Quantitative computed tomography applied to interstitial lung diseases

Martin Obert; Marian Kampschulte; Rebekka Limburg; Stefan Barańczuk; Gabriele A Krombach

doi:10.1016/j.ejrad.2018.01.018

Quantitative computed tomography applied to interstitial lung diseases

Eur J Radiol. 2018 Mar:100:99-107. doi: 10.1016/j.ejrad.2018.01.018. Epub 2018 Jan 31.

Authors

Martin Obert¹, Marian Kampschulte², Rebekka Limburg², Stefan Barańczuk³, Gabriele A Krombach²

Affiliations

¹ Department of Radiology, University Hospital Giessen, Justus-Liebig-University Giessen, Klinikstrasse 33, 35392 Giessen, Germany Members of The German Center for Lung Research (DZL e. V.). Electronic address: martin.obert@radiol.med.uni-giessen.de.
² Department of Radiology, University Hospital Giessen, Justus-Liebig-University Giessen, Klinikstrasse 33, 35392 Giessen, Germany Members of The German Center for Lung Research (DZL e. V.).
³ Faculty of Mathematics and Computer Science, Adam Mickiewicz University, Umultowska 87, 61-614 Poznań, Poland.

PMID: 29496086
DOI: 10.1016/j.ejrad.2018.01.018

Abstract

Objectives: To evaluate a new image marker that retrieves information from computed tomography (CT) density histograms, with respect to classification properties between different lung parenchyma groups. Furthermore, to conduct a comparison of the new image marker with conventional markers.

Materials and methods: Density histograms from 220 different subjects (normal = 71; emphysema = 73; fibrotic = 76) were used to compare the conventionally applied emphysema index (EI), 15^th percentile value (PV), mean value (MV), variance (V), skewness (S), kurtosis (K), with a new histogram's functional shape (HFS) method. Multinomial logistic regression (MLR) analyses was performed to calculate predictions of different lung parenchyma group membership using the individual methods, as well as combinations thereof, as covariates. Overall correct assigned subjects (OCA), sensitivity (sens), specificity (spec), and Nagelkerke's pseudo R² (NR²) effect size were estimated. NR² was used to set up a ranking list of the different methods.

Results: MLR indicates the highest classification power (OCA of 92%; sens 0.95; spec 0.89; NR² 0.95) when all histogram analyses methods were applied together in the MLR. Highest classification power among individually applied methods was found using the HFS concept (OCA 86%; sens 0.93; spec 0.79; NR² 0.80). Conventional methods achieved lower classification potential on their own: EI (OCA 69%; sens 0.95; spec 0.26; NR² 0.52); PV (OCA 69%; sens 0.90; spec 0.37; NR² 0.57); MV (OCA 65%; sens 0.71; spec 0.58; NR² 0.61); V (OCA 66%; sens 0.72; spec 0.53; NR² 0.66); S (OCA 65%; sens 0.88; spec 0.26; NR² 0.55); and K (OCA 63%; sens 0.90; spec 0.16; NR² 0.48).

Conclusion: The HFS method, which was so far applied to a CT bone density curve analysis, is also a remarkable information extraction tool for lung density histograms. Presumably, being a principle mathematical approach, the HFS method can extract valuable health related information also from histograms from complete different areas.

Keywords: CT histogram analysis; Image marker; Multinomial logistic regression; Quantitative image analysis; Radiomics.

Publication types

Evaluation Study

MeSH terms

Evaluation Studies as Topic
Female
Humans
Lung / diagnostic imaging
Lung Diseases, Interstitial / diagnostic imaging*
Male
Middle Aged
Reproducibility of Results
Retrospective Studies
Sensitivity and Specificity
Tomography, X-Ray Computed / methods*