Evaluation of a Novel Content-Based Image Retrieval System for the Differentiation of Interstitial Lung Diseases in CT Examinations

To evaluate the reader’s diagnostic performance against the ground truth with and without the help of a novel content-based image retrieval system (CBIR) that retrieves images with similar CT patterns from a database of 79 different interstitial lung diseases. We evaluated three novice readers’ and...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Tobias Pogarell, Nadine Bayerl, Matthias Wetzl, Jan-Peter Roth, Christoph Speier, Alexander Cavallaro, Michael Uder, Peter Dankerl
Formato:	article
Lenguaje:	EN
Publicado:	MDPI AG 2021
Materias:	content-based image retrieval computer-aided diagnosis interstitial lung disease computed tomography Medicine (General) R5-920
Acceso en línea:	https://doaj.org/article/3bee5105821f42f39cea4e626e2a6524
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:3bee5105821f42f39cea4e626e2a6524
record_format	dspace
spelling	oai:doaj.org-article:3bee5105821f42f39cea4e626e2a65242021-11-25T17:21:44ZEvaluation of a Novel Content-Based Image Retrieval System for the Differentiation of Interstitial Lung Diseases in CT Examinations10.3390/diagnostics111121142075-4418https://doaj.org/article/3bee5105821f42f39cea4e626e2a65242021-11-01T00:00:00Zhttps://www.mdpi.com/2075-4418/11/11/2114https://doaj.org/toc/2075-4418To evaluate the reader’s diagnostic performance against the ground truth with and without the help of a novel content-based image retrieval system (CBIR) that retrieves images with similar CT patterns from a database of 79 different interstitial lung diseases. We evaluated three novice readers’ and three resident physicians‘ (with at least three years of experience) diagnostic performance evaluating 50 different CTs featuring 10 different patterns (e.g., honeycombing, tree-in bud, ground glass, bronchiectasis, etc.) and 24 different diseases (sarcoidosis, UIP, NSIP, Aspergillosis, COVID-19 pneumonia etc.). The participants read the cases first without assistance (and without feedback regarding correctness), and with a 2-month interval in a random order with the assistance of the novel CBIR. To invoke the CBIR, a ROI is placed into the pathologic pattern by the reader and the system retrieves diseases with similar patterns. To further narrow the differential diagnosis, the readers can consult an integrated textbook and have the possibility of selecting high-level semantic features representing clinical information (chronic, infectious, smoking status, etc.). We analyzed readers’ accuracy without and with CBIR assistance and further tested the hypothesis that the CBIR would help to improve diagnostic performance utilizing Wilcoxon signed rank test. The novice readers demonstrated an unassisted accuracy of 18/28/44%, and an assisted accuracy of 84/82/90%, respectively. The resident physicians demonstrated an unassisted accuracy of 56/56/70%, and an assisted accuracy of 94/90/96%, respectively. For each reader, as well as overall, Sign test demonstrated statistically significant (<i>p</i> < 0.01) difference between the unassisted and the assisted reads. For students and physicians, Chi²-test and Mann-Whitney-U test demonstrated statistically significant (<i>p</i> < 0.01) difference for unassisted reads and statistically insignificant (<i>p</i> > 0.01) difference for assisted reads. The evaluated CBIR relying on pattern analysis and featuring the option to filter the results of the CBIR by predominant characteristics of the diseases via selecting high-level semantic features helped to drastically improve novices’ and resident physicians’ accuracy in diagnosing interstitial lung diseases in CT.Tobias PogarellNadine BayerlMatthias WetzlJan-Peter RothChristoph SpeierAlexander CavallaroMichael UderPeter DankerlMDPI AGarticlecontent-based image retrievalcomputer-aided diagnosisinterstitial lung diseasecomputed tomographyMedicine (General)R5-920ENDiagnostics, Vol 11, Iss 2114, p 2114 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	content-based image retrieval computer-aided diagnosis interstitial lung disease computed tomography Medicine (General) R5-920
spellingShingle	content-based image retrieval computer-aided diagnosis interstitial lung disease computed tomography Medicine (General) R5-920 Tobias Pogarell Nadine Bayerl Matthias Wetzl Jan-Peter Roth Christoph Speier Alexander Cavallaro Michael Uder Peter Dankerl Evaluation of a Novel Content-Based Image Retrieval System for the Differentiation of Interstitial Lung Diseases in CT Examinations
