CN-121971067-A - Talus health monitoring and recognition system integrating image features and biomechanical parameters

CN121971067ACN 121971067 ACN121971067 ACN 121971067ACN-121971067-A

Abstract

The application relates to the technical field of auxiliary diagnosis and discloses a talus health monitoring and identifying system integrating image features and biomechanical parameters, which comprises an image processing module, a grid construction module, a finite element solving module, an evaluation output module and a local tissue vulnerability index, wherein the image processing module is used for extracting a talus cartilage geometric mask and reconstructing boundary signals and outputting a pure biochemical voxel set, the grid construction module is used for generating an individual calculation grid by utilizing non-rigid registration and transmitting an anatomical partition label, the construction mapping module is used for constructing a non-uniform dynamic viscoelastic stiffness matrix based on biochemical values and partition strain rates, the finite element solving module is used for calculating stress distribution fields under physiological loads, and the evaluation output module is used for determining local yield strength thresholds according to the biochemical values and comparing the local tissue vulnerability indexes with the stress fields. The application realizes the deep coupling analysis of cartilage biochemical degeneration and mechanical response, can accurately identify the recessive high risk area caused by tissue degeneration, and provides reliable quantitative basis for talus health monitoring.

Inventors

HUANG JIE
ZHANG YANJING
LIU MIN
Wei Zhouhao
DUAN ZHIQING
LIU JIE
CHEN YONG
ZHAO JIAN
Yao Wanzhen
ZHOU JING
DING JIANPING
DAI SIYU
PAN SHINONG

Assignees

杭州师范大学附属医院(杭州市第二人民医院)

Dates

Publication Date: 20260505
Application Date: 20260126

Claims (10)

1. A talus health monitoring and recognition system that fuses image features with biomechanical parameters, comprising: the image processing module is used for extracting a geometric mask of talus cartilage based on multi-mode magnetic resonance imaging data of the subject and outputting a pure biochemical voxel set reconstructed by the boundary signals; The grid construction module is used for mapping a standard reference finite element model carrying an anatomical partition label to a subject space based on the geometric mask, and generating an individualized calculation grid containing the anatomical partition label; The system global stiffness matrix comprises dynamic viscoelasticity characteristics according to T2 values in the pure biochemical voxel set and strain rate characteristics corresponding to the anatomical partition labels; The finite element solving module is used for solving a balance equation under a preset physiological load based on the system global stiffness matrix to obtain an equivalent von mises stress distribution field of the talus cartilage of the subject; An evaluation output module configured to determine a local yield strength threshold from the T2 values and calculate a local tissue vulnerability index at voxel level by comparing the local yield strength threshold with the equivalent von mises stress distribution field.
2. The talus health monitoring and recognition system of claim 1, wherein the image processing module is configured to process the geometric mask with a three-dimensional morphological erosion operator to strip outermost voxels, generate a core region mask, and define a boundary region mask by boolean differential operation; The image processing module is further configured to search for spatially adjacent reference voxels in the core region mask for voxels in the boundary region mask, and reconstruct biochemical values of the boundary voxels using an inverse distance weighted interpolation algorithm to eliminate signal noise generated by partial volume effects, and generate the clean biochemical voxel set.
3. The talus health monitoring and recognition system of claim 1, wherein the mesh construction module is configured to perform non-rigid registration using a B-spline based free-form deformation algorithm, driving the control lattice of the standard reference finite element model to deform by minimizing an objective function comprising a data matching term and a smoothing regularization term; The grid construction module is configured to communicate label attributes of six anatomical partitions of the lateral anterior, lateral medial, lateral posterior, medial anterior, medial, and medial posterior in the standard reference finite element model to the personalized computing grid based on an index invariance principle of the finite element grid.
4. The talus health monitoring and recognition system of claim 1, wherein the constitutive mapping module is configured to build a spatial mapping of the pure set of biochemical voxels and the personalized computational grid using a k-d tree spatial index structure; The constitutive mapping module is configured to calculate a static elastic modulus by adopting a negative exponential decay model, wherein the static elastic modulus monotonically decreases along with the increase of the T2 value, and the calculation formula comprises a rigidity scaling coefficient, a decay rate constant and a minimum limit modulus.
5. The talus health monitoring and recognition system of claim 4, wherein the constitutive mapping module is configured to perform dynamic viscoelasticity compensation, and specifically comprises obtaining a characteristic strain rate of an anatomical partition to which a current grid cell belongs in a preset characteristic strain rate lookup table, calculating a dynamic elastic modulus according to the static elastic modulus, the characteristic strain rate and a viscoelasticity hardening coefficient by using a logarithmic hardening model, and assembling the system global stiffness matrix.
6. The talus health monitoring and recognition system of claim 1, wherein the finite element solution module is configured to screen mesh nodes less than a set threshold from subchondral bone surface as bone-cartilage interfaces and impose full-fixation constraints, and establish virtual reference points at joint interfaces, applying external physiological loads to the nodes of the joint interfaces by distributing coupling constraints; The finite element solution module is configured to solve the system balance equation including geometric nonlinearity and material nonlinearity using a newton-raphson iteration method.
7. The talus health monitoring and recognition system of claim 1, wherein the assessment output module is configured to calculate the local yield strength threshold using a linear injury decay model; The local yield strength threshold is inversely related to the T2 value, and the calculated parameters comprise theoretical ultimate compressive strength of healthy cartilage, a reference T2 lower limit value of healthy cartilage, a T2 upper limit reference value of severely degenerated cartilage and a strength attenuation factor.
8. The talus health monitoring and recognition system of fused image features and biomechanical parameters of claim 1, wherein the local tissue vulnerability index is defined as a ratio of equivalent von mises stress of the grid cells to the local yield strength threshold; The evaluation output module is further configured to count high risk units within each anatomical partition having a local tissue vulnerability index greater than or equal to 1.0 and calculate a regional risk severity index for each anatomical partition.
9. The talus health monitoring and recognition system of claim 8, wherein said regional risk severity index is a generalized injury energy density calculated based on a nonlinear weighted calculation of the total volume of anatomical partitions and the relative magnitude of said high risk units exceeding said local yield strength threshold; The calculation formula of the regional risk severity index comprises a damage sensitivity index, wherein the damage sensitivity index is used for amplifying the weight of a high vulnerability unit.
10. The talus health monitoring and recognition system of claim 8, wherein the assessment output module is configured to generate a six-axis radar map comprising six dimensions, each dimension of the six-axis radar map exhibiting the regional risk severity index for each anatomical partition normalized and displaying superimposed health reference range shadows based on health crowd database statistics.

