Method for preparing gastrointestinal transmission marker by using barium sulfate developing yarn and product

Abstract

The invention discloses a method for preparing a gastrointestinal transmission marker by using barium sulfate developing yarn, a product and an image processing method, and belongs to the technical field of medical instruments. The image processing method comprises the following steps: acquiring a medical image of the abdomen of a subject; identifying a first marker with a first form and a second marker with a second form in the medical image to obtain a preliminary counting result; counting the number of the first markers and the number of the second markers in preset gastrointestinal tract segmented areas to obtain count values of the markers in various forms in the segmented areas; based on the count value, one or more parameters characterizing gastrointestinal transport function of the subject are determined. According to the method, the barium sulfate developing yarn which is common clinically is utilized, so that the distinguishing and standardization problems of multi-day markers are solved, the inspection cost is reduced, and the accessibility and accuracy of diagnosis are improved.

Description

Technical Field

[0001] This invention relates to the field of medical technology, and in particular to a method and product for preparing gastrointestinal transit markers using barium sulfate imaging yarn. Background Technology

[0002] Gastrointestinal transit dysfunction, particularly slow colonic transit, is one of the core pathophysiological mechanisms of common digestive system diseases such as chronic constipation. Accurate measurement of colonic transit time (CTT) is of crucial clinical significance for the diagnosis, classification, and treatment guidance of these diseases. Currently, the gold standard method for CTT measurement is the radionuclide method, but its clinical application is limited due to its high equipment requirements, complex operation, and radiation exposure. In contrast, the radioisotope labeling method has gained widespread use due to its simplicity, non-invasiveness, and low radiation dose.

[0003] This method typically requires patients to swallow capsules containing a specific number of radioisotope markers over several consecutive days, followed by an abdominal X-ray on a designated date. The total CTT and segmental CTT are calculated by counting the number of residual markers in the body and their distribution in different segments of the colon (e.g., ascending colon, transverse colon, descending colon-sigmoid colon region). However, existing commercially available radioisotope marker products (such as Sitzmarks®) are expensive and have limited availability, making this test difficult to implement in primary healthcare institutions and economically underdeveloped areas, resulting in many patients not receiving accurate diagnoses. In the absence of commercially available products, some clinical units have attempted to use temporary alternatives, such as cut urinary catheters or infusion tubing. However, these alternatives are difficult to standardize in terms of size, density, and morphology, leading to poor reproducibility of measurements and making data comparison between different medical institutions impossible, severely impacting diagnostic reliability. Furthermore, accurate segmental CTT assessment requires the use of radiographically distinguishable multi-day markers, which existing temporary alternatives can hardly meet.

[0004] Therefore, there is an urgent need in this field for a gastrointestinal transit marker preparation scheme that is extremely low-cost, easy to obtain, standardized in operation, and capable of meeting the diagnostic needs of segmented transit time, as well as a reliable image analysis method to match it. Summary of the Invention

[0005] The purpose of this application is to provide a method, product, and image processing method for preparing gastrointestinal transit markers using barium sulfate imaging yarn, aiming to solve the technical problems of high cost, poor accessibility, and inaccurate diagnostic results caused by non-standard substitutes in the prior art.

[0006] In a first aspect, this application provides an image processing method for evaluating gastrointestinal transit function, comprising:

[0007] Obtain a medical image of the abdomen of a subject, wherein the medical image contains at least two radiographically distinguishable markers, the markers being prepared from the same medical-grade barium sulfate imaging yarn through different physical shaping methods, and being swallowed by the subject on different preset dates;

[0008] Identify a first marker having a first morphology and a second marker having a second morphology in the medical image;

[0009] The number of the first marker and the second marker in the preset gastrointestinal segment regions are counted respectively to obtain the count value of each morphological marker in each segment region;

[0010] Based on the count values, one or more parameters characterizing the subject's gastrointestinal transit function are determined.

[0011] In one possible implementation of the first aspect, the first form is a linear form, and the second form is a knot-like form or a V-shaped form.

[0012] In one possible implementation of the first aspect, the medical image further includes a third marker having a third form that is radiographically distinguishable from both the first and second forms;

[0013] The identification step further includes: identifying the third marker;

[0014] The counting step further includes: counting the number of the third markers within the preset gastrointestinal segment region;

[0015] The step of determining the parameters is based on the count values ​​of the first, second, and third markers.

[0016] In one possible implementation of the first aspect, the predefined gastrointestinal segmentation region includes the right colon region, the left colon region, and the sigmoid colon-rectum region.

