A device, system, and image processing method for real-time detection and segmentation of whole lesions in dynamic breast ultrasound video based on artificial intelligence

Abstract

The invention discloses an artificial intelligence-based system for real-time detection and segmentation of whole lesions in dynamic breast ultrasound video. The system includes at least one ultrasound machine and an AI server. Wherein the AI ​​server is provided with an AI-based dynamic breast ultrasound video detection and segmentation device, which at least includes (1) system robustness design module, (2) data preprocessing module, (3) data amplification module, (4) ) a lesion detection module, and (5) a lesion segmentation module. The system of the present invention can realize automatic real-time detection of all lesions involved in dynamic breast ultrasound video images without changing the ultrasound machine and the existing diagnostic process, and intelligently segment and measure the detected lesions, improving efficiency At the same time as the accuracy rate, it can effectively help doctors reduce missed diagnoses.

Description

technical field [0001] The invention relates to the field of medical images, in particular to an artificial intelligence (AI)-based device, system and image processing method for real-time detection and segmentation of whole lesions in dynamic breast ultrasound video. Background technique [0002] Breast cancer is a malignant tumor of the breast. According to the data released by the National Cancer Center, breast cancer ranks first in the incidence of female malignant tumors in my country and seriously threatens women's health. Professor Hillman of the University of Virginia in the United States wrote an article on "NEngl J Med" in 2010 and pointed out that early accurate diagnosis can increase the 5-year survival rate of breast cancer patients from 25% to 99%. [0003] Breast ultrasound technology has the advantages of being non-invasive, fast, and highly reproducible. It can clearly display the changes in the shape, internal structure, and adjacent tissues of soft tissues...

AI Technical Summary

Problems solved by technology

[0010] Most of the existing breast ultrasound detection and segmentation studies only focus on tumors, and most of them are based on two-dimensional images. Clinical diagnosis usually requires a comprehensive judgment based on the information of the front and rear frames. Therefore, artificial intelligence based on two-dimensional images is easy to use clinically. In order to effectively solve the current situation that the existing methods have poor clinical usability and serious lack of clinical help, the present invention proposes an AI-based dynamic breast ultrasound video of all lesions in real time. The detection and segmentation device, system and detection method are used to solve the clinical problem of missed diagnosis due to visual fatigue and insufficient visual sensitivity, and to improve the efficiency of doctors' diagnosis

Images