Multi-modal deep temporal adversarial network based on multi-head self-attention for breast cancer survival prediction

Breast cancer is the most commonly diagnosed malignant tumor among women worldwide. Since breast cancer is highly heterogeneous, it is difficult to determine the cancer type and treatment plan. Accurate prognostic prediction is crucial for breast cancer management. However, existing prediction methods often overlook survival-related information, affecting multi-modal feature extraction and fusion. To address these limitations, we propose a novel Multi-modal Deep Temporal Adversarial Network (MDTAN) designed to predict breast cancer prognosis with greater accuracy. We evaluated our model on the METABRIC dataset using various metrics. Compared with other methods, our model shows superior prediction performance.