Variable wall thickness in-situ artificial anus, anti-falling plug and matching design method thereof

Abstract

This invention discloses a variable-wall-thickness in-situ artificial anus, an anti-fall plug, and their matching design method. The artificial anus comprises a variable-wall-thickness double-layer structure, with an overall cone-shaped form. This structure uses a flexible medical-grade impermeable material. The double-layer structure includes an inner layer and an outer layer, forming a buffer cavity between them to absorb dynamic loads and adapt to tissue deformation. A spherical contraction region is provided on the inner side of the tube opening near the outer end of the anus; this region protrudes inward to form an elastic constriction structure. The matching anti-fall plug is spindle-shaped; after being inserted into the artificial anus with an interference fit, it is elastically held within the contraction region, achieving reliable anti-fall and sealing. This invention utilizes a tomographic scanning fusion network embedded with physical constraints for structural parameter design, which can adapt to different cone angles, lengths, diameters, and uneven variable-wall-thickness distributions according to the patient's in-situ anal shape. It has support in a specific direction, and the insertion and removal structure has strong cooperative adaptability.

Description

Technical Field

[0001] This invention belongs to the field of medical device technology, specifically relating to a variable wall thickness in-situ artificial anus, an anti-fall plug, and a matching design method thereof. Background Technology

[0002] Currently, various types of artificial anal devices exist, such as silicone cannulas, expandable annular supports, or magnetic closure devices. Existing anal plug structures include a method for fixing a perianal support device described in patent (David D. Blurton. METHOD OF SECURING PERIANALS UPPORT DEVICE: US Patent US11406422B2 [P]. 2022-08-09). This method includes a support with a central perianal tissue compression member for providing support to perianal soft tissue. Patent (Fred G. DEVICE FOR SEALING AND DRAINAGE OF ANATURAL OR ARTIFICIAL INTESTINAL OUTLET: US Patent US8529429B2 [P]. 2013-09-10) describes a medical device structure for closing and draining an artificial stoma. This structure achieves stoma closure through a radially expanding rectal device within the rectum or intestine. A method for assisting sphincter control is described in the patent (Frederick E. METHOD FOR ASSISTINGA SPHINCTER: US Patent US10405865B2 [P]. 2019-09-10). It includes an annular member fixed to surrounding tissue by sharp anchors. The member contains multiple valves that open and close under magnetic force.

[0003] These devices generally have the following problems: (1) The structure is too rigid, making it difficult to adapt to the anatomical differences of the anal tissues of different patients, and it is difficult to meet the support in a specific direction, which can easily cause local compression, friction damage or displacement; (2) There is a lack of effective anti-fall design, and the device is easy to fall out during the patient's activity, coughing or defecation, which can lead to fecal leakage or infection risk; (3) Most of the existing enterostomy plugs are independent plugs, which have no structural coordination with the artificial anal body, have weak spatial curvature matching, and insufficient sealing and stability.

[0004] Therefore, there is an urgent need for an in-situ artificial anus that is structurally sound, biocompatible, adaptable to individual differences, and has reliable anti-fall function, as well as a method that can design support in specific directions according to the physiological structure of different patients. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a variable-wall-thickness in-situ artificial anus, a fall-prevention plug, and a matching design method. The method proposed in this invention can adaptively adjust its shape according to the individual anatomical characteristics of the patient, better restoring the patient's original structure, and achieving reliable fall prevention and sealing functions through structural synergy, thereby improving user comfort and safety.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] In a first aspect, the present invention proposes a design method for a variable wall thickness in-situ artificial anus and a fall-prevention plug, the method comprising the following steps:

[0008] Step 1: Obtain the patient's human feature parameters and construct the basic feature vector and the inner contour features of the affected area;

[0009] Step 2: Input the basic feature vector into the pre-trained artificial anal parameter prediction model. The model outputs the artificial anal geometric parameter vector. Based on the artificial anal geometric parameter vector, construct a basic model of uniform wall thickness for the artificial anal.

[0010] Step 3: Input the inner contour features of the affected area into the pre-trained artificial anal inner contour parameter prediction model. The model outputs the in-situ inner contour features. Combined with the basic model of uniform wall thickness of the artificial anus, the model of variable wall thickness of the artificial anus is generated.

[0011] Step 4: Design the corresponding anti-fall plug based on the artificial anal wall thickness model.

[0012] Furthermore, the pre-training process of the artificial anal parameter prediction model is as follows:

[0013] For each training iteration of the artificial anal parameter prediction model, the basic feature vector of each training sample is input into the model, and the model outputs the corresponding artificial anal geometric parameter vector. Based on the current artificial anal geometric parameter vector, an artificial anal mesh model is generated, and the physical displacement field of the current artificial anal mesh model under abdominal pressure compression is calculated. Then, the strain field is solved based on the current physical displacement field. and stress field Based on the strain field obtained from the solution and stress field Calculate the total physical loss, adjust the parameters of the artificial anal parameter prediction model based on the current total physical loss, and complete one training cycle.

[0014] Furthermore, the total physical loss Satisfy the following formula:

[0015]

[0016]

[0017] in, The residuals of the equilibrium equation; The residuals of the geometric equations; The constitutive equation residuals; For boundary condition residuals; External force; This represents the stress field divergence. The gradient of the displacement field; T represents the transpose operation; The strain energy of a micro-element in an elastic material; This represents the desired displacement field of the artificial anal mesh model.