[0017] In one possible implementation of the first aspect, the step of determining one or more parameters characterizing the subject's gastrointestinal transit function includes:

[0018] The total colonic transit time of the subject is calculated based on the total values ​​of all morphological markers across all predefined gastrointestinal segments; and / or,

[0019] Based on the count values ​​of all morphological markers in each of the predefined gastrointestinal segments, the segmental colonic transit time of the subject is calculated.

[0020] In one possible implementation of the first aspect, after the identification step, it further includes:

[0021] Based on the location information of all markers and their corresponding ingestion dates in the preliminary counting results, a marker movement probability map is generated for each ingestion date.

[0022] For each marker in the preliminary counting results, its morphological confidence score and kinematic confidence score are obtained, wherein the kinematic confidence score is determined based on the marker's location and the marker's motion probability map corresponding to the ingestion date;

[0023] Based on the morphological confidence and the kinematic confidence, the preliminary counting results are adaptively validated to generate a final counting value;

[0024] The step of determining one or more parameters characterizing the subject's gastrointestinal transit function is performed based on the final count value.

[0025] In one possible implementation of the first aspect, the step of generating a marker motion probability map for each ingestion date includes:

[0026] Extract the location data set of all markers for each ingestion date;

[0027] Kernel density estimation is performed independently for each set of location data points to generate the marker motion probability map.

[0028] In one possible implementation of the first aspect, the step of adaptively verifying the preliminary counting result includes:

[0029] For each marker, its morphological confidence is multiplied by its kinematic confidence to obtain a coupled confidence score;

[0030] The coupling confidence score is compared with a preset high confidence threshold and a preset low confidence threshold to accept, reject or label the marker.

[0031] Secondly, this application provides a method for preparing gastrointestinal transit markers using barium sulfate imaging yarn, comprising:

[0032] Prepare a plurality of first markers, wherein each of the first markers is a barium sulfate developing yarn segment having a first preset length and exhibiting a first morphology;

[0033] Multiple second markers are prepared, wherein each second marker is prepared by physically shaping a barium sulfate developing yarn segment to present a second morphology that is radiographically distinguishable from the first morphology.

[0034] The plurality of first markers are encapsulated in a first set of capsules, and the plurality of second markers are encapsulated in a second set of capsules, such that the first set of capsules and the second set of capsules can be used to assess the gastrointestinal transit function of the same subject on different dates.

[0035] In one possible implementation of the second aspect, the first form is a linear form.

[0036] In one possible implementation of the second aspect, the second morphology is a knotted morphology; the step of preparing multiple second markers includes: knotting multiple barium sulfate developing yarn segments having a second preset length to form the knotted morphology.

[0037] In one possible implementation of the second aspect, the second shape is a V-shape; the step of preparing multiple second markers includes: folding multiple barium sulfate developing yarn segments having a third preset length in half to form the V-shape.

[0038] In one possible implementation of the second aspect, the number of first markers in the first set of capsules is the same as the number of second markers in the second set of capsules.

[0039] Thirdly, this application provides a product for preparing gastrointestinal transit markers using barium sulfate developing yarn, comprising:

[0040] The marker raw material module comprises one or more rolls of medical-grade barium sulfate imaging yarn;

[0041] The marker forming and packaging module includes a cutting tool for cutting the barium sulfate developing yarn and at least two sets of hollow capsules for packaging.

[0042] The operating procedure instruction module includes an operating manual, which guides users to prepare at least two types of morphologically distinguishable marker capsules using the marker raw material module and the marker forming and packaging module.

[0043] In one possible implementation of the third aspect, the barium sulfate developing yarn is pre-printed with length markings.

[0044] In one possible implementation of the third aspect, the outer packaging of the at least two sets of empty capsules has visual markings for distinguishing different preparation dates.

[0045] In one possible implementation of the third aspect, a sterility assurance module is also included, in which the marker raw material module, the marker forming and dispensing module, and the operating procedure instruction module are all encapsulated. Attached Figure Description

[0046] Figure 1 This is a schematic diagram of the structure of a product for preparing a gastrointestinal transport marker according to an embodiment of this application.

[0047] Figure 2 This is a clinical application flowchart provided according to an embodiment of the present application, which includes the preparation method and image processing method described in the present application.

[0048] Figure 3 This is an exemplary abdominal X-ray image containing three different morphological markers in one embodiment of this application. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with the accompanying drawings and specific embodiments. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0050] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0051] This embodiment provides a product 100 for preparing gastrointestinal transit markers using barium sulfate-developable yarn. (Refer to...) Figure 1 The product 100 is a multi-module integrated system, including a marker raw material module 110, a marker forming and packaging module 120, an operating procedure instruction module 130, and a sterility assurance module 140.

