Skull dissection method and system of diffusion weighted image and readable storage medium

By employing the OTSU segmentation method with weighted fusion, filtering, and Gaussian smoothing, combined with a hole-filling algorithm, the robustness and noise sensitivity issues of traditional methods in diffusion-weighted image processing are resolved. This achieves high-precision skull dissection results and reduces the cost of AI-based data collection and computation.

CN120931687APending Publication Date: 2025-11-11宁波慧沣生物科技有限公司
5 Cites 0 Cited by

Patent Information

Application Number
CN202511068524.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-07-31
Publication Date
2025-11-11

AI Technical Summary

Technical Problem

Traditional thresholding methods are not robust to diffusion-weighted images, are difficult to adapt to different imaging conditions, have poor edge preservation, are prone to oversegmentation or undersegmentation, and are sensitive to noise. AI-based segmentation methods are difficult to collect data, have high training and inference costs, and are difficult to deploy quickly.

Method used

We employ weighted fusion, filtering, Gaussian smoothing-based OTSU segmentation, connected component processing, and hole-filling algorithms to process the diffusion-weighted image, generating an isotropic diffusion-weighted fused image. We optimize the segmentation results through adaptive window midpoint filtering and Gaussian smoothing, and improve segmentation accuracy and robustness by combining the hole-filling algorithm.

Benefits of technology

It achieves high-precision and robust skull dissection under different imaging conditions, reduces the impact of noise, improves segmentation results, reduces the requirements for data collection and computing resources, and supports rapid deployment.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention relates to a skull dissection method and system based on diffusion weighted images and a readable storage medium, and the method comprises the steps: respectively collecting the diffusion weighted images of a target object along a plurality of gradient directions, carrying out the weighted fusion of the collected diffusion weighted images, and generating an isotropic diffusion weighted fusion image; performing filtering processing on the diffusion weighted fusion image to obtain a filtered image; carrying out segmentation processing on the filtered image by adopting an OTSU segmentation method based on Gaussian smoothing to obtain a binary input image; performing connected region processing and target extraction on the binary input image to obtain a binary mask image; and performing hole filling on the binary mask image layer by layer by adopting a hole filling algorithm to obtain a final segmented mask image. Compared with a traditional threshold segmentation method and an existing AI-based segmentation method, the skull dissection method provided by the invention has the advantages of high adaptability, high precision, good robustness and the like.
Need to check novelty before this filing date? Find Prior Art