Mammary gland focus positioning method and system

Abstract

The invention provides a mammary gland focus positioning method and device. The mammary gland focus positioning method splits the video sequences mainly based on mammary gland ultrasonic video sequence data, so as to process the video sequences into continuous image frame sequences, thus compared with a traditional ultrasonic image, increasing more information amount, and avoiding the condition that the sample is too good due to manual selection. Besides, the mammary gland focus positioning method adopts a deep learning method to train a model, so as to realize end-to-end positioning of test set data while the breast lesion area can be automatically detected, and then information of a subject is utilized to the maximum extent, and an effective auxiliary effect is provided for diagnosis ofdoctors, and the workload of the doctors is reduced.

Description

technical field [0001] The invention relates to the technical field of ultrasonic examination, in particular to a method and system for locating breast lesions. Background technique [0002] With the rapid development of modern society, women's work and life pressure is becoming more and more serious, and the incidence of breast cancer is also increasing. According to statistics, breast cancer is still the most common malignant tumor in women. In recent years, the incidence of breast cancer in my country has shown a rapid upward trend. Among all kinds of malignant tumors, breast cancer is one with better curative effect, but the overall survival rate is still unsatisfactory. Early detection and early treatment are better. [0003] Currently, breast cancer screening mainly includes clinical breast examination, breast B-ultrasound and mammography X-ray examination. Breast clinical examination is convenient, cheap, and non-invasive, but it is easy to miss or misdiagnose, and ...

AI Technical Summary

Problems solved by technology

Traditional target detection is mainly based on the sliding window method, but the region selection strategy based on the sliding window is not targeted, the time complexity is high, the window is redundant, and the manually designed features are not very robust to changes in diversity.

For some specific target detection, such as face detection and pedestrian detection, there are already very mature technologies. For medical images, the background noise is complex and there are only grayscale images, making it difficult for traditional target detection methods to obtain better results. Effect

Images