[0018] Furthermore, the pre-training process of the artificial anal contour parameter prediction model is as follows:

[0019] For each training iteration of the artificial anal internal contour parameter prediction model, the internal contour features of the affected area for each training sample are first input into the model, and the model outputs the corresponding artificial anal internal contour parameter vector. Then, a reconstructed internal contour is generated based on this parameter vector. Next, the internal contour in the uniform wall thickness model of the artificial anus is replaced with the reconstructed internal contour to obtain a variable wall thickness model. Finally, the physical displacement field of the current variable wall thickness model under abdominal pressure compression is calculated. Then, the strain field is solved based on the current physical displacement field. and stress field Based on the strain field obtained from the solution and stress field Calculate the total physical loss, and adjust the parameters of the artificial anal contour parameter prediction model based on the current total physical loss to complete one training cycle.

[0020] Furthermore, in step 4, the maximum outer diameter D1 of the anti-fall plug satisfies the following formula:

[0021] D1 = (1.03~1.08) × d

[0022] Where d is the small diameter of the contraction zone in the artificial anal wall thickness model.

[0023] Furthermore, the artificial anus includes a connected tissue portion and an anal plug portion. The tissue portion includes an inner layer and an outer layer. One end of both the inner and outer layers is connected to the anal plug portion, and the other end of the inner and outer layers is used to connect to the patient's tissue. The cavity between the inner and outer layers is referred to as the buffer cavity. The internal cavity of the anal plug portion is referred to as the spherical contraction area. The spherical contraction area is used to place an anti-fall plug, and the inner layer is connected to the spherical contraction area.

[0024] Furthermore, the anti-fall plug includes a connected insertion structure and a flange structure; the insertion structure can be embedded in the spherical contraction area of ​​the artificial anus.

[0025] Secondly, this invention proposes a design system for a variable-wall-thickness in-situ artificial anus and a fall-prevention plug, the system comprising:

[0026] The human body parameter acquisition unit is used to acquire the human body characteristic parameters of the subject.

[0027] The feature construction unit is used to construct basic feature vectors and inner contour features of the affected area based on the acquired human body feature parameters of the patient.

[0028] The first prediction unit is used to input the basic feature vector into the pre-trained artificial anal parameter prediction model. The model outputs the artificial anal geometric parameter vector, and constructs a basic model of uniform wall thickness of the artificial anal based on the artificial anal geometric parameter vector.

[0029] The second prediction unit is used to input the inner contour features of the affected area into the pre-trained artificial anal inner contour parameter prediction model. The model outputs the in-situ inner contour features, which, combined with the basic model of uniform wall thickness of the artificial anus, generate the artificial anal variable wall thickness model.

[0030] An anal plug design unit is used to design corresponding anti-fall plugs based on the artificial anal wall thickness model.

[0031] Thirdly, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method described in the first aspect above.

[0032] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described in the first aspect above.

[0033] The beneficial effects of this invention are as follows:

[0034] The artificial anus of this invention uses medical-grade impermeable material and features a flexible structure with variable wall thickness, combining flexibility and support to facilitate tissue ingrowth and long-term stability. The spherical contraction zone of the artificial anus, in conjunction with the matching anal plug, employs an interference fit design to ensure reliable anti-fall function.

[0035] This invention uses a tomographic fusion neural network with embedded physical constraints to extract the mapping relationship between the basic structure of the artificial anus and the anal type characteristics of the patient. For different size parameters of patients with different anal types, a basic model of uniform wall thickness of the artificial anus can be quickly generated.

[0036] This invention utilizes a personalized custom network to extract the mapping relationship between the original lesion and the tissue after lesion removal, enabling rapid personalized matching design for patients with different anal types. The resulting variable wall thickness anal canal has a supporting function in a specific direction. Attached Figure Description

[0037] Figure 1 This is a schematic diagram of the uniform wall thickness in-situ artificial anus of the present invention.

[0038] Figure 2 These are schematic diagrams of in-situ artificial anuses with varying wall thickness corresponding to three typical anal types; (a) is the in-situ artificial anuse for tall and thin patients; (b) is the in-situ artificial anuse for neutral patients; and (c) is the in-situ artificial anuse for short and stout patients.

[0039] Figure 3 This is a schematic diagram of the anti-fall plug of the present invention.

[0040] Figure 4 This is an assembly diagram of the variable wall thickness in-situ artificial anus and the anti-fall plug of the present invention; wherein, (a) is a schematic diagram before the combination of the uniform wall thickness in-situ artificial anus and the anti-fall plug; and (b) is a schematic diagram after the combination of the uniform wall thickness in-situ artificial anus and the anti-fall plug.

[0041] Figure 5 These are structural diagrams of an in-situ artificial anus with uniform wall thickness; where (a) is a front view of the in-situ artificial anus with uniform wall thickness; (b) is a sectional view of the in-situ artificial anus with uniform wall thickness along the AA plane; and (c) is a sectional view of the in-situ artificial anus with uniform wall thickness along the BB plane.