Description

Talus health monitoring and recognition system integrating image features and biomechanical parameters Technical Field The invention relates to the technical field of auxiliary diagnosis, in particular to a talus health monitoring and identifying system for fusing image features and biomechanical parameters. Background The talus serves as a core hub for ankle weight transfer, and the integrity of its articular cartilage is critical to maintaining lower limb motor function. However, cartilage damage to the talus is often irreversible and can be very prone to develop traumatic arthritis if not found and intervened in time at an early stage. Whereas current clinical diagnostics rely primarily on Magnetic Resonance Imaging (MRI) or Computed Tomography (CT) to observe morphological changes in cartilage. Although quantitative magnetic resonance sequences such as T2 mapping can reflect changes in biochemical components (such as collagen arrangement and water content) inside cartilage, these biochemical indexes are usually only analyzed as independent parameters, and it is difficult to directly determine whether a specific region is at the edge of structural failure under real exercise load. On the other hand, finite element analysis techniques are widely used to evaluate the mechanical response of joints. However, existing conventional modeling methods often have idealized assumptions that generally treat cartilage as a homogeneous, isotropic linear elastic material and impart uniform material properties. This treatment ignores the non-uniformity of the cartilage tissue in terms of spatial distribution and fails to take into account the pathological fact that biochemical degeneration of the tissue leads to a significant decrease in local mechanical properties (such as modulus of elasticity and yield strength). In addition, in the process of constructing a model based on images, due to the limitation of imaging resolution, partial volume effect often exists on the cartilage surface, so that boundary signals are blurred, and if modeling is directly performed by using original data, larger boundary errors are introduced. More importantly, the prior art generally uses a fixed stress threshold based on health demographics as a criterion when assessing cartilage health risk. This static evaluation criterion ignores the tolerance differences of the tissue itself, i.e. the limit load that it can withstand has been greatly reduced for areas where biochemical degeneration has occurred. Thus, even though the calculated stress values for certain regions are within conventional safety limits, the local yield limit of the degenerated tissue may actually have been exceeded. Due to the lack of a comprehensive evaluation mechanism for fusing the biochemical characteristics of images, dynamic mechanical response and local tolerance depth of tissues, the existing monitoring means can not accurately identify hidden risk areas with normal stress levels, which are endangered by fragile tissues. Disclosure of Invention Aiming at the defects of the prior art, the invention provides a talus health monitoring and identifying system fusing image characteristics and biomechanical parameters, which solves the technical problems that the cartilage mechanical bearing capacity cannot be directly quantified by simply relying on image biochemical indexes, the simulation precision is insufficient due to neglect of tissue biochemical non-uniformity and boundary partial volume effect in conventional finite element modeling, and the hidden damage risk of reduced tolerance caused by tissue degeneration is difficult to identify by adopting a fixed stress threshold evaluation mode in the prior art. In order to achieve the above object, the invention is realized by the following technical scheme that the talus health monitoring and identifying system for fusing image characteristics and biomechanical parameters comprises: the image processing module is used for extracting a geometric mask of talus cartilage based on multi-mode magnetic resonance imaging data of the subject and outputting a pure biochemical voxel set reconstructed by the boundary signals; The grid construction module is used for mapping a standard reference finite element model carrying an anatomical partition label to a subject space based on the geometric mask, and generating an individualized calculation grid containing the anatomical partition label; The system global stiffness matrix comprises dynamic viscoelasticity characteristics according to T2 values in the pure biochemical voxel set and strain rate characteristics corresponding to the anatomical partition labels; The finite element solving module is used for solving a balance equation under a preset physiological load based on the system global stiffness matrix to obtain an equivalent von mises stress distribution field of the talus cartilage of the subject; An evaluation output module configured to determine a local yield stren