Dynamic ultrasonic breast nodule real-time segmentation and recognition method based on deep learning

Abstract

The invention relates to the technical field of medical image processing, and aims to provide a dynamic ultrasonic breast nodule real-time segmentation and recognition method based on deep learning. The method comprises the following steps: collecting ultrasonic mammary gland images and videos with nodules and case data with operative pathology results, constructing a data set, constructing a static image nodule segmentation network, and training the static image nodule segmentation model on an original image; predicting an intermediate frame nodule probability by using an LSTM layer, constructing a video dynamic segmentation network, and training a dynamic segmentation model; constructing a benign and malignant identification network structure by using a basic network, and training a benign and malignant identification model; and outputting nodule position information in real time, using the benign and malignant recognition model for recognizing benign and malignant nodules of each frame, and outputting the number of output nodules and the comprehensive benign and malignant probability after examination is finished. Information incompleteness of a single image can be avoided, error detection is reduced, missing small nodules are reduced, and the nodule benign and malignant identification accuracy is improved.

Description

technical field [0001] The invention relates to the technical field of medical image processing, in particular to a method for real-time segmentation of dynamic ultrasound breast nodules and identification of benign and malignant nodules based on deep learning. Background technique [0002] The incidence of breast cancer is high worldwide, and early detection of breast cancer can greatly improve the survival rate and quality of life, and improve the prognosis. Non-invasive auxiliary examination methods commonly used in breast cancer screening mainly include mammography, ultrasound and magnetic resonance. Among the three examination methods, mammography is cheap and mainly uses X-ray film, which has the most advantage in finding tiny calcifications, and then can find asymptomatic or impalpable tumors, and its diagnostic efficiency is even higher than that of MRI. The advantage of ultrasound is that there is no radioactivity, convenient inspection, low price, and repeated ins...

AI Technical Summary

Problems solved by technology

From the practice of video segmentation tasks, it can be seen that LSTM achieves better results than traditional time series models, but the basic network for practical application of LSTM is very shallow. If there are more training samples, a deeper network needs to be designed to improve model performance.

On the other hand, due to the black-box nature of deep learning, a large number of experiments are required to adjust parameters, such as the number of LSTM network layers, sequence length, training batch size, etc. These parameters may have a great impact on the results. Lack of effective theoretical guidance

Images