[0042] Figure 6 These are structural diagrams of an in-situ artificial anus with variable wall thickness; where (a) is a front view of the in-situ artificial anus with variable wall thickness; (b) is a cross-sectional view of the in-situ artificial anus with variable wall thickness along the CC plane; and (c) is a cross-sectional view of the in-situ artificial anus with uniform wall thickness along the DD plane.

[0043] Figure 7 Sagittal magnetic resonance imaging of the anorectal rectum in the patient.

[0044] Figure 8 This is a flowchart of the method of the present invention.

[0045] Figure 9 This is a six-iteration model in the iterative process of a personalized variable wall thickness artificial anus model; where (a) is the cross-sectional view of the model generated after the first iteration based on the changed inner wall points and the corresponding physical loss, (b) is the cross-sectional view of the model generated after the second iteration based on the changed inner wall points and the corresponding physical loss, (c) is the cross-sectional view of the model generated after the third iteration based on the changed inner wall points and the corresponding physical loss, (d) is the cross-sectional view of the model generated after the fourth iteration based on the changed inner wall points and the corresponding physical loss, (e) is the cross-sectional view of the model generated after the fifth iteration based on the changed inner wall points and the corresponding physical loss, and (f) is the cross-sectional view of the model generated after the sixth iteration based on the changed inner wall points and the corresponding physical loss.

[0046] Figure 10This study compares the inner contour parameters of 11-layer slices from a basic model with uniform wall thickness for artificial anus and a customized model with variable wall thickness for artificial anus. Figure 1 Among them, (a) is a graph showing the change in the center distance of the sampling points obtained by sampling along 360° of the inner wall of the pre-trained basic model (i.e., the constant wall thickness model) and the final variable wall thickness model obtained by personalized color design; (b) is a graph showing the change in the area enclosed by the inner rim width of different layers of the pre-trained basic model (i.e., the constant wall thickness model) and the final variable wall thickness model obtained by personalized design.

[0047] Figure 11 This study compares the inner contour parameters of 11-layer slices from a basic model with uniform wall thickness for artificial anus and a customized model with variable wall thickness for artificial anus. Figure 2 Among them, (a) is a bar graph showing the changes in the center distance of the sampling points obtained by sampling along 360° of the inner wall of the pre-trained basic model (i.e., the equal wall thickness model) and the final variable wall thickness model obtained by personalized design, and the deviation of the standard deviation of the deviation between the center distance of the sampling points obtained by sampling along 360° of the inner wall of the pre-trained basic model (i.e., the equal wall thickness model) and the final variable wall thickness model obtained by personalized design, and the mean deviation of the center distance of the sampling points obtained by sampling along 360° of the inner wall of the final variable wall thickness model obtained by sampling along 360° of the inner wall of the final variable wall thickness model, and the deviation of the center distance between the center distance of the sampling points of the final variable wall thickness model obtained by sampling along 360° of the inner wall of the final variable wall thickness model, and the deviation of the deviation ...;

[0048] Figure 12 A comparison of parameters between a basic model of a uniformly thick artificial anus and a customized model of a variable-thickness artificial anus. Figure 3 Among them, (a) is a histogram of the deviation between the area enclosed by the inner contours of different layers of the pre-trained basic model (i.e., the isothickness model) and the area enclosed by the inner contours of different layers of the patient's in-situ anal contour; (b) is a histogram and error bars of the mean deviation between the area enclosed by the inner contours of different layers of the pre-trained basic model (i.e., the isothickness model) and the area enclosed by the inner contours of different layers of the patient's in-situ anal contour.

[0049] Figure 13 Comparison of anti-fall plugs for different artificial anuses; where (a) is the first example model of the anti-fall plug, (b) is the second example model of the anti-fall plug, (c) is the third example model of the anti-fall plug, (d) is the fourth example model of the anti-fall plug, (e) is the fifth example model of the anti-fall plug, and (f) is the sixth example model of the anti-fall plug.

[0050] In the diagram: 1. Artificial anus; 2. Inner layer; 3. Outer layer; 4. Buffer cavity; 5. Spherical contraction zone; 6. Insertion structure; 7. Flange structure. Detailed Implementation

[0051] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0052] The design goal of the artificial anus and the anti-fall plug is to simulate the natural anatomical shape and mechanical behavior of the human anal canal to the greatest extent possible while replacing or assisting the lost function of the anal sphincter.

[0053] like Figure 1 As shown, the artificial anus 1 can be made of flexible medical materials, and is an overall cone-shaped hollow tube with a distal port diameter larger than the proximal port diameter. The artificial anus 1 includes a connected tissue portion and an anal plug portion. The tissue portion includes an inner layer 2 and an outer layer 3. One end of both the inner layer 2 and the outer layer 3 is connected to the anal plug portion, and the other end of the inner layer 2 and the outer layer 3 serves as the distal port for connecting to the patient's tissue. The inner layer 2 and the outer layer 3 are coaxially arranged. The cavity between the inner layer 2 and the outer layer 3 is designated as the buffer cavity 4. The anal plug portion serves as the proximal port, and the internal cavity of the anal plug portion is designated as the spherical contraction area 5. The spherical contraction area 5 is used to place an anti-fall plug, and this area protrudes inward to form an elastic constriction structure. The inner layer 2 communicates with the spherical contraction area 5, forming a flow channel structure that gradually narrows from the inside out within the artificial anus 1.

