Chemotherapy-induced cardiotoxicity is a significant complication that can severely impact cancer patients’ cardiac health, leading to conditions such as heart failure, myocardial infarction, and arrhythmias. Early detection of cardiotoxicity is critical, yet traditional methods often struggle to identify it at a sufficiently early stage. In this study, we propose a multimodal deep learning model enhanced with generative AI to improve the early detection of chemotherapy-induced cardiotoxicity. The model integrates clinical data—such as age, weight, height, treatment specifics, and cardiovascular metrics—with functional data from Tissue Doppler Imaging (TDI), which captures myocardial velocity during the cardiac cycle. Using Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs), the model combines these multimodal features to enhance prediction accuracy. Additionally, Generative Adversarial Networks (GANs) are employed to augment the training data, improving the model’s robustness and generalizability. Key findings demonstrate that the proposed model achieved an accuracy of 97.9% and an AUC-ROC of 0.99, significantly outperforming traditional methods such as logistic regression 88% and CNNs 87%. These results suggest that our approach holds great potential for clinical applications, supporting early detection, personalized treatment plans, and improved patient care in oncology. The proposed model demonstrates superior performance over traditional diagnostic methods. Key performance metrics include enhanced accuracy, sensitivity, and specificity in predicting the onset of cardiotoxic effects due to chemotherapy. The effectiveness of the model is quantified through rigorous validation protocols, which show marked improvement in early detection capabilities.