Deep Learning Approaches for Disease Prediction: A Survey of Techniques and Applications in Healthcare

Deep learning (DL) is revolutionizing two fundamental aspects of disease: early diagnosis and prediction. Some of these models such as AIs for medical imaging (for example, CNNs), electronic health records (for example, RNNs/LSTMs) and medical natural language processing and multimodal learning (for example, transformers) are presented in the following survey papers. Though these models have great potential to spot diseases like
COVID-19, cancer and heart disease, there are still questions
about a shortage of data, interpretability by researchers and
ethical concerns. This article shares an overview of the future of
predictive healthcare by exploring the latest technologies
including wearable tight integration, federated learning and
explainable AI