[0054] In one feasible implementation, the spherical contraction region can be formed into an elastic necking structure by local heat setting and increasing the number of braided layers. The heat setting treatment causes the polypropylene fibers to form a permanent crimped structure in this region. The inner surface of the spherical contraction region 5 is smooth and can be coated with a hydrophilic lubricating coating to reduce the insertion and removal resistance of the anal plug. The core function of the spherical contraction region is to form an elastic snap-fit ​​with the matching anti-fall plug. Figure 4 (a) and Figure 4 As shown in (b), when the anti-fall plug is inserted, the artificial anal structure deforms and expands outward, and the spherical contraction area generates a circumferential clamping force due to interference, which stably locks the anti-fall plug inside and effectively prevents it from falling out.

[0055] Optionally, the cone shape of the artificial anus 1 is not a simple frustum, but a non-linear taper curve optimized based on human anatomical data, with its generatrix being a slightly concave curve to better conform to the physiological curvature of the anal canal-rectum junction.

[0056] Optionally, the flexible medical material is a medical waterproofing material. This material has good biocompatibility and tissue ingrowth ability.

[0057] The artificial anus 1 has a hollow buffer cavity. This absorbs momentary impact loads when the patient coughs, jumps, or defecates, preventing rigidity transmission to surrounding tissues. The artificial anus 1's double-layer structure provides anisotropic flexibility—axially flexible to conform to the perineal curvature, and radially pressure-resistant to maintain lumen openness and prevent collapse leading to defecation difficulties. The outer layer of the artificial anus 1 directly contacts the surrounding muscles and skin, while the inner layer contacts the intestinal contents; a physical barrier is formed between the two layers, reducing the risk of infection spread. Preferably, both the inner and outer layers are made of the same flexible medical material, but the stiffness distribution can be controlled by locally creating density differences. For example, the distal region can be made denser to enhance support, while the proximal region can be made looser to improve flexibility.

[0058] like Figure 3 As shown, the anti-fall plug comprises an interconnected insertion structure 6 and a flange structure 7. The anti-fall plug is geometrically symmetrical, without sharp edges, and its surface is polished or coated to reduce frictional damage to the anal canal mucosa during insertion / removal. The insertion structure 6 can be embedded in the spherical contraction zone 5 of the artificial anus 1. The insertion structure 6 is generally spherical or ellipsoidal, with its maximum outer diameter larger than the inner diameter of the spherical contraction zone, allowing the anti-fall plug to be elastically held within the spherical contraction zone after insertion, achieving anti-fall fixation and sealing.

[0059] In one feasible implementation, the parameters of the artificial anus and anti-fall plug include the cone angle, axial length, distal inner diameter, and proximal inner diameter. The cone angle is the angle between the generatrices on both sides of the axis. A smaller cone angle is suitable for patients with wide pelvis or significant rectal ampulla bulge, reducing the distensible stress on surrounding tissues at the distal end; a larger cone angle is beneficial for children or thin patients, achieving a more compact implantation. The axial length refers to the straight-line distance from the distal opening to the proximal opening. The axial length needs to be anastomosed with the anal canal to low rectum segment (approximately 40–100 mm) in most adults and adolescents, ensuring that the distal end can be safely anastomosed to the rectal stump, and the proximal end is located precisely at the level of the original anal opening.

[0060] Generally, the cone angle θ of an artificial anus is 60°–100°, the axial length is 45–70 mm, the distal inner diameter is 70–120 mm, and the proximal inner diameter is 35–50 mm. Adjustments can be made to fit the patient's anal type. Patients' anal types include tall and thin, medium, or short and stout, and the corresponding artificial anus body uses different combinations of cone angles, lengths, and diameters to achieve anatomical fit.

[0061] Clinically, there are three typical anal type populations: tall and thin, medium-sized, and short and stout.

[0062] Figure 2(a) shows an in-situ artificial anus suitable for tall and thin patients. Tall and thin patients typically have a longer pelvic cavity, a narrower perineal distance, and a smaller volume of perianal tissue. For this population, the artificial anus is designed with a smaller cone angle, a longer axial length, and a smaller diameter to avoid excessive stretching of surrounding tissue and reduce implantation tension. Designable configuration parameters: cone angle θ = 64°, axial length = 60 mm, distal inner diameter = 71 mm, proximal inner diameter = 31 mm. The artificial anus is based on a cylindrical structure; therefore, the distal inner diameter refers to the diameter of the larger end of the cone, and the proximal diameter refers to the diameter of the smaller end. This configuration allows the artificial anus to closely conform to the steeper anal canal of tall and thin patients, reducing stress concentration due to excessive length, while the smaller diameter reduces the risk of compression on weak sphincter remnants.

[0063] Figure 2 (b) is an in-situ artificial anus suitable for patients with a neutral body mass index (BMI) of 18.5–24.9, characterized by a symmetrical anal sphincter, representing the majority of clinical cases. For these patients, the artificial anus employs balanced geometric parameters, balancing support, flexibility, and sealing. Designable configuration parameters include: a cone angle θ of 77°, an axial length of 50 mm, a distal inner diameter of 76 mm, and a proximal inner diameter of 36 mm. This configuration maintains sufficient lumen opening to ensure unobstructed defecation while effectively absorbing dynamic loads during daily activities through a double-layered buffer structure.