[0052] The aseptic protection module 140 is the outermost structure of the product 100, used to establish and maintain an internal environment that meets the aseptic requirements of medical devices until the product is opened by the end user. In a specific preferred implementation, the aseptic protection module 140 is a pre-formed aseptic barrier system conforming to ISO 11607-1. Specifically, it is constructed as a packaging bag formed by heat-sealing a 70-micron thick multilayer co-extruded transparent composite film (e.g., PET / PE structure) of DuPont™ Tyvek® 1073B medical packaging material. The Tyvek® material provides excellent microbial barrier performance and gas permeability, allowing for ethylene oxide (EO) sterilization. In some embodiments, the sterilization cycle requires rigorous validation according to ISO 11135 standards. For example, the EO gas concentration is 600±50 mg / L, the temperature is 55±2°C, the relative humidity is 60±10%, the sterilization time is 4 hours, followed by at least 12 hours of forced ventilation and desorption to ensure that the residual ethylene oxide content is below the safety limit specified in ISO 10993-7. The heat-sealed edge width of the packaging bag is not less than 6 mm, and its seal strength needs to be verified by tensile peel test, with the peel force controlled within the range of 1.5±0.5 N / 15 mm to ensure seal integrity and ease of opening for the user. In some embodiments, a Class 4 chemical indicator conforming to ISO 11140-1 may also be printed on the packaging. This indicator will irreversibly change from its initial pink color to brown after a complete EO sterilization cycle, providing confirmation of the sterility status.

[0053] The following functional modules are encapsulated within the aseptic protection module 140:

[0054] The marker raw material module 110 may include a roll of individually aseptically packaged (e.g., placed in a smaller internal sterilization bag) medical-grade barium sulfate imaging yarn 111. To achieve the technical objectives of this application, the matrix material of the yarn 111 is a high-count (e.g., 60S) medical degreased cotton yarn made of long-staple cotton, validated by ISO 10993-5 cytotoxicity testing, or a viscose-based synthetic fiber with equivalent biocompatibility. The diameter of the yarn 111 can be monitored using a laser diameter gauge, with a preferred nominal diameter range of 0.3 mm to 0.8 mm. Exemplarily, a nominal diameter of 0.5 mm is used. This specification balances image visibility, operational flexibility, and structural strength. The loading process of barium sulfate (BaSO4) in the yarn 111 can employ a continuous production process of wet impregnation-calendering-drying. The barium sulfate powder used is micronized to meet the United States Pharmacopeia (USP) standard, with a median particle size (D50) of 1.5 ± 0.5 micrometers. This particle size ensures that the barium sulfate can uniformly and firmly penetrate and adhere to the fiber matrix. The final barium sulfate loading (on a dry weight basis) is quality controlled by thermogravimetric analysis and limited to the range of 30% to 60% (w / w). For example, the loading specification in this application is 45%, thereby ensuring that a 0.5 mm diameter yarn can produce an image contrast exceeding the background structure by at least five gray levels at X-ray energies of 70-90 kVp peak. To standardize subsequent operations, the surface of the yarn 111 can be printed with length markings using high-precision rotary gravure printing technology. The ink used is FDA-approved medical-grade gentian violet solution (Gentian Violet USP 1%), which has good fiber adhesion and does not leach out in the in vivo environment. The width of the scale marking lines is controlled to be less than 0.1 mm, with a short marking line of 1 mm length printed every 5.0 mm (±0.1 mm) and a long marking line of 2 mm length printed every 10.0 mm (±0.1 mm).

[0055] The marker forming and packaging module 120 includes a cutting tool 121 and hollow capsules 122. In one embodiment, the cutting tool 121 is a disposable precision iris scissor made of 316L medical-grade stainless steel. The capsules are size 0 (international standard) with a nominal volume of 0.68 ml, capable of holding approximately 20 shaped marker segments. In one embodiment, the capsule material is bovine bone gelatin conforming to pharmacopoeia standards, with a disintegration time of less than 15 minutes in water at 37°C; in another optional embodiment, it is hydroxypropyl methylcellulose (HPMC) to meet the needs of specific patient groups. Each set of hollow capsules 122 (20 capsules per set, plus 1 spare) is sealed in an individual aluminum-plastic composite bag with an easy-tear opening. A unique multimodal visual identifier 123 is printed on the outer surface of the aluminum-plastic composite bag. The visual identifier 123 is designed to prevent confusion through redundant information. For example, the first set of labels has a white background with the words "Day 1" printed in bold black font and a large Arabic numeral "1"; the second set of labels has a blue background with the words "Day 2" printed in bold white font and a large Arabic numeral "2"; and the third set of labels has a green background with the words "Day 3" printed in bold white font and a large Arabic numeral "3". In addition, each label also has a two-dimensional barcode printed on it containing the product batch number, production date, and expiration date, which can be scanned and traced by a medical information system.

