Image segmentation method and system

Abstract

The invention relates to an image segmentation method and system. The method comprises the following steps: initially positioning medical images to obtain a positioning area; preprocessing the positioning area to obtain a target area, wherein the target area contains a nodular area and a background area, and the nodular area consists of a solid area and a surrounding area which surrounds the solid area; processing the target area in combination with a Gaussian mixture model to obtain a probability graph of the target area; processing the probability graph according to a morphological model to determine the solid area and the surrounding area of a nodule; and fusing the solid area and the surrounding area to obtain an image segmentation result. According to the method and system provided by the invention, different types of nodules can be accurately segmented, and the follow-up diagnostic analysis can be effectively improved.

Description

[0001] 【Technical Field】 [0002] The present invention relates to the field of medical image processing, in particular to an image segmentation method and system. [0003] 【Background technique】 [0004] Lung cancer is the cancer with the highest mortality rate in the world. Although the level of medical diagnosis and treatment continues to improve, the five-year survival rate of lung cancer is only about 15%. Early detection and early treatment are the main ways to improve the cure rate of lung cancer. Lung nodules are an early manifestation of lung cancer. The accurate segmentation results of lung nodules can effectively reflect the pathological and morphological characteristics of the nodules, thereby helping users to diagnose and analyze the lesions. The key technology to realize the automatic analysis and recognition of lung nodules is the research and application of a series of image processing, analysis and understanding algorithms such as lung nodule segmentation, detectio...

AI Technical Summary

Problems solved by technology

[0005] In the prior art, the level set algorithm or multi-scale threshold method is used for nodule segmentation, but pulmonary nodules have different shapes, sizes, distribution locations, and are easily linked with other tissues, and their density is similar to that of some lung tissues. For example, pulmonary nodules have various forms such as solid nodules, mixed ground-glass nodules, and ground-glass nodules (GGN), and it is impossible to accurately identify pulmonary nodules, especially ground-glass nodules, simply by morphological methods. As a type of nodule with the highest possibility of malignancy, nodules have irregular edges in shape, and present fuzzy and thin shadows in CT images. The HU (Hounsfield Unit, Hounsfield Unit) value in CT is widely distributed and the texture morphology is variable. It is difficult to accurately identify pulmonary nodules by edge-based algorithms such as Level Set

In addition, the gray value of pulmonary nodules in the CT image does not obey the Gaussian distribution, and the multi-threshold method is easy to cause leakage and result in inaccurate segmentation results.

Images