Since joining into CIDI team, I was primarily interested in detection and quantification of an abnormal imaging pattern pertaining infectious disease, named Tree-in-bud (TIB), and observed in lung CTs. In 2011, we have published our initial detection algorithm in MICCAI, then an alternative method with comparison in IEEE EMBC. More recently, we have added computational quantification methods into the work and partially published in IEEE ISBI 2012. I will be presenting work in Barcelona in May-2012.
We have combined all lung segmentation, detection, and computer quantification methods into a framework where we have done TIB detection and quantification in a fully automated way. Overall process written in journal article has been accepted today into IEEE Transaction on Biomedical Engineering Journal.
Here is the title, author list, and abstract of the paper. The full links to the pdf of the paper will be available from publication link.
Automatic Detection and Quantification of Tree-in-Bud (TIB) Opacities from CT Scans
U. Bagci, J. Yao, A. Wu, J.Caban, T. Palmore, A. Suffredini, O. Aras, and D. J. Mollura.
Abstract—This study presents a novel computer assisted detection
(CAD) system for automatically detecting and precisely quantifying
abnormal nodular branching opacities in chest computed tomography
(CT), termed tree-in-bud (TIB) opacities by radiology literature. The
developed CAD system in this study is based on (1) fast localization
of candidate imaging patterns using local scale information of the
images, and (2) Mobius invariant feature extraction method based on
learned local shape and texture properties of TIB patterns. For fast
localization of candidate imaging patterns, we use ball-scale (b-scale)
filtering and, based on the observation of the pattern of interest, a
suitable scale selection is used to retain only small size patterns. Once
candidate abnormality patterns are identified, we extract proposed shape
features from regions where at least one candidate pattern occupies. The
comparative evaluation of the proposed method with commonly used
CAD methods is presented with a data set of 60 chest CTs (laboratory
confirmed 39 viral bronchiolitis human parainfluenza (HPIV) CTs and
21 normal chest CTs). The quantitative results are presented as the
area under the receiver operator characteristics (ROC) curves and a
computer score (volume affected by TIB) provided as an output of
the CAD system. In addition, a visual grading scheme is applied to
the patient data by three well trained radiologists. Inter-observer and
observer-computer agreements are obtained by the relevant statistical
methods over different lung zones. Experimental results demonstrate
that the proposed CAD system can achieve high detection rates with an
overall accuracy of 90.96%. Moreover, correlations of observer-observer
(R2 = 0.8848, p < 0.01) and observer-CAD agreements (R2 = 0.824,
p < 0.01) validate the feasibility of the use of the proposed CAD system
in detecting and quantifying TIB patterns.