm segment coverage map. The structural disturbance path graphics include a layer boundary drift map, a boundary extension consistency map, and a path direction extension map.

3. The method for analyzing pharmacological image sequences of endocrine-related anorectal traditional Chinese medicines according to claim 1, characterized in that, The unidirectional evolution layer refers to the process of obtaining the positional distribution of mucosal structure lines in each frame during image sequence tracking, analyzing the spatial orientation according to the frame sequence direction, comparing the extension orientation of the boundary lines in adjacent frames, and determining whether they continue to shift in the same direction. The tissue region refers to the area selected around the central axis of the image frame through which the traditional Chinese medicine is transported, and the structural region with continuous boundary signal response and complete outline is selected.

4. The method for analyzing pharmacological image sequences of endocrine-related anorectal traditional Chinese medicines according to claim 1, characterized in that, The mucosal boundary line refers to the edge with prominent gray-scale changes identified within the tissue area. The outer contour of the mucosa in each frame image is located, the position coordinates are extracted frame by frame, the boundary direction and displacement are continuously compared, and its movement trend on the time axis is analyzed. The rhythmic segment image refers to an image segment that is aligned with the hormone release cycle based on hormone rhythm data, and the boundary response and rhythm changes are synchronized to determine whether the image falls into the rhythmic cycle response segment.

5. The method for analyzing pharmacological image sequences of endocrine-related anorectal traditional Chinese medicines according to claim 1, characterized in that, The specific steps of S1 are as follows: S101: Obtain continuous frames in the anorectal image sequence, collect the contour coordinate point set of the intestinal ring tissue boundary in each frame, continuously track the boundary point position in adjacent frames, determine whether the contour structure repeats the same trend on the time axis, and number the frame sequence of the repeated structure in continuous frames to obtain the set of boundary contour repeated frame sequence numbers. S102: Based on the set of repeated frame numbers of the boundary contour, filter the direction path of the boundary contour in the corresponding frame image, analyze whether the contour trend in the frame sequence continues in the same direction in the spatial position, remove the frame sequence of the region where the direction has not changed but the position overlaps, and obtain a set of image frames with non-overlapping directions. S103: Based on the set of image frames with non-overlapping directions, extract the motion direction information of the boundary contours in consecutive frames in chronological order, filter the layers whose directions have not reversed and continue to advance in the frame sequence, extract the frames with continuous changes during the process, and obtain the set of image frames with repeated morphological regions removed.

6. The method for analyzing pharmacological image sequences of endocrine-related anorectal traditional Chinese medicines according to claim 1, characterized in that, The specific steps of S2 are as follows: S201: Based on the position of the traditional Chinese medicine action frame in the image frame set after removing the morphological repetition region, extract the tissue region through which the traditional Chinese medicine path passes in the frame, analyze the mucosal contour line of the same region in adjacent frames, filter the continuously appearing boundary line segments along the path direction, and obtain a set of continuous boundary lines of the path region. S202: Based on the morphological distribution of the mucosal contour in each frame of the continuous boundary line set of the path region, calculate the boundary morphological projection value, compare the direction of change of the projection value in adjacent frames, filter the boundary layer whose projection value continues to change along the path direction, and obtain the contour shift sequence. S203: Based on the contour shift sequence, analyze the morphological shift region along the path direction in consecutive frames, extract the layer segments whose boundaries maintain the continuity of the path direction in the frame sequence, identify the regions in the layer whose morphological trends are coherent and consistent with the path direction, and obtain the image of the structural region of the Chinese medicine intervention response.

7. The method for analyzing pharmacological image sequences of endocrine-related anorectal traditional Chinese medicines according to claim 1, characterized in that, The specific steps for S3 are as follows: S301: Based on the time period of the image of the structural region of the Chinese medicine intervention response, extract the boundary structure of the target region in the frames before and after the intervention, determine whether there is a break or direction deflection in the contour trend in the continuous frames, filter the region with continuous change in shape contour, and obtain the continuous change structure segment. S302: Based on the frame number and hormone release cycle time point in the continuously changing structural segment, calculate the corresponding value between the frame number and hormone release time point, determine whether the regional continuity is synchronously arranged in the hormone rhythm, and obtain the segment sequence corresponding to the hormone rhythm. S303: Based on the frame index in the segment sequence corresponding to the hormone rhythm, filter the image frames in the image sequence that respond continuously within the corresponding rhythm segment, analyze whether the boundary trend is continuously preserved in the layer, and obtain the response area image segment.

8. The method for analyzing pharmacological image sequences of endocrine-related anorectal traditional Chinese medicines according to claim 1, characterized in that, The specific steps of S4 are as follows: S401: Based on the continuous image frames in the response region image segment, extract the tissue boundary line of each frame, track the position change of the boundary between adjacent frames in the layer sequence, analyze whether the boundary line shows directional drift under the time axis, and obtain the tissue boundary drift trajectory. S402: Based on the tissue boundary drift trajectory, compare the transmission path of the Chinese medicine intervention area, locate the boundary line segment whose drift direction is consistent with the path direction, continuously track whether the direction continues in the frame sequence, and obtain the path direction extended boundary area. S403: Based on the path direction, extend the boundary region, identify the extension trend of the boundary shape in the transmission direction in the previous and next frames, track the position changes of the boundary line in the layer sequence, analyze the continuous movement behavior of the shape along the path direction, and obtain the structural disturbance path graphic.

9. The method for analyzing pharmacological image sequences of endocrine-related anorectal traditional Chinese medicines according to claim 1, characterized in that, The method further includes: S5: Based on the structural perturbation path graph, extract the position of the central axis structure on both sides of the path, compare the extension trend of the structural layer in the frame before and after hormone cycle release, identify whether the tissue layers continuously shift along the path direction, extract the sequential distribution segment of each frame's shift segment, and obtain the traditional Chinese medicine pharmacology linkage image response trajectory structure. The image response trajectory structure linked to the pharmacology of traditional Chinese medicine includes a central axis structure position map, a structure extension direction map, a tissue layer continuous offset map, and a structure offset segment distribution map.

10. The method for analyzing pharmacological image sequences of endocrine-related anorectal traditional Chinese medicines according to claim 1, characterized in that, The specific steps of S5 are as follows: S501: Based on the structural perturbation path graph, extract continuous image frames of the central axial tissue region on both sides of the path in the intestinal image sequence, track the extension direction of the structural line in the layer in adjacent frames, determine whether it continues to change along the path direction, and obtain the extension direction of the central axial structure. S502: Based on the extension direction of the central axis structure, compare the positional changes of the corresponding structural regions in the frames before and after hormone release, analyze whether the consistent offset trend in the path direction is maintained in the continuous image frames, identify the tissue layer that is continuously shifting, and obtain the continuously offset tissue region within the path. S503: Based on the continuous offset tissue regions within the path, follow the start and end positions of the offset of each frame structure in the image sequence, determine whether the region changes move forward according to the time sequence, track the continuous transfer positions of the response region in the sequence, and obtain the image response trajectory structure of traditional Chinese medicine pharmacology linkage.

Images