Disease detection is an emerging research area in the field of medical informatics. Disease detection aims to detect potential health disorders for providing efficient treatment. The significant goal of disease detection is early detection and surveillance to reduce disease risk. Compared to classic learning models, deep learning models are more effective in disease detection and diagnosis. Deep learning models utilize large neural architectures, which have the ability to modify their hyper-parameters based on new data updation and automatically extract high-level features. The drawback of the deep learning model in disease detection is a huge time and computational resources to train the model with massive parameters. Deep transfer learning is introduced to solve such issues, which combine transfer learning and deep learning model.
Transfer learning refers to improving the performance of the target domain by transferring the knowledge to different source domains. Deep transfer learning trains a neural network for a task, and it is reused as starting point of the next task. Deep learning-based disease detection magnifies feature extraction capability and reduces the computational complexity of the model without losing the discriminant formation.