[0064] Figure 2 (c) is an in-situ artificial anus suitable for obese patients. Obese patients often have a short pelvis, wide perineum, thick subcutaneous fat, and a large rectal ampulla volume. If a standard model is used, it is easy to cause excessive tension at the distal anastomosis or insufficient proximal exposure. Therefore, this embodiment adopts a design with a larger cone angle, a longer axial length, and a larger tube diameter. Designable configuration parameters: cone angle θ is 100°, axial length is 50 mm, distal inner diameter is 116 mm, and proximal inner diameter is 46 mm. The larger cone angle makes the tube body more gentle, adapting to the more horizontal anorectal angle of obese patients; the extended design ensures that the distal end can be anastomosed with the high rectal stump without tension, while the larger inner diameter matches the possible enlargement of the rectal ampulla and prevents fecal retention.

[0065] Generally, artificial anuses are designed with a uniform wall thickness, such as Figure 5 of (a), Figure 5 (b) and Figure 5 As shown in (c). The actual wall thickness of the human anal tissue is uneven, as... Figure 6 of (a), Figure 6 (b) and Figure 6As shown in (c). To achieve better functional replacement, the wall thickness of the column and cone of the artificial anus is not uniform. The so-called non-uniformity includes the uneven change in wall thickness in the longitudinal section (CC) from the side of the artificial anus facing the rectal stump to the side of the main body facing outward, and the uneven change in wall thickness around the same section in the cross section (DD).

[0066] like Figure 8 As shown, the design method of a variable wall thickness in-situ artificial anus and anti-fall plug proposed in this invention includes the following steps:

[0067] Step 1: Obtain the patient's human feature parameters and construct the basic feature vector and the inner contour features of the affected area;

[0068] In one feasible implementation, the basic feature vector Including the patient's anal canal length The major diameter of the ellipse is similar to that of the rectal stump. (i.e., the length of the semi-major axis of the ellipse) and minor diameter (i.e., the length of the minor semi-axis of the ellipse), the patient's height H and weight W.

[0069] In one feasible implementation, the inner contour features of the affected area include inner contour slice features at different heights of the patient's affected area. These inner contour slice features at different heights of the patient's affected area are generated based on a three-dimensional model of the patient's affected area.

[0070] In one feasible implementation, the inner contour slice features at different heights of the patient's affected area are generated as follows:

[0071] For a 3D model of the patient's affected area, slice the model along the height direction. Set a corresponding axial control vector for each m slice layer, which is a total of... Axial control vectors Thus, the control vector set is obtained. T represents the transpose operation.

[0072] Each axial control vector is constructed as follows:

[0073] Sampling points are taken from the slice layer corresponding to the current axial control vector. The coordinate data is stored in the current axial control vector. That is, each axial control vector contains all the sampling point data of a single-layer contour, and its point set coordinate data can be represented as:

[0074]

[0075]

[0076]

[0077] Furthermore, each axial control vector can be simplified using Fourier descriptors to... Fourier coefficients of order This allows for the acquisition of a new axial control vector.

[0078] Step 2: Input the basic feature vector into the pre-trained artificial anus parameter prediction model. The model outputs an artificial anus geometric parameter vector. The artificial anus geometric parameter vector includes the cone angle. axial length Major diameter of the distal inner contour (i.e., the length of the semi-major axis of the ellipse) and minor diameter (i.e., the length of the minor semi-axis of the ellipse), the major diameter of the proximal inner contour and trails And the major diameter D and minor diameter d of the contraction zone. Based on this geometric parameter vector of the artificial anus, a basic model of the artificial anus with uniform wall thickness is constructed. Step 2 is referred to as the basic design stage.

[0079] In one feasible implementation, the pre-training process of the artificial anal parameter prediction model is as follows:

[0080] A first training dataset is constructed, containing a large number of data samples covering different anal shapes and design parameters. An artificial anal parameter prediction model is used to establish the mapping between the basic feature vector, the geometric parameter vector of the artificial anus, and the physical displacement field. For each training iteration of the artificial anal parameter prediction model, the basic feature vector of each training sample is input into the model, and the model outputs the corresponding geometric parameter vector of the artificial anus. Based on the current geometric parameter vector, an artificial anal mesh model is generated, and the physical displacement field of the current artificial anal mesh model under abdominal pressure compression is calculated. The physical displacement field represents the displacement of different particles in the artificial anal mesh model under abdominal pressure, including axial displacement. radial displacement and tangential displacement , represented as .

[0081] Then, the strain field is solved based on the current physical displacement field. and stress field Based on the strain field obtained from the solution and stress field Calculate the total physical loss, adjust the parameters of the artificial anal parameter prediction model based on the current total physical loss, and complete one training cycle.

[0082] In one feasible implementation, the total physical loss Satisfy the following formula:

[0083]

[0084]

[0085] in, The residuals of the equilibrium equation; The residuals of the geometric equations; The constitutive equation residuals; For boundary condition residuals; External force refers to the force applied externally to the artificial anal mesh model. This represents the stress field divergence. The gradient of the displacement field; T represents the transpose operation; The strain energy of a micro-element in an elastic material; This represents the desired displacement field of the artificial anal mesh model.

[0086] For a stable structure, the resultant force of all infinitesimal elements is zero, and the internal stress field changes with external forces. The sum should be 0, which means That is, the equilibrium equation, thus constructing the equilibrium equation residuals.

