Cross-modality Sub-image Retrieval Using Contrastive Multimodal Image Representations

This article discusses the use of multimodal imaging and computational advances in tissue characterization and cancer diagnostics. It proposes a new content-based image retrieval system that combines deep learning and robust feature extraction for efficient and reliable retrieval. The system is shown to outperform other cross-modality CBIR methods in a study using a dataset of microscopy images.