Tim Disney

A Lookup and Reference Tool for Pulmonary Computed Tomography Nodules

Picture archive and communication systems (PACS) often contain millions of images related to clinical studies performed at an institution. In the current standard of practice, these images are often not annotated in a standards-based, structured fashion which would allow retrieval based on image features identified by an expert observer. The integration of content-based image retrieval (CBIR) into radiologist workstations is important for situations when relevant images cannot be retrieved by the coarse search of patient and study related information alone. A content-based medical image retrieval system could be used extensively for reference and teaching by providing retrieval of cases similar to a given image. This kind of system could also be used for decision support by retrieving cases with similar features and providing assisted diagnosis based on known pathology in retrieved cases. This paper describes a demonstration CBIR system for pulmonary nodule lookup