Speaker: Raúl San José Estépar, PhD, Brigham and Women’s Hospital, Harvard Medical School,
Date: Wednesday, July 8, 2020
Time: 2:00PM – 3:00PM Central Time
Location: Web conference (a link will be sent out)
Title: “Artificial Intelligence in Quantitative Imaging of Chronic Lung Injury: Enabling Clinical and Genetic Discovery”
Synopsis: Chronic Obstructive Pulmonary Disease (COPD) is a chronic, inflammatory lung disease that arises from exposure to cigarette smoke and other inhaled toxins. It is the 3rd leading cause of death worldwide, affecting about 10% of the general population, but its prevalence among heavy smokers can reach 50%. The main clinical feature of COPD is a limitation of airflow that is not fully reversible. Despite the simplicity of the clinical definition, the heterogeneity of this disease makes the clinical assessment and treatment complex. In-vivo Quantitative Chest Computed Tomography provides high-resolution structural information of the lung that enables better characterization of patients suffering from COPD. In this talk, I will give an overview of the image-based
biomarkers that are being developed to phenotype the different pathophysiological components of COPD: airway disease, emphysema, and pulmonary vascular disease. I will present the current paradigm to develop image-based biomarkers in lung diseases, and I will introduce some of the new deep learning approaches that are allowing us to perform end-to-end automatic imaging phenotyping in large populations. Finally, I will show some of the applications of these new approaches for clinical and genetic discovery.
Bio: Raúl is co-director of the Applied Chest Imaging Laboratory, lead scientist at Brigham and Women’s Hospital and Associate Professor of Radiology at Harvard Medical School.
Raúl is focused on the development of quantitative imaging biomarkers to define novel primary or secondary endpoints in lung clinical investigations, specif
ically Chronic Obstructive Pulmonary Disease (COPD), Pulmonary Vascular Disease (PVD) and Interstitial Lung Disease (ILD). In particular, his group is heavily involved in the development of novel imaging computational techniques that form the basis of the biomarker discovery pipeline to establish the clinical and genetics significant of chest diseases. His group has served as imaging core for the COPDGene study and the Framingham Heart Study Pulmonary Research Center among others. The current research lines span from the quantitative study of pulmonary vascular remodeling to the subtyping of parenchymal lung injury and the prediction of outcomes directly from imaging by means of artificial intelligence techniques.
Prior areas of interest included the development of techniques for the processing of Diffusion Tensor Imaging (DTI). He also designed and implementated image-guided approaches for laparoscopic (IRLUS) and endoscopic (IRGUS) interventions guided by ultrasound and its application to Natural Transluminal Endoscopic Surgery (NOTES). His image analysis interest includes low level operators based on tensor image analysis, image segmentation and registration and the application of deep learning approaches to medical imaging applications.
Raúl is also an affiliated member of the Laboratory of Mathematics in imaging (LMI), the Surgical Planning Laboratory (SPL), Clinical Image Guidance Laboratory (CIGL) and the Image Processing Laboratory (LPI). Raúl received his M.Sc. and Ph.D. in Telecommunications Engineering from the University of Valladolid, Spain and conducted his post-doctoral training at Harvard Medical School. He has co-authored over 100 peer-reviewed manuscripts, and he is currently the Principal Investigator of three NIH NHLBI awards. He is a member of the IMPACT program at MIT and the Fleischner Society.