Chest CT image processing system

By employing intensity clipping, smoothing bias field optimization, threshold segmentation, and multi-scale structural enhancement in chest CT image processing, combined with multi-resolution fusion technology, the problems of tissue contrast distortion, inaccurate segmentation, and high resampling overhead in chest CT image processing are solved, achieving efficient image processing and diagnostic support.

CN122289165APending Publication Date: 2026-06-26THE AFFILIATED HOSPITAL OF QINGDAO UNIV
0 Cites 0 Cited by

Patent Information

Application Number
CN202610355802.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-03-23
Publication Date
2026-06-26

AI Technical Summary

Technical Problem

Existing technologies for processing chest CT images suffer from tissue contrast distortion due to inconsistency between the bias field and the lung field size (HU), coarse lung field segmentation leading to missed or misclassified lungs, insufficient ability to distinguish between blood vessels and nodules, high computational and storage overhead for resampling strategies, and artifact problems.

Method used

We employ intensity clipping and linear normalization, smooth bias field optimization, thresholding and connectivity segmentation, nonlocal mean denoising, and multi-scale structure enhancement, combined with multi-scale Laplacian and Gaussian pyramid fusion, to perform multi-resolution resampling and weighted synthesis through voxel-level content importance mapping.

Benefits of technology

It restores the real tissue contrast, improves the accuracy of lung field and lesion boundaries, preserves and magnifies diagnostic details, reduces computational and storage overhead, avoids resolution switching artifacts, and improves diagnostic visibility and processing efficiency.

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

Abstract

The invention discloses a chest CT image processing system. The chest CT image processing system comprises a data acquisition module, an image enhancement module, a multi-resolution fusion module and a fidelity export module. The invention relates to the technical field of CT (Computed Tomography) image intelligent processing, in particular to a chest CT image processing system, which is characterized in that intensity cutting and normalization are carried out on a CT image, and tissue comparison is recovered by combining regularization iteration correction of a logarithm domain bias field; firstly, a lung field is positioned, variational segmentation is adopted to refine a boundary, and mask constraint non-local mean denoising is used in a lung to reserve a small focus; improving Frangi measurement in combination with a Hesi matrix to perform structure enhancement, fusing an original image, denoising and enhancing a result; voxel-level importance mapping is constructed based on structural response and gradient, high and low resolution resampling is carried out, seamless synthesis is carried out according to weight, high resolution and details are reserved in a key area, lung field and focus boundary precision is improved, calculation and storage overhead is reduced, and artifacts are effectively avoided.
Need to check novelty before this filing date? Find Prior Art