Heterogeneity Index for Spatial Distribution in CT Images and Correlation with Symptom Scores

• Activity Number: 122
• Type: Topic Contributed
• Date/Time: Monday, August 5, 2013 : 8:30 AM to 10:20 AM
• Sponsor: Biometrics Section
• Author(s): Eran Barnoy*+ and Hyun (Grace) Kim and Jonathan Goldin and Eric Kleerup and Peiyun Lu
• Companies: UCLA Biostatistics and UCLA and UCLA Radiology and UCLA Division of Pulmonary and Critical Care and UCLA Radiological Science
• Keywords: heterogeneity ; asthma ; spatial distribution ; air trapping ; APLE ; imaging

Abstract

CT images help visualize air trapping in lungs of asthma patients. The images also allow us to quantify air trapping, and so measure a degree of spatial organization. The heterogeneity of the spatial distribution, combined with disease severity, is indicative of overall disease status. In this talk, we discuss indices for spatial heterogeneity and their association with asthma symptom scores. We indicate proximity of segments by treating two segments in the same lobe as neighbors. Then, two indices, the Moran's I and the approximate profile-likelihood estimator (APLE), can be determined in order to describe heterogeneity. However, due to the sensitivity of these measures to the degree of air trapping, their values can give false indications of disease status. To adjust for the problem and make a more clinically relevant parameter, the APLE has been multiplied by the average air trapping for a given patient, normalized by the sample maximum. The new index has been calculated for asthma patients and compared with symptom scores. We find that the proposed heterogeneity index and scores associate well, indicating a possible quantitative replacement to the more subjective score value.