[0087] The geometric equation residuals are used to ensure that the displacement of the artificial anal mesh model corresponds to a reasonable strain, that is, the strain tensor is half of the symmetrical part of the displacement gradient. That is, geometric equations.

[0088] For easily deformable elastic materials, the increment of strain energy of a differential element should be equal to the product of the stress and the increment of strain on the differential element, i.e. That is, the constitutive equation, thus constructing the constitutive equation residual.

[0089] The connection between the artificial anal structure and the human body must not separate during deformation; that is, the boundary condition is... The boundary condition residual function calculates the difference between the displacement field corresponding to the current artificial anal mesh model and the expected displacement field of the artificial anal model.

[0090] The artificial anal parameter prediction model is a physical information machine learning model. Its core advantage lies in embedding the governing equations of solid mechanics into its loss function. Specifically, during training, the model not only requires the output to have the smallest error compared to the sample data (if available), but also requires the output to satisfy the laws of mechanics. By incorporating the residuals of the mechanical equations as regularization terms into the loss function, the model can efficiently explore the design space without relying on a large amount of training data, finding the parameter combination that achieves the most uniform stress distribution, the most reliable fall protection function, and the best tissue adaptability. This process can be completed automatically, eliminating the need for repeated manual trial and error and significantly shortening the customization cycle.

[0091] Step 3: Input the internal contour features of the affected area into the pre-trained artificial anus internal contour parameter prediction model. The model outputs the in-situ internal contour features, which, combined with the basic model of uniform wall thickness of the artificial anus, generate a model of variable wall thickness of the artificial anus. Step 3 is designated as the personalized design stage.

[0092] In one feasible implementation, the pre-training process of the artificial anal contour parameter prediction model is as follows:

[0093] A second training dataset is constructed, containing a large number of training samples consisting of a basic model of uniform wall thickness of the artificial anus, internal contour features of the affected area, and in-situ internal contour features. For each training iteration of the artificial anus internal contour parameter prediction model, the internal contour features of the affected area of ​​each training sample are first input into the artificial anus internal contour parameter prediction model, and the model outputs the corresponding artificial anus internal contour parameter vector. Then, a reconstructed internal contour is generated based on this artificial anus internal contour parameter vector. After replacing the internal contour in the basic model of uniform wall thickness of the artificial anus with the reconstructed internal contour, a model of variable wall thickness of the artificial anus is obtained. The physical displacement field of the current model of variable wall thickness of the artificial anus under abdominal pressure compression is calculated. Then, the strain field is solved based on the current physical displacement field. and stress field Based on the strain field obtained from the solution and stress field Calculate the total physical loss, and adjust the parameters of the artificial anal contour parameter prediction model based on the current total physical loss to complete one training cycle. The total physical loss in the training of the artificial anal contour parameter prediction model is the same as the total physical loss in the training of the artificial anal parameter prediction model.

[0094] In one feasible implementation, the construction process of the first training dataset is as follows:

[0095] First, a parameter space composed of different basic feature parameters is constructed, and as much patient information as possible is selected to ensure that the sampling covers the parameter space. For patients with different basic feature vectors, the displacement field of the uniform wall thickness basic model of the artificial anus is calculated under abdominal pressure. This allows for the design of a suitable uniform wall thickness basic model for each patient, which is denoted as the standard model. The patient's basic feature vector and the corresponding geometric parameter vector of the standard model are used as the first training dataset.

[0096] In one feasible implementation, the construction process of the second training dataset is as follows:

[0097] Three-dimensional scans were performed on the physiological structures near the anus after the anus was removed and in situ, thereby constructing corresponding three-dimensional models. Figure 7Sagittal magnetic resonance imaging of the rectum in a patient's anorectal region was performed with a slice thickness of 3.5000 mm. Each pixel can store 4096 different gray levels (intensity values), and the resolution can be as fine as 0.325 mm. Through the neighborhood topology 3D reconstruction technology of medical images, the hierarchical structure of the rectal wall and perianal region (mucosa, submucosa, muscle layer) can be obtained. The near-circular shape of the in situ anal canal and the non-uniform wall thickness distribution characteristics in different directions can be obtained. A tomographic scan slice database of the in situ artificial anus 3D model is constructed for training a tomographic scan fusion network with embedded physical constraints.

[0098] For the patient's 3D model, along the axial direction (i.e., the height direction), at the same height, tomographic scan slices of the 3D model of the physiological structures near the anus were taken at equal angles around the axis, both after anal resection and in situ. The coordinates of the contour edges of the slices were recorded as axial control vectors. Then, Fourier transform was performed on the coordinates recorded in each axial control vector to obtain Fourier descriptors for describing the contour, thus obtaining the final axial control vector. Finally, the axial control vectors of different slice layers in the 3D model of the physiological structures near the anus after anal resection and in situ were summarized to obtain the internal contour features of the affected area and the internal contour features of the in situ, which were then combined as a second dataset.

[0099] This invention utilizes a tomographic fusion network-based custom design system based on a physical information machine learning model. This system enables intelligent and automated customization of the entire process from patient images to the finished implant, greatly improving the individual adaptability and clinical efficacy of artificial anus.

[0100] Step 4: Design the corresponding anti-fall plug based on the artificial anal wall thickness model.