[0056] The operating procedure instruction module 130 includes an operating manual 131. The content of the operating manual 131 is organized according to the operating flow. For example, under the section "4.0 Marker Preparation," there are sub-sections such as "4.1 Day 1 (Linear) Marker Preparation," "4.2 Day 2 (Nodular) Marker Preparation," and "4.3 Day 3 (V-shaped) Marker Preparation." Each sub-section is described in a fixed format: "Preparing Materials," "Step 1: Cutting," "Step 2: Shaping," "Step 3: Counting and Inspection," and "Step 4: Packaging," ensuring the standardization of the operating flow.

[0057] This application also provides a method for preparing gastrointestinal transit markers using barium sulfate-developable yarn. For example... Figure 2 As shown, the method includes:

[0058] S110. Prepare a plurality of first markers, wherein each of the first markers is a barium sulfate developing yarn segment having a first preset length and exhibiting a first morphology.

[0059] On a sterile work surface, the operator removes the first set of hollow capsules 122 marked with the visual identifier "Day 1" from product 100. Then, the barium sulfate developing yarn 111 is unfolded from the marker raw material module 110. The operator holds the yarn with sterile forceps in their left hand and holds the cutting tool 121 in their right hand, aligning the center line of its cutting edge with the center line of a 5 mm pre-printed length scale line on the yarn 111. A single vertical cut is made to obtain a yarn segment approximately 5.0 mm long (tolerance ±0.1 mm). This yarn segment undergoes no additional physical shaping after cutting; its naturally extended, approximately one-dimensional geometric shape under gravity is defined as the first morphology, i.e., the linear morphology. The above cutting process is repeated to generate twenty first markers. Before packaging, the twenty first markers can be quality inspected, including quantity verification and length sampling (e.g., randomly select three segments and confirm that their length meets the requirement of 5.0 ± 0.1 mm against a graduated background).

[0060] S120. Prepare a plurality of second markers, wherein each of the second markers is prepared by physically shaping a barium sulfate developing yarn segment to present a second morphology that is radiographically distinguishable from the first morphology.

[0061] In one specific implementation, the second form is defined as a knotted form. The operator removes the second set of hollow capsules 122 marked with the visual identifier "Day 2". Similarly, twenty segments of barium sulfate-developable yarn, each 5.0 ± 0.1 mm in length, are cut. For each individual 5 mm yarn segment, a standardized knotting process is performed. The knotting process, in a specific operational procedure, includes the following sub-steps: (a) holding one end of the yarn segment (approximately 0.5 mm); (b) wrapping the free end of the yarn segment around the tip of another closed toothless tweezers to form a loop (approximately 1.0 mm in diameter); (c) passing the free end completely through the loop; (d) holding both ends of the yarn with two tweezers and applying uniform force in opposite directions until a tight, irreversible single knot is formed. This physical shaping process is repeated once or multiple times, allowing the one-dimensional linear object to be topologically transformed into a point-like object in three-dimensional space, so that it stably presents as a high-density image of an approximately circular or elliptical shape under two-dimensional X-ray projection. Testing was conducted using an in vitro X-ray model. In some embodiments, the projection diameter of the nodular markers prepared by this method on X-ray films was controlled within the range of 1.5 mm to 2.5 mm. This size and high density characteristics enabled them to be clearly visually distinguished from linear markers with a length of 5 mm, gas images within the intestine, and low-density fecal artifacts.

[0062] In an optional extended implementation for the three-day assay, the preparation of a third marker is also included. The morphology of the third marker is defined as a V-shape or X-shape. The operator removes the third set of hollow capsules 122 marked with the visual identifier "Day 3". Twenty yarn segments of 10.0 ± 0.1 mm in length are cut along the 10 mm pre-printed length markings on the yarn 111. For each 10 mm yarn segment, a folding and fixing process is performed. The process specifically involves: (a) using tweezers to precisely fold the yarn segment along its geometric midpoint (5.0 mm), so that the two 5 mm long portions form a V-shape; (b) to ensure that the V-shape maintains its geometric stability during gastrointestinal peristalsis and compression, a fixing operation is performed after folding. In one embodiment, the fixation procedure is as follows: using a short piece of medical-grade 7-0 polydioxanone (PDS) absorbable suture, approximately 2 cm in length, the two parallel strands of suture are bound together twice at a point approximately 1 mm from the bend, and secured with a surgical square knot. Excess suture ends are then trimmed. In another optional embodiment, the fixation procedure is as follows: using an applicator, a microdrop (approximately 0.5-1.0 μL) of medical-grade N-butyl cyanoacrylate tissue adhesive is applied to the inside of the bend, and the area is held in place for 10 seconds until fully cured.

