<p>Constant advances and innovation in technology, and particularly in medical imaging modalities, have enhanced diagnostic capabilities, enabling medical experts and professionals to visualize and diagnose a variety of medical conditions in order to provide better patient care. Recently, to exploit these advantages, multi-source medical image fusion has been widely applied in two highly recommended situations, namely classification-based computer-aided diagnosis (CAD) and retrieval-based CAD. In fact, the concept of multi-source medical imaging fusion has been used successfully, particularly in the context of breast cancer, cardiac magnetic resonance imaging and pulmonary nodules. In this study, we present a comprehensive survey of the latest trends in multi-source medical imaging methods in the context of classification (normal <i>vs.</i> abnormal) and content-based retrieval. In addition, we describe the most relevant multi-source medical image datasets that are publicly available.</p>