Benchmarking Medical Image Databases

Abstract An increase in the use of digital images in medicine and medical research has resulted in petabytes of digital images acquired each year. As more information is collected, there is a greater need to store, share, and retrieve images from various medical imaging modalities. Web-based medical image data management systems have been created to solve many issues caused by the Picture Archiving and Communication System (PACS). Five such medical image database systems are reviewed in this work. Though these systems have various features, they seem to fail to capture experimental design specifications necessary for the proper interpretation of the associated images. We summarize the paper by indicating a wish list of the features that are missing from these systems.