"Pathological Lung Segmentation and Computer Aided Diagnosis Systems for Pulmonary Abnormalities"
Dr. Ulas Bagci
National Institutes of Health
Tuesday, November 4, 2014 · 2:00PM · HEC 101
Accurate segmentation of pathological lungs from computed tomography (CT) scans remains unsolved because available methods in the literature fail to provide a reliable
generic solution for a wide spectrum of lung abnormalities. In this talk, I will present a novel pathological lung segmentation method that takes into account neighbor
prior constraints and a novel pathology recognition system. This rudimentary, but intelligent lung volume estimation system allows comparison of volume differences
between rib cage and initial lung volume measurements. Significant volume difference indicates the presence of pathology, which invokes the second stage of the proposed
framework for the refinement of segmented lung based on machine learning algorithms as well as neighbor prior constraints.
In second part of my talk, I will describe various abnormal imaging markers for pulmonary diseases (consolidations, ground glass, interstitial thickening, tree-in-bud,
honeycombing, nodules, and micro-nodules) and then I will focus on the automatic detection of some of those patterns, particularly the ones with varying shape and
Accuracy and efficiency of the proposed method was tested on more than 400 CT scans with the presence of a wide spectrum of abnormalities. This is the first study to
evaluate all abnormal imaging patterns in a single segmentation framework. The quantitative results show that the pathological lung segmentation method improves on
current standards because of its high sensitivity and specificity and may have considerable potential to enhance the performance of routine clinical tasks.
Dr. Bagci is a Staff Scientist in NIH's Center for Infectious Disease Imaging (CIDI) Lab, in Radiology and Imaging Sciences (RAD&IS), where he is conducting research
focused on medical image processing, analysis, and computer vision fields. Dr. Bagci began his work at NIH in 2010 when he received an Imaging Science Training Program
(ISTP) Fellowship in 2010-2012. Prior to joining CIDI in 2010, he was a Marie Curie Research Fellow in the Collaborative Medical Image Analysis Group at the University
of Nottingham where he was awarded his PhD in April 2010. In 2009, he was a visiting research scholar in the Medical Image Processing Group at the University of
Pennsylvania under the supervision of Prof. Jayaram K. Udupa. He received his BSc and MSc degrees in the Electrical and Electronics Engineering Department at Bilkent
University, and in Electrical and Computer Engineering from Koc University, Turkey, in 2003 and 2005, respectively. He was the winner of NIH Fellow Award for Research
Excellence (FARE) awards in 2012 and 2011. He has received several awards including Best Poster Prize in Molecular Imaging of Infectious Diseases, RSNA Education
exhibit, IEEE - Best Student Paper in the IEEE Conference on SIU, 2006. Dr Bagci is a regular reviewer for IEEE TMI, TBME, TIP, Computerized Medical Imaging and
Graphics, Computer Vision and Image Understanding, and Computers in Biology and Medicine. He is also serving as an editorial board member on several journals for image
processing and computer vision.