A Single-arm, Prospective, Multi-center Cohort Study Based on Deep Learning-based cfDNA Fragment Omics to Verify the TuFEst Model for the Staging Diagnosis of Breast Cancer Lesions and Lymph Nodes

Overview

Through the research of this project, we expect to achieve the cfDNA fragment omics liquid biopsy technology based on deep learning, verify the accuracy of the TuFEst model in predicting the tumor burden status of breast cancer lesions and lymph nodes in newly diagnosed breast cancer patients and those receiving neoadjuvant therapy, and provide a theoretical basis for large-scale clinical application in the future

1. Based on the previously established TuFEst model, the cfDNA fragment omics liquid biopsy technology based on deep learning is utilized to predict the tumor burden status of breast cancer lesions and lymph nodes, thereby enhancing the accuracy of early diagnosis of breast cancer: This study will collect and analyze blood samples from breast cancer patients at different stages, and use deep learning-based cfDNA fragment omics liquid biopsy technology to extract tumor-related cfDNA fragments and construct a cfDNA fragment omics feature library. Predictions are made based on the TuFEst model. Then, accuracy matching and evaluation are carried out according to the prediction results and the actual breast cancer lesion and lymph node tumor burden status. Verify the efficacy of the TuFEst model in the staging diagnosis of breast cancer. 2. To evaluate the sensitivity, specificity, accuracy and repeatability of the TuFEst model to determine its reliability in clinical application: This study will collect a larger number of blood samples from breast cancer patients based on the previous retrospective cohort to assess the performance of the model in a larger sample prospective cohort. This study will also explore the application of this technology in the monitoring of neoadjuvant therapy for breast cancer, specifically evaluating its application in post-treatment staging diagnosis of breast cancer, prediction of treatment effects, and monitoring of tumor recurrence.

Conditions