[0101] In one feasible implementation, the maximum outer diameter of the anti-fall plug and minimum outer diameter Satisfy the following formula:

[0102]

[0103]

[0104] Where d is the small diameter of the contraction zone in the artificial anal wall thickness model.

[0105] The interference fit design ensures that after the anal plug is inserted, the spherical contraction area tightly covers the equatorial region of the anal plug due to elastic deformation, forming a uniform circumferential contact pressure. This is sufficient to resist the dislodgement force caused by gravity and abdominal pressure, without causing tissue ischemia or pain.

[0106] For a given patient, the basic model parameters for the uniform wall thickness of the artificial anus are: cone angle. The angle is 100°, the axial length is 50mm, the distal inner diameter is 116mm, and the proximal inner diameter is 46mm. For example... Figure 9 As shown, with model iteration, the physical loss corresponding to the inner wall points decreases. The model with uniform wall thickness and the model with variable wall thickness of the artificial anus are compared. The dimensional data collected from the model are compared with the patient's original anal structure using 100 horizontal sections. Eleven sections are taken and the comparison results of these eleven sections are plotted. The results are as follows... Figure 9 , Figure 10 and Figure 11 As shown in Table 1, for slices of the same height, the area enclosed by their inner walls was calculated, and a sampling point was taken every 1° around the circumference, for a total of 360 coordinate points. The Euclidean distances from the center were compared, and the results are shown in Table 1.

[0107] Table 1. Area enclosed by sampling points on the inner wall of each layer ( )

[0108]

[0109] The statistical results of the data deviation between different artificial anal models and the original anal structure are shown in Tables 2, 3 and 4.

[0110] For the area enclosed by the inner wall, the variance of the data for the basic model of uniform wall thickness of the artificial anus is 7.28 × 10⁻⁶. 3 The variance of the artificial anal wall thickness model is 2.34 × 10⁻⁶. 3 . Figure 10 The changes in the center distance of sampling points and the changes in the area of ​​different layers were recorded in both the variable wall thickness model and the uniform wall thickness model of the artificial anus. For example... Figure 10 As shown in (a), the data fluctuation of the artificial anus variable wall thickness model is significantly greater than that of the artificial anus uniform wall thickness basic model, such as Figure 10 (b) shows that the area change curves between the sampling layers in the artificial anal variable wall thickness model are closer to a straight line.

[0111] Table 2. Statistical analysis of the deviation between the model and the in situ anal structure data.

[0112]

[0113] Table 3. Statistical analysis of distance deviation for different models (mm)

[0114]

[0115] Data was also collected from the three-dimensional anal anatomy of the patient in situ, and the differences in data obtained from the same sampling points in the same layer were calculated. The average distance deviation of the basic model scheme with uniform wall thickness of artificial anus was 1.0666 mm, and the average distance deviation of the model scheme with variable wall thickness of artificial anus was 0.4596 mm, with a distance deviation optimization rate of 56.91%. Figure 11 The distribution of the mean and standard deviation of distance deviation at the same sampling angle in different layers of the artificial anal variable wall thickness model and the artificial anal uniform wall thickness basic model was recorded, as well as the distribution of the mean and standard deviation of distance deviation at different sampling points within the same layer as the number of layers. For example... Figure 11 As shown in (a), the standard deviation band of the variable wall thickness model of the artificial anus at different sampling angles is significantly narrower than that of the basic model with uniform wall thickness of the artificial anus, indicating that its structure is closer to the patient's original situ size data and can better replace the function of the original artificial anus. Figure 11 As shown in (b), the mean standard of the distance deviation of the artificial anal variable wall thickness model at different sampling layers is significantly smaller than that of the artificial anal uniform wall thickness basic model, and the data distribution is also more concentrated.

[0116] Table 4 Statistical analysis of area deviation for different models (mm) 2 )

[0117]

[0118] Figure 12 The area deviations, mean values, and standard deviations of different sampling layers in both the variable wall thickness model and the uniform wall thickness model of the artificial anus were recorded. The average area deviation of the uniform wall thickness model of the artificial anus was 54.9435 mm. 2 The average area deviation of the artificial anal wall thickness model was 7.6566 mm. 2 The area deviation optimization rate was 86.06%. The area difference values ​​of the artificial anal variable wall thickness model were significantly smaller and more concentrated.

[0119] Based on clinical anthropometric data, the key dimensions of the fall arrestor are designed as follows:

[0120] The maximum outer diameter is 40mm, the axial length is 8mm, and the spherical cap height is 45mm. The maximum outer diameter D1 refers to the diameter of the anal plug's equatorial surface (i.e., the largest cross-section perpendicular to the axis); the spherical cap height refers to the height of the curved surface from the equatorial surface to the apex, used to control the insertion depth and contact area. For example... Figure 13 As shown, different sizes of anti-fall plugs are used to match different artificial anuses. The maximum outer diameter of the anti-fall plugs varies in the six sets of examples, and the actual dimensions are shown in Table 5.

[0121] Table 5 Figure 13 Maximum and minimum outer diameters of different sizes of anti-fall plugs

[0122]

[0123] During use, the anal plug is slowly pushed in along the axis of the artificial anus. When its equatorial plane passes through the spherical contraction zone, the contraction zone temporarily yields due to elastic expansion. Once the anal plug is fully inserted, the contraction zone rebounds, "locking" it in place on the proximal inner side. At this point, the spherical crown of the anal plug is located inside the artificial anus, while the structure with the largest outer diameter at the tail is exposed outside the body, facilitating operation.

