Radiological Image and Information Processing Lab (RIIPL)
The practice of Radiology is undergoing a radical transformation from one in which the primary result of an imaging examination is a written report addressing the reasons that the examination was ordered, to one in which the output is a (set of) quantitative measurement(s) with links to knowledge that could affect treatment. For example, while a traditional report might have said “there is a mass in the right upper lobe of the lung,” the report of the future might say “The mass in the right upper lobe of the lung has grown by 25% since the last examination 3 months ago; it now measures 60 cc and has imaging features consistent with adenocarcinoma with an EGFR mutation that has has a favorable response to TK inhibitors. Click these links for similar cases and their clinical history. See references [1-4] for the latest articles of relevance.” Our lab, in collaboration with other IBIIS labs, radiologists, and other clinicians, and other collaborators from the School of Medicine, is involved in many aspects of creating that future, including advanced software for image visualization and quantitative analysis, image segmentation software that isolates regions within images for further analysis, software that extracts imaging features (e.g., shape, size, margin sharpness, pixel texture) within these regions, and algorithms for computing similarity between images and between patients as expressed by their images, demographic and clinical data.
Sandy Napel, PhD
Professor of Radiology (General Radiology) and, by courtesy, of Medicine (Medical Informatics) and of Electrical Engineering
My primary interests are in developing diagnostic and therapy-planning applications and strategies for the acquisition, visualization, and quantitation of multi-dimensional medical imaging data. Examples are: creation of three-dimensional images of blood vessels using CT, visualization of complex flow within blood vessels using MR, computer-aided detection and characterization of lesions (e.g., colonic polyps, pulmonary nodules) from cross-sectional image data, visualization and automated assessment of 4D ultrasound data, and fusion of images acquired using different modalities (e.g., CT and MR). I have also been involved in developing and evaluating techniques for exploring cross-sectional imaging data from an internal perspective, i.e., virtual endoscopy (including colonoscopy, angioscopy, and bronchoscopy), and in the quantitation of structure parameters, e.g., volumes, lengths, medial axes, and curvatures. I am also interested in creating workable solutions to the problem of "data explosion," i.e., how to look at the thousands of images generated per examination using modern CT and MR scanners. My most recent focus includes making image features computer-accessible, to facilitate content-based retrieval of similar lesions, and prediction of molecular phenotype, response to therapy, and prognosis from imaging features. I am co-director of the Radiology 3D and Quantitative Imaging Lab, providing clinical service to the Stanford and local community, and co-Director of IBIIS (Integrative Biomedical Imaging Informatics at Stanford), whose mission is to advance the clinical and basic sciences in radiology, while improving our understanding of biology and the manifestations of disease, by pioneering methods in the information sciences that integrate imaging, clinical and molecular data.