Thyroid ultrasound image nodule analysis method based on deep learning network and shallow texture feature fusion

Abstract

The invention relates to a thyroid ultrasound image nodule analysis method based on a deep learning network and shallow texture feature fusion, and belongs to the field of computer thyroid ultrasoundimage auxiliary analysis. The method comprises the following steps: preprocessing a thyroid ultrasound image; performing deep feature extraction based on a GooLeNet deep learning network; shallow texture feature extraction based on a rotation invariance local binary pattern; fusing the depth features and the shallow features; and classifying thyroid ultrasound images based on a cost-sensitive random forest. The thyroid nodule benign and malignant classification method can rapidly and accurately classify thyroid nodules by establishing the thyroid ultrasound image nodule analysis method.

Description

technical field [0001] The invention relates to a thyroid ultrasound image nodule analysis method based on deep learning network and shallow texture feature fusion, and belongs to the field of computer-assisted analysis of thyroid ultrasound images. Background technique [0002] In thyroid ultrasound images, the analysis of benign and malignant thyroid nodules is of great significance for the early diagnosis of thyroid cancer. With the development of medical imaging, most of the early thyroid nodules can be accurately detected in ultrasound images, but there is still a lack of accurate judgment on the nature of the nodules. Because most nodules are benign or inert, accurate judgment of benign and malignant nodules can reduce patients' medical risks and a large number of medical and health costs caused by acupuncture testing. Therefore, how to achieve a more accurate early diagnosis of benign and malignant thyroid nodules by ultrasonic images, avoid unnecessary acupuncture o...

AI Technical Summary

Problems solved by technology

However, the traditional method requires manual design of complex feature extraction methods, the extracted feature dimensions are high, the workload is large and the efficiency is low

Since ordinary networks are not sensitive to the features of ultrasound images of thyroid nodules, it is difficult to extract features that can represent the properties of thyroid nodules with ordinary networks, and the deep features extracted by networks are not enough to accurately describe the properties of thyroid nodules

Images