Clinical Study of Imaging Genomics Based on Machine Learning for BCIG

1. Identify the imaging features of breast cancer with different molecular types 2. Reveal the association between hormone receptor positive/HER2 negative breast cancer and imaging histology, Oncotype Dx recurrence score 3. Combine genomics and imaging to establish a predictive model for the sensitivity of HER2-positive breast cancer targeted therapy 4. Establish an imaging genomics prediction model for triple-negative breast cancer molecular subtypes, and clarify the imaging genomics characteristics of the therapeutic targets of each subtype