Cargando…

Reliability as a Precondition for Trust—Segmentation Reliability Analysis of Radiomic Features Improves Survival Prediction

Machine learning results based on radiomic analysis are often not transferrable. A potential reason for this is the variability of radiomic features due to varying human made segmentations. Therefore, the aim of this study was to provide comprehensive inter-reader reliability analysis of radiomic fe...

Descripción completa

Detalles Bibliográficos
Autores principales: Müller-Franzes, Gustav, Nebelung, Sven, Schock, Justus, Haarburger, Christoph, Khader, Firas, Pedersoli, Federico, Schulze-Hagen, Maximilian, Kuhl, Christiane, Truhn, Daniel
Formato: Online Artículo Texto
Lenguaje:English
Publicado: MDPI 2022
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8871487/
https://www.ncbi.nlm.nih.gov/pubmed/35204338
http://dx.doi.org/10.3390/diagnostics12020247
_version_ 1784657007386558464
author Müller-Franzes, Gustav
Nebelung, Sven
Schock, Justus
Haarburger, Christoph
Khader, Firas
Pedersoli, Federico
Schulze-Hagen, Maximilian
Kuhl, Christiane
Truhn, Daniel
author_facet Müller-Franzes, Gustav
Nebelung, Sven
Schock, Justus
Haarburger, Christoph
Khader, Firas
Pedersoli, Federico
Schulze-Hagen, Maximilian
Kuhl, Christiane
Truhn, Daniel
author_sort Müller-Franzes, Gustav
collection PubMed
description Machine learning results based on radiomic analysis are often not transferrable. A potential reason for this is the variability of radiomic features due to varying human made segmentations. Therefore, the aim of this study was to provide comprehensive inter-reader reliability analysis of radiomic features in five clinical image datasets and to assess the association of inter-reader reliability and survival prediction. In this study, we analyzed 4598 tumor segmentations in both computed tomography and magnetic resonance imaging data. We used a neural network to generate 100 additional segmentation outlines for each tumor and performed a reliability analysis of radiomic features. To prove clinical utility, we predicted patient survival based on all features and on the most reliable features. Survival prediction models for both computed tomography and magnetic resonance imaging datasets demonstrated less statistical spread and superior survival prediction when based on the most reliable features. Mean concordance indices were C(mean) = 0.58 [most reliable] vs. C(mean) = 0.56 [all] (p < 0.001, CT) and C(mean) = 0.58 vs. C(mean) = 0.57 (p = 0.23, MRI). Thus, preceding reliability analyses and selection of the most reliable radiomic features improves the underlying model’s ability to predict patient survival across clinical imaging modalities and tumor entities.
format Online
Article
Text
id pubmed-8871487
institution National Center for Biotechnology Information
language English
publishDate 2022
publisher MDPI
record_format MEDLINE/PubMed
spelling pubmed-88714872022-02-25 Reliability as a Precondition for Trust—Segmentation Reliability Analysis of Radiomic Features Improves Survival Prediction Müller-Franzes, Gustav Nebelung, Sven Schock, Justus Haarburger, Christoph Khader, Firas Pedersoli, Federico Schulze-Hagen, Maximilian Kuhl, Christiane Truhn, Daniel Diagnostics (Basel) Article Machine learning results based on radiomic analysis are often not transferrable. A potential reason for this is the variability of radiomic features due to varying human made segmentations. Therefore, the aim of this study was to provide comprehensive inter-reader reliability analysis of radiomic features in five clinical image datasets and to assess the association of inter-reader reliability and survival prediction. In this study, we analyzed 4598 tumor segmentations in both computed tomography and magnetic resonance imaging data. We used a neural network to generate 100 additional segmentation outlines for each tumor and performed a reliability analysis of radiomic features. To prove clinical utility, we predicted patient survival based on all features and on the most reliable features. Survival prediction models for both computed tomography and magnetic resonance imaging datasets demonstrated less statistical spread and superior survival prediction when based on the most reliable features. Mean concordance indices were C(mean) = 0.58 [most reliable] vs. C(mean) = 0.56 [all] (p < 0.001, CT) and C(mean) = 0.58 vs. C(mean) = 0.57 (p = 0.23, MRI). Thus, preceding reliability analyses and selection of the most reliable radiomic features improves the underlying model’s ability to predict patient survival across clinical imaging modalities and tumor entities. MDPI 2022-01-19 /pmc/articles/PMC8871487/ /pubmed/35204338 http://dx.doi.org/10.3390/diagnostics12020247 Text en © 2022 by the authors. https://creativecommons.org/licenses/by/4.0/Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
spellingShingle Article
Müller-Franzes, Gustav
Nebelung, Sven
Schock, Justus
Haarburger, Christoph
Khader, Firas
Pedersoli, Federico
Schulze-Hagen, Maximilian
Kuhl, Christiane
Truhn, Daniel
Reliability as a Precondition for Trust—Segmentation Reliability Analysis of Radiomic Features Improves Survival Prediction
title Reliability as a Precondition for Trust—Segmentation Reliability Analysis of Radiomic Features Improves Survival Prediction
title_full Reliability as a Precondition for Trust—Segmentation Reliability Analysis of Radiomic Features Improves Survival Prediction
title_fullStr Reliability as a Precondition for Trust—Segmentation Reliability Analysis of Radiomic Features Improves Survival Prediction
title_full_unstemmed Reliability as a Precondition for Trust—Segmentation Reliability Analysis of Radiomic Features Improves Survival Prediction
title_short Reliability as a Precondition for Trust—Segmentation Reliability Analysis of Radiomic Features Improves Survival Prediction
title_sort reliability as a precondition for trust—segmentation reliability analysis of radiomic features improves survival prediction
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8871487/
https://www.ncbi.nlm.nih.gov/pubmed/35204338
http://dx.doi.org/10.3390/diagnostics12020247
work_keys_str_mv AT mullerfranzesgustav reliabilityasapreconditionfortrustsegmentationreliabilityanalysisofradiomicfeaturesimprovessurvivalprediction
AT nebelungsven reliabilityasapreconditionfortrustsegmentationreliabilityanalysisofradiomicfeaturesimprovessurvivalprediction
AT schockjustus reliabilityasapreconditionfortrustsegmentationreliabilityanalysisofradiomicfeaturesimprovessurvivalprediction
AT haarburgerchristoph reliabilityasapreconditionfortrustsegmentationreliabilityanalysisofradiomicfeaturesimprovessurvivalprediction
AT khaderfiras reliabilityasapreconditionfortrustsegmentationreliabilityanalysisofradiomicfeaturesimprovessurvivalprediction
AT pedersolifederico reliabilityasapreconditionfortrustsegmentationreliabilityanalysisofradiomicfeaturesimprovessurvivalprediction
AT schulzehagenmaximilian reliabilityasapreconditionfortrustsegmentationreliabilityanalysisofradiomicfeaturesimprovessurvivalprediction
AT kuhlchristiane reliabilityasapreconditionfortrustsegmentationreliabilityanalysisofradiomicfeaturesimprovessurvivalprediction
AT truhndaniel reliabilityasapreconditionfortrustsegmentationreliabilityanalysisofradiomicfeaturesimprovessurvivalprediction