Abstract

Info

Title:: Improving the diagnostic efficacy of breast cancer by integrating breast ultrasound and magnetic resonance imaging with multimodal fusion technology

Author(s):: DAI Nina; ZHANG Wenjun; Department of Ultrasound Medicine, Taihe Hospital Affiliated to Hubei University of Medicine, Shiyan 442000, China

Keywords:: Keywords: breast cancer multimodal fusion breast ultrasound magnetic resonance imaging

Abstract:: Abstract: Objective To propose a multimodal data fusion method based on the deep residual network ResNet-101 for enhancing the diagnostic efficiency of breast ultrasound combined with magnetic resonance imaging (MRI) in breast cancer based on multimodal fusion technology, and further evaluate the performance of various models under multimodal data fusion. Methods The breast ultrasound and MRI images were collected from 154 patients for data fusion and analysis. Multimodal fusion was performed using the ultrasound and MRI data, and the deep residual network models were utilized for classification and diagnosis. Finally, the diagnostic performance of each model was compared. Results The ResNet-101 method demonstrated the best performance in breast cancer diagnosis using multimodal fusion data, which was significantly superior to that of single-modal imaging and other models. Compared with traditional convolutional neural network and ResNet50 model, the proposed method significantly improved diagnostic accuracy, sensitivity, and specificity. Conclusion Multimodal fusion technology can effectively enhance the diagnostic efficiency of breast cancer. Specifically, the deep residual network ResNet-101 model significantly improves the accuracy and robustness of diagnosis, indicating that multimodal data fusion technology has important application prospects in breast cancer diagnosis.