[0124] The above specific embodiments are used to explain and illustrate the present invention, but not to limit the present invention. Any modifications and changes made to the present invention within the spirit and scope of the claims shall fall within the protection scope of the present invention.

Claims

1. A design method for a variable-wall-thickness in-situ artificial anus and anti-fall plug, characterized in that, The artificial anus includes a connected tissue portion and an anal plug portion. The tissue portion includes an inner layer (2) and an outer layer (3). One end of the inner layer (2) and the outer layer (3) are connected to the anal plug portion, and the other end of the inner layer (2) and the outer layer (3) is used to connect to the patient's tissue. The cavity between the inner layer (2) and the outer layer (3) is called the buffer cavity (4). The internal cavity of the anal plug portion is called the spherical contraction area (5). The spherical contraction area (5) is used to place the anti-fall plug. The inner layer (2) is connected to the spherical contraction area (5). Includes the following steps: Step 1: Obtain the patient's human feature parameters and construct the basic feature vector and the inner contour features of the affected area; Step 2: Input the basic feature vector into the pre-trained artificial anal parameter prediction model. The model outputs the artificial anal geometric parameter vector. Based on the artificial anal geometric parameter vector, construct a basic model of uniform wall thickness for the artificial anal. The pre-training process of the artificial anal parameter prediction model is as follows: For each training iteration of the artificial anal parameter prediction model, the basic feature vector of each training sample is input into the model, and the model outputs the corresponding artificial anal geometric parameter vector. Based on the current artificial anal geometric parameter vector, an artificial anal mesh model is generated, and the physical displacement field of the current artificial anal mesh model under abdominal pressure compression is calculated. Then, the stress field is solved based on the current physical displacement field. and strain field Based on the stress field obtained from the solution and strain field Calculate the total physical loss, adjust the parameters of the artificial anal parameter prediction model based on the current total physical loss, and complete one training cycle; The total physical loss Satisfy the following formula: in, The residuals of the equilibrium equation; The residuals of the geometric equations; The constitutive equation residuals; For boundary condition residuals; External force; This represents the stress field divergence. The gradient of the displacement field; T represents the transpose operation; The strain energy of a micro-element in an elastic material; This represents the desired displacement field of the artificial anal mesh model; Step 3: Input the inner contour features of the affected area into the pre-trained artificial anal inner contour parameter prediction model. The model outputs the in-situ inner contour features. Combined with the basic model of uniform wall thickness of the artificial anus, the model of variable wall thickness of the artificial anus is generated. Step 4: Design the corresponding anti-fall plug based on the artificial anal wall thickness model; In step 4, the maximum outer diameter D1 of the anti-fall plug satisfies the following formula: D1 = (1.03~1.08) × d Where d is the small diameter of the contraction zone in the artificial anal wall thickness model.

2. The design method of a variable wall thickness in-situ artificial anus and anti-fall plug according to claim 1, characterized in that, The pre-training process of the artificial anal contour parameter prediction model is as follows: For each training iteration of the artificial anal internal contour parameter prediction model, the internal contour features of the affected area for each training sample are first input into the model, and the model outputs the corresponding artificial anal internal contour parameter vector. Then, a reconstructed internal contour is generated based on this parameter vector. Next, the internal contour in the uniform wall thickness model of the artificial anus is replaced with the reconstructed internal contour to obtain a variable wall thickness model. Finally, the physical displacement field of the current variable wall thickness model under abdominal pressure compression is calculated. Then, the stress field is solved based on the current physical displacement field. and strain field Based on the stress field obtained from the solution and strain field Calculate the total physical loss, and adjust the parameters of the artificial anal contour parameter prediction model based on the current total physical loss to complete one training cycle.

3. The design method of a variable wall thickness in-situ artificial anus and anti-fall plug according to claim 1, characterized in that, The anti-fall plug includes a connected insertion structure (6) and a flange structure (7).

4. A design system for a variable wall thickness in-situ artificial anus and anti-fall plug for implementing the method of claim 1, characterized in that, include: The human body parameter acquisition unit is used to acquire the human body characteristic parameters of the subject. The feature construction unit is used to construct basic feature vectors and inner contour features of the affected area based on the acquired human body feature parameters of the patient. The first prediction unit is used to input the basic feature vector into the pre-trained artificial anal parameter prediction model. The model outputs the artificial anal geometric parameter vector, and constructs a basic model of uniform wall thickness of the artificial anal based on the artificial anal geometric parameter vector. The second prediction unit is used to input the inner contour features of the affected area into the pre-trained artificial anal inner contour parameter prediction model. The model outputs the in-situ inner contour features, which, combined with the basic model of uniform wall thickness of the artificial anus, generate the artificial anal variable wall thickness model. An anal plug design unit is used to design corresponding anti-fall plugs based on the artificial anal wall thickness model.

5. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the design method for a variable wall thickness in-situ artificial anus and anti-fall plug as described in any one of claims 1 to 3.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the design method of a variable wall thickness in-situ artificial anus and anti-fall plug as described in any one of claims 1 to 3.

Images