[0063] S130. The plurality of first markers are encapsulated in a first set of capsules, and the plurality of second markers are encapsulated in a second set of capsules.

[0064] The set of twenty first markers (linear) prepared in step S110 is placed into a capsule body taken from the first group of hollow capsules (size 0). The set of twenty second markers (knot-like) prepared in step S120 is placed into a capsule body taken from the second group of hollow capsules. If a third marker is prepared, the set of twenty third markers (V-shaped) is placed into a capsule body taken from the third group of hollow capsules and locked.

[0065] This application also provides an image processing method for evaluating gastrointestinal transit function, such as... Figure 2 As shown, the method includes:

[0066] S210. Obtain a medical image of the abdomen of a subject, wherein the medical image contains at least two radiographically distinguishable markers, the markers being prepared from the same medical-grade barium sulfate imaging yarn through different physical shaping methods and being swallowed by the subject on different preset dates.

[0067] In one clinical procedure, subjects swallow capsules labeled "Day 1," "Day 2," and "Day 3," prepared in step S130, at the same time each day for three consecutive days (e.g., Monday, Tuesday, and Wednesday) at 8:00 AM. On the target examination date, the fourth day (Thursday), at 8:00 AM, the subjects undergo a standard abdominal aneroposterior (AP) standing X-ray. The X-ray tube voltage for this examination can be set within the range of 70-90 kV peak voltage, and the product of tube current and exposure time (mAs) is automatically controlled for exposure (AEC) based on the subject's body thickness to ensure consistent image density. This examination allows for the acquisition of a diagnostic-quality digital medical image (DICOM format).

[0068] S220A, Identify a first marker having a first morphology and a second marker having a second morphology in the medical image to obtain a preliminary counting result.

[0069] In one embodiment, a radiologist opens the medical image 400 on a professional medical monitor that meets diagnostic requirements and has a brightness of at least 350 cd / m², and defines pre-defined gastrointestinal segment regions on the image. This definition process can be based on recognized anatomical landmarks: such as drawing a midline along the spinous processes of the spine; drawing a horizontal line along the highest points of the iliac crests on both sides; and drawing a second horizontal line along the fifth lumbar vertebra. These lines divide the colonic region of the abdominal cavity into three parts: the right colonic region (RC) 401, the left colonic region (LC) 402, and the sigmoid colon-rectum region (RS) 403.

[0070] Each partition is examined one by one, and all markers within the partition are identified and classified for counting.

[0071] Example 1 (Patient Zhang San, chief complaint: chronic constipation):

[0072] First marker (linear) count: 2 in RC area, 5 in LC area, and 8 in RS area.

[0073] Second marker (nodular) count: 6 in RC region, 10 in LC region, and 4 in RS region.

[0074] The third marker (V-shape) count is 15 in the RC area, 5 in the LC area, and 0 in the RS area.

[0075] Example 2 (Patient Li Si, chief complaint: difficulty in defecation, feeling of incomplete evacuation):

[0076] First marker (linear) count: 0 in RC area, 1 in LC area, and 3 in RS area.

[0077] Second marker (nodular) count: 1 in RC region, 4 in LC region, and 10 in RS region.

[0078] The third marker (V-shape) count is 2 in the RC area, 5 in the LC area, and 13 in the RS area.

[0079] In other embodiments, this step is performed by a computer device, the purpose of which is also to analyze the medical images acquired in S210 to generate a preliminary marker count result that may contain false positives and missed detections.

[0080] In a specific implementation, this step includes a series of sequentially executed image processing sub-steps.

[0081] S220B1. Preprocess the input digital medical image. Exemplarily, the preprocessing includes applying a contrast-limited adaptive histogram equalization (CLAHE) algorithm to enhance the local contrast of the image, making the grayscale difference between the marker and surrounding tissue more significant; and applying a 5x5 Gaussian filter to smooth the image to suppress random noise.

[0082] S220B2, Perform image segmentation. For example, a global thresholding segmentation method based on the Otsu algorithm is used to convert the preprocessed image into a binary mask, where regions with a pixel value of 1 represent potential high-density targets (i.e., candidate marker objects), and regions with a pixel value of 0 represent the background.

[0083] S220B3, Perform connected component analysis on the binary mask. This analysis is used to separate each individual candidate object and assign a unique identifier to each object. Simultaneously, morphological opening operations (erosion followed by dilation) are applied to eliminate tiny isolated pixels caused by noise.

