A limb volume correction method based on one-dimensional convolutional neural network
By using a limb volume correction method based on a one-dimensional convolutional neural network, the problem of volume measurement error in infrared grating measurement technology was solved, enabling accurate diagnosis of lymphedema and improving the accuracy of volume measurement and early detection capability.
Patent Information
- Application Number
- CN202610231208.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2026-02-27
- Publication Date
- 2026-07-24
- Estimated Expiration
- 2046-02-27
AI Technical Summary
Existing infrared grating measurement technology suffers from systematic errors in volume measurement due to sensor physical blind spots and geometric simplification algorithms. This makes it difficult to ensure both the accuracy of absolute volume measurement and the identification of local morphological changes while maintaining convenience, thus failing to effectively support the accurate diagnosis of early lymphedema.
A limb volume correction method based on a one-dimensional convolutional neural network is adopted. By constructing a dynamic feature matrix and training dataset, the original length and width data sequences are collected using an infrared measurement grating. A one-dimensional convolutional neural network regression model is trained to correct the initial measured volume data. The volume correction is then performed by combining the elliptical cylinder accumulation algorithm.
It achieves error compensation for sensor physical blind spots and geometric simplification algorithms, improves the accuracy of volume measurement and the ability to identify local morphological changes, reduces the risk of false positive diagnosis, and enhances the detection sensitivity and diagnostic reliability of early lymphedema.