description	To evaluate the reader’s diagnostic performance against the ground truth with and without the help of a novel content-based image retrieval system (CBIR) that retrieves images with similar CT patterns from a database of 79 different interstitial lung diseases. We evaluated three novice readers’ and three resident physicians‘ (with at least three years of experience) diagnostic performance evaluating 50 different CTs featuring 10 different patterns (e.g., honeycombing, tree-in bud, ground glass, bronchiectasis, etc.) and 24 different diseases (sarcoidosis, UIP, NSIP, Aspergillosis, COVID-19 pneumonia etc.). The participants read the cases first without assistance (and without feedback regarding correctness), and with a 2-month interval in a random order with the assistance of the novel CBIR. To invoke the CBIR, a ROI is placed into the pathologic pattern by the reader and the system retrieves diseases with similar patterns. To further narrow the differential diagnosis, the readers can consult an integrated textbook and have the possibility of selecting high-level semantic features representing clinical information (chronic, infectious, smoking status, etc.). We analyzed readers’ accuracy without and with CBIR assistance and further tested the hypothesis that the CBIR would help to improve diagnostic performance utilizing Wilcoxon signed rank test. The novice readers demonstrated an unassisted accuracy of 18/28/44%, and an assisted accuracy of 84/82/90%, respectively. The resident physicians demonstrated an unassisted accuracy of 56/56/70%, and an assisted accuracy of 94/90/96%, respectively. For each reader, as well as overall, Sign test demonstrated statistically significant (<i>p</i> < 0.01) difference between the unassisted and the assisted reads. For students and physicians, Chi²-test and Mann-Whitney-U test demonstrated statistically significant (<i>p</i> < 0.01) difference for unassisted reads and statistically insignificant (<i>p</i> > 0.01) difference for assisted reads. The evaluated CBIR relying on pattern analysis and featuring the option to filter the results of the CBIR by predominant characteristics of the diseases via selecting high-level semantic features helped to drastically improve novices’ and resident physicians’ accuracy in diagnosing interstitial lung diseases in CT.
format	article
author	Tobias Pogarell Nadine Bayerl Matthias Wetzl Jan-Peter Roth Christoph Speier Alexander Cavallaro Michael Uder Peter Dankerl
author_facet	Tobias Pogarell Nadine Bayerl Matthias Wetzl Jan-Peter Roth Christoph Speier Alexander Cavallaro Michael Uder Peter Dankerl
author_sort	Tobias Pogarell
title	Evaluation of a Novel Content-Based Image Retrieval System for the Differentiation of Interstitial Lung Diseases in CT Examinations
title_short	Evaluation of a Novel Content-Based Image Retrieval System for the Differentiation of Interstitial Lung Diseases in CT Examinations
title_full	Evaluation of a Novel Content-Based Image Retrieval System for the Differentiation of Interstitial Lung Diseases in CT Examinations
title_fullStr	Evaluation of a Novel Content-Based Image Retrieval System for the Differentiation of Interstitial Lung Diseases in CT Examinations
title_full_unstemmed	Evaluation of a Novel Content-Based Image Retrieval System for the Differentiation of Interstitial Lung Diseases in CT Examinations
title_sort	evaluation of a novel content-based image retrieval system for the differentiation of interstitial lung diseases in ct examinations
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/3bee5105821f42f39cea4e626e2a6524
work_keys_str_mv	AT tobiaspogarell evaluationofanovelcontentbasedimageretrievalsystemforthedifferentiationofinterstitiallungdiseasesinctexaminations AT nadinebayerl evaluationofanovelcontentbasedimageretrievalsystemforthedifferentiationofinterstitiallungdiseasesinctexaminations AT matthiaswetzl evaluationofanovelcontentbasedimageretrievalsystemforthedifferentiationofinterstitiallungdiseasesinctexaminations AT janpeterroth evaluationofanovelcontentbasedimageretrievalsystemforthedifferentiationofinterstitiallungdiseasesinctexaminations AT christophspeier evaluationofanovelcontentbasedimageretrievalsystemforthedifferentiationofinterstitiallungdiseasesinctexaminations AT alexandercavallaro evaluationofanovelcontentbasedimageretrievalsystemforthedifferentiationofinterstitiallungdiseasesinctexaminations AT michaeluder evaluationofanovelcontentbasedimageretrievalsystemforthedifferentiationofinterstitiallungdiseasesinctexaminations AT peterdankerl evaluationofanovelcontentbasedimageretrievalsystemforthedifferentiationofinterstitiallungdiseasesinctexaminations
_version_	1718412464976035840

Evaluation of a Novel Content-Based Image Retrieval System for the Differentiation of Interstitial Lung Diseases in CT Examinations

Ejemplares similares