[0084] S220B4, for each independent candidate object, extract a set of feature vectors to describe its geometric and morphological characteristics. For example, the feature vectors include the following 12 quantitative indicators: area (total number of pixels), perimeter, eccentricity (describing the degree to which the object deviates from a circle), compactness (…), and density (…). ), principal axis length, minor axis length, and seven invariant moments (M1 to M7) calculated from Hu moments.

[0085] S220B5 inputs the feature vector of each candidate object into a pre-trained classifier for morphological classification.

[0086] In one embodiment, the classifier is a support vector machine (SVM) with a radial basis function (RBF) as its kernel. The SVM classifier is trained offline on a dataset containing over 10,000 manually labeled samples (covering linear, knotted, V-shaped, and various artifacts). For each input feature vector, the SVM classifier outputs a classification label (e.g., 'linear,' 'knotted,' 'V-shaped,' or 'noise / artifact') and a corresponding confidence probability. The output of this step is a preliminary counting result list, where each list item contains the ID of an object identified as a marker and its centroid coordinates. Its classification labels and its morphological confidence level .

[0087] S230. After the identification step, an adaptive verification is performed to generate a final count value.

[0088] The purpose of this step is to cross-validate and correct the preliminary counting results generated by S220 in order to improve the accuracy of the final count.

[0089] S231. Based on the location information of all markers and their corresponding ingestion date information in the preliminary counting results, generate a marker movement probability map for each ingestion date.

[0090] This sub-step aims to transform the "time domain" information of the markers ingested over multiple days into a "spatial domain" probability distribution. The system first assigns the centroid coordinates of all objects identified as markers to three sets of location data points corresponding to their ingestion dates, based on the classification labels in the preliminary counting results of S220:

[0091]

[0092]

[0093]

[0094] Subsequently, the system processed these three independent location data point sets ( Each kernel density estimation (KDE) is performed independently to generate three consecutive marker motion probability maps with the same dimensions as the original X-ray images: In one specific implementation, a two-dimensional Gaussian kernel function is used. To better simulate the tubular movement patterns of substances within the colon, the bandwidth matrix of this Gaussian kernel... It is set to anisotropic. Specifically, the system first performs principal component analysis (PCA) on the overall distribution of all marker point clouds to determine the principal axis direction vectors of the macroscopic colonic orientation. and the secondary axis direction vector orthogonal to it Then, the bandwidth matrix Set as ,in It is by The rotation matrix formed, It is a larger bandwidth along the main axis (e.g., 50 pixels). This is a smaller bandwidth (e.g., 20 pixels) along the secondary axis. This setting makes the generated probability density smoother in the longitudinal direction of the colon and more concentrated in the transverse direction. Each probability map ultimately generated, for example... any point on it pixel values This represents the probability density of the marker ingested on the first day appearing at that spatial location.

[0095] S232. For each marker in the preliminary counting results, obtain its morphological confidence score and kinematic confidence score, wherein the kinematic confidence score is determined based on the marker's location and the marker's motion probability map corresponding to its ingestion date. For each marker in the preliminary counting results of S220... The system obtains two independent confidence metrics:

[0096] Morphological confidence This value comes directly from the output of the SVM classifier in step S225, and its value ranges from [0-1]. An example value is 0.98.

[0097] Kinematic confidence This value is obtained by querying its centroid coordinates. The normalized probability value is obtained from the motion probability map corresponding to the ingestion date. For example, if If the category label is 'V-shaped' (corresponding to the third day), then... That is In coordinates The value at the specified location is normalized to make its range [0-1].

[0098] S233. Based on the morphological confidence and the kinematic confidence, the preliminary counting result is adaptively validated. This sub-step integrates information from the two dimensions and performs an adaptive threshold decision to generate a validated final count value.

[0099] For each marker Calculate a coupling confidence score:

[0100]

[0101] Subsequently, the coupling confidence score is compared with a set of preset thresholds, and the decision logic is executed:

[0102] High confidence level acceptance: for all The system accepts the marker as a true positive. The high confidence threshold... For example, it is set to 0.8.

[0103] Low confidence rejection: For all The system identifies markers as artifacts or serious misidentifications and removes them from the counting results. The low confidence threshold... For example, it is set to 0.2.

[0104] Adaptation of fuzzy intervals: For markers located in fuzzy intervals ( The system activates an adaptation mechanism. The rule for this adaptation is: if the kinematic confidence of a marker... Above a preset rescue threshold (For example, set to 0.9), then it is forced to be accepted as a valid count regardless of its coupling confidence score. This mechanism is used to retain those with low morphological confidence due to severe obscuring by feces (e.g., However, its location perfectly matches the true marker of the motion probability model corresponding to its ingestion date. For example, the rescue threshold...

[0105] It is set to 0.9 (dimensionless).

