ENHANCED GENERATIVE MODEL FOR UNSUPERVISED DISCOVERY OF SPATIALLY-INFORMED MACROSCOPIC EMPHYSEMA: THE MESA COPD STUDY

Yu Gan; Jie Yang; Benjamin Smith; Pallavi Bake; Eric Hoffman; Christine Hendon; R. Graham Barr; Andrew F. Laine; Elsa D. Angelini

doi:10.1109/ISBI.2019.8759243

Conference proceeding

ENHANCED GENERATIVE MODEL FOR UNSUPERVISED DISCOVERY OF SPATIALLY-INFORMED MACROSCOPIC EMPHYSEMA: THE MESA COPD STUDY

Yu Gan, Jie Yang, Benjamin Smith, Pallavi Bake, Eric Hoffman, Christine Hendon, R. Graham Barr, Andrew F. Laine and Elsa D. Angelini

2019 IEEE 16th International Symposium on Biomedical Imaging (ISBI 2019), pp.1212-1215

IEEE International Symposium on Biomedical Imaging

01/01/2019

DOI: 10.1109/ISBI.2019.8759243

View Online

Abstract

Pulmonary emphysema, overlapping with Chronic Obstructive Pulmonary Disorder (COPD), contributes to a significant amount of morbidity and mortality annually. Computed tomography is used for in viva quantification of emphysema and labeling into three standard subtypes at a macroscopic level. Unsupervised learning of texture patterns has great potential to discover more radiological emphysema subtypes. In this work, we improve a probabilistic Latent Dirichlet Allocation (IDA) model to discover spatially-informed lung macroscopic patterns (sLMPs) from previously learned spatially informed lung texture patterns (sLTPs). We exploit a specific reproducibility metric to empirically tune the number of sLMPs and the size of patches. Experimental results on the MESA COPD cohort show that our algorithm can discover highly reproducible sLMPs, which are able to capture relationships between sLTPs and preferred localizations within the lung. The discovered sLMPs also achieve higher prediction accuracy of three standard emphysema subtypes than in our previous implementation.

Engineering

Technology

Engineering, Biomedical

Life Sciences & Biomedicine

Radiology, Nuclear Medicine & Medical Imaging

Science & Technology

Details

Title: Subtitle: ENHANCED GENERATIVE MODEL FOR UNSUPERVISED DISCOVERY OF SPATIALLY-INFORMED MACROSCOPIC EMPHYSEMA: THE MESA COPD STUDY
Creators: Yu Gan - Columbia University
Jie Yang - Columbia University
Benjamin Smith - Columbia University Irving Medical Center
Pallavi Bake - Columbia Univ, Dept Med, Med Ctr, New York, NY USA
Eric Hoffman - University of Iowa
Christine Hendon - Columbia University
R. Graham Barr - Columbia University Irving Medical Center
Andrew F. Laine - Columbia University
Elsa D. Angelini - Columbia University
Resource Type: Conference proceeding
Publication Details: 2019 IEEE 16th International Symposium on Biomedical Imaging (ISBI 2019), pp.1212-1215
Publisher: IEEE
Series: IEEE International Symposium on Biomedical Imaging
DOI: 10.1109/ISBI.2019.8759243
ISSN: 1945-7928
eISSN: 1945-8452
Number of pages: 4
Grant note: R01-HL121270; R01-HL077612; RC1-HL100543; R01-HL093081; N01-HC095159; N01-HC-95169; UL1-RR-024156; UL1-RR-025005 / NIH/NHLBI; United States Department of Health & Human Services; National Institutes of Health (NIH) - USA; NIH National Heart Lung & Blood Institute (NHLBI)
Language: English
Date published: 01/01/2019
Academic Unit: Roy J. Carver Department of Biomedical Engineering; Internal Medicine; Radiology
Record Identifier: 9984318723502771

Metrics

10 Record Views