Breast cancer, as one of the most common and significant cancers, plays a major role in increasing mortality rates among women. Currently, multi-modal cancer-related data, including genomic details, mammography images, and clinical information, are available, which has led to increased focus on developing advanced deep-learning models for predicting breast cancer survival. This study proposes a deep learning model based on convolutional neural networks (CNN) for predicting breast cancer survival using multi-modal data. This model extracts crucial features to enhance prediction accuracy. The proposed method comprises two phases: in the first phase, a CNN is used to extract effective features, and in the second phase, these features are employed to predict breast cancer survival. The experimental results demonstrated that the proposed method achieved an accuracy of 98.34% when clinical data were utilized, and 97.21% without clinical data. This indicates that the proposed model significantly outperforms the compared models. Moreover, the results show that incorporating clinical data is instrumental in improving the accuracy and performance of the proposed model.