[0106] S240. Count the number of the first marker and the second marker in the preset gastrointestinal segment regions respectively to obtain the count value of each morphological marker in each segment region, and determine one or more parameters characterizing the gastrointestinal transit function of the subject based on the count value.

[0107] This step utilizes the final count values ​​generated in S230 (or, if S230 is not performed, the preliminary count values ​​from S220) for final quantitative analysis. Each valid marker in the final count value list is assigned to a predefined gastrointestinal segment region (Right Colon RC, Left Colon LC, Sigmoid Colon-Rectum RS) based on its centroid coordinates. Then, each type of marker within each region is categorized and summarized to obtain the final count value. Finally, based on the count values, standard calculation formulas are applied to determine parameters such as segmentation and total colonic transit time.

[0108] Before performing specific numerical calculations, the standard calculation formula described in this embodiment is defined. This standard calculation formula is a recognized method in the art for calculating colonic transit time from the results of radioisotope marker counting (e.g., based on the Metcalf method).

[0109] Formula for calculating segmental colonic transit time (CTT):

[0110] The colonic transit time for a given segment is calculated based on the total number of all residual markers within that segment. The general formula is:

[0111] (Hour)

[0112] in:

[0113] Colonic transit time representing a specific segment (e.g., the right colonic region RC).

[0114] This represents the total number of all morphological markers counted within that specific segmented region.

[0115] This represents the total number of markers ingested daily (in this example, the value is 20).

[0116] Total CTT is the sum of the transit times of all segments of the colon, and its general formula is:

[0117] (Hour)

[0118] in, , , These represent the segmental colonic transit times for the right colonic region, left colonic region, and sigmoid colon-rectum region, respectively.

[0119] The Regional Transit Time Ratio (RTR) is used to further analyze the location of major transit barriers. It calculates the ratio of transit time in each segment to the total colonic transit time. The general formula is:

[0120]

[0121] For example, assuming the image of Example 1 (patient Zhang San) is being processed, the automated identification step of S220 generates the following preliminary count result list containing errors due to interference factors in the image:

[0122] A false positive: A small gallbladder calcification, located at the lower margin of the liver projection in the right hemicolon (RC), was incorrectly identified as a "nodular" marker by the SVM classifier due to its dense morphology. Its morphological confidence level... The value is medium (for example, 0.75).

[0123] A potential false negative: Deep in the sigmoid-rectal region (RS), a true "linear" marker becomes blurred due to severe obscuring by large amounts of feces and gas, leading to a lower morphological confidence score from the SVM classifier. Extremely low (exemplarily, 0.35), placing it on the edge of being rejected by the low confidence threshold of S225.3.

[0124] Therefore, the preliminary counting result output by S220 is:

[0125] First marker (linear) count: 2 in RC region, 5 in LC region, 8 in RS region (1 of which is...) (Only 0.35).

[0126] Second marker (nodular) count: 7 in the RC region (1 of which was a false positive). (0.75), 10 in the LC area, and 4 in the RS area.

[0127] The third marker (V-shape) count is 15 in the RC area, 5 in the LC area, and 0 in the RS area.

[0128] Step S230 receives the aforementioned flawed preliminary count results and performs verification on each of the markers:

[0129] For the misidentified "nodular" marker (calcification point) in the RC region, the system calculates its coupling confidence. Because this point is far from... The high probability region of the (second day marker motion probability map) has a kinematic confidence level. Extremely low (exemplarily, 0.01). Therefore, its coupling confidence is... This value is far below the preset low confidence threshold. (0.2), therefore, the marker is identified as an artifact by the system and rejected, and is not included in the final count value.

[0130] For the occluded "linear" marker in the RS region, its morphological confidence level Although it is only 0.35, its position is exactly in The high-probability core region of the (first day marker motion probability map) has a kinematic confidence level. Extremely high (exemplary, 0.99). Its coupling confidence level The value falls within a fuzzy range (between 0.2 and 0.8). At this point, the system activates the "rescue" mechanism because... (0.99) is far higher than the preset rescue threshold. (0.9), the marker is determined by the system to be a real marker and is forcibly accepted as a valid count, and is included in the final count value.

[0131] After verification by S230, the generated final count list was corrected. The system then assigns each valid marker in the list to a predefined gastrointestinal segment region (RC, LC, RS) based on its centroid coordinates, and categorizes and summarizes the markers of each morphology within each region to obtain the final count matrix.

[0132] First marker (linear) count: 2 in RC area, 5 in LC area, and 8 in RS area (rescued markers are confirmed as valid).

[0133] Second marker (nodular) count: 6 in RC region (false positives were eliminated), 10 in LC region, and 4 in RS region.

[0134] The third marker (V-shape) count is 15 in the RC area, 5 in the LC area, and 0 in the RS area.

[0135] The final quantitative analysis was performed using the revised final count matrix.

[0136] The final total number of residuals in each segment is as follows: RC segment = 2 + 6 + 15 = 23; LC segment = 5 + 10 + 5 = 20; RS segment = 8 + 4 + 0 = 12.

[0137] The final segmented CTT (hours): RC segment = (23 / 20) * 24 = 27.6 h; LC segment = (20 / 20) * 24 = 24.0 h; RS segment = (12 / 20) * 24 = 14.4 h.

[0138] The final total CTT (hours): 27.6 + 24.0 + 14.4 = 66.0 h.

[0139] Additional parameters (regional transit time ratios): RC / total CTT = 41.8%; LC / total CTT = 36.4%; RS / total CTT = 21.8%. This distribution suggests slow transit throughout the colon.

[0140] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit described above can be implemented in hardware.

[0141] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. An image processing method for evaluating gastrointestinal transit function, characterized in that, include: Obtain a medical image of the abdomen of a subject, wherein the medical image contains at least two radiographically distinguishable markers, the markers being prepared from the same medical-grade barium sulfate imaging yarn through different physical shaping methods, and being swallowed by the subject on different preset dates; Identify a first marker with a first morphology and a second marker with a second morphology in the medical image to obtain a preliminary counting result; The number of the first marker and the second marker in the preset gastrointestinal segment regions are counted respectively to obtain the count value of each morphological marker in each segment region; Based on the count values, one or more parameters characterizing the subject's gastrointestinal transit function are determined.

2. The method according to claim 1, characterized in that, The first form is a linear form, and the second form is a knot-like or V-shaped form.

3. The method according to claim 1, characterized in that, The medical image also includes a third marker having a third form that is radiographically distinguishable from both the first and second forms; The identification step further includes: identifying the third marker; The counting step further includes: counting the number of the third markers within the preset gastrointestinal segment region; The step of determining the parameters is based on the count values ​​of the first, second, and third markers.

4. The method according to claim 1, characterized in that, The step of determining one or more parameters characterizing the subject's gastrointestinal transit function includes: The total colonic transit time of the subject is calculated based on the total values ​​of all morphological markers across all predefined gastrointestinal segments; and / or, Based on the count values ​​of all morphological markers in each of the predefined gastrointestinal segments, the segmental colonic transit time of the subject is calculated.

5. The method according to claim 1, characterized in that, Following the identification step, the method further includes: Based on the location information of all markers and their corresponding ingestion dates in the preliminary counting results, a marker movement probability map is generated for each ingestion date. For each marker in the preliminary counting results, its morphological confidence score and kinematic confidence score are obtained, wherein the kinematic confidence score is determined based on the marker's location and the marker's motion probability map corresponding to its ingestion date; Based on the morphological confidence and the kinematic confidence, the preliminary counting results are adaptively validated to generate a final counting value; The step of determining one or more parameters characterizing the subject's gastrointestinal transit function is performed based on the final count value.

6. A method for preparing gastrointestinal transport markers using barium sulfate developing yarn, characterized in that, include: Prepare a plurality of first markers, wherein each of the first markers is a barium sulfate developing yarn segment having a first preset length and exhibiting a first morphology; Multiple second markers are prepared, wherein each second marker is prepared by physically shaping a barium sulfate developing yarn segment to present a second morphology that is radiographically distinguishable from the first morphology. The plurality of first markers are encapsulated in a first set of capsules, and the plurality of second markers are encapsulated in a second set of capsules, such that the first set of capsules and the second set of capsules can be used to assess the gastrointestinal transit function of the same subject on different dates.

7. The method according to claim 6, characterized in that, The first form is a linear form, and the second form is a knot-like form; The step of preparing multiple second markers includes: knotting multiple barium sulfate developing yarn segments with a second preset length to form the knotted shape.

8. The method according to claim 7, characterized in that, The second shape is a V-shape; the step of preparing multiple second markers includes: folding multiple barium sulfate developing yarn segments with a third preset length in half to form the V-shape.

9. The method according to claim 6, characterized in that, The number of first markers in the first group of capsules is the same as the number of second markers in the second group of capsules.

10. A product for preparing gastrointestinal transit markers using barium sulfate developing yarn, characterized in that, include: The marker raw material module comprises one or more rolls of medical-grade barium sulfate imaging yarn; The marker forming and packaging module includes a cutting tool for cutting the barium sulfate developing yarn and at least two sets of hollow capsules for packaging. The operating procedure instruction module includes an operating manual, which guides users to prepare at least two types of morphologically distinguishable marker capsules using the marker raw material module and the marker forming and packaging module.

Images