AI and High-Grade Glioma for Diagnosis and Outcome Prediction: Do All Machine Learning Models Perform Equally Well?

Radiomic models outperform clinical data for outcome prediction in high-grade gliomas (HGG). However, lack of parameter standardization limits clinical applications. Many machine learning (ML) radiomic models employ single classifiers rather than ensemble learning, which is known to boost performanc...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Luca Pasquini, Antonio Napolitano, Martina Lucignani, Emanuela Tagliente, Francesco Dellepiane, Maria Camilla Rossi-Espagnet, Matteo Ritrovato, Antonello Vidiri, Veronica Villani, Giulio Ranazzi, Antonella Stoppacciaro, Andrea Romano, Alberto Di Napoli, Alessandro Bozzao
Formato:	article
Lenguaje:	EN
Publicado:	Frontiers Media S.A. 2021
Materias:	glioblastoma machine learning radiomics survival high-grade glioma (HGG) genetics Neoplasms. Tumors. Oncology. Including cancer and carcinogens RC254-282
Acceso en línea:	https://doaj.org/article/41f565c7ad4b436b98ef650d7c372b72
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:41f565c7ad4b436b98ef650d7c372b72
record_format	dspace
spelling	oai:doaj.org-article:41f565c7ad4b436b98ef650d7c372b722021-11-30T12:10:40ZAI and High-Grade Glioma for Diagnosis and Outcome Prediction: Do All Machine Learning Models Perform Equally Well?2234-943X10.3389/fonc.2021.601425https://doaj.org/article/41f565c7ad4b436b98ef650d7c372b722021-11-01T00:00:00Zhttps://www.frontiersin.org/articles/10.3389/fonc.2021.601425/fullhttps://doaj.org/toc/2234-943XRadiomic models outperform clinical data for outcome prediction in high-grade gliomas (HGG). However, lack of parameter standardization limits clinical applications. Many machine learning (ML) radiomic models employ single classifiers rather than ensemble learning, which is known to boost performance, and comparative analyses are lacking in the literature. We aimed to compare ML classifiers to predict clinically relevant tasks for HGG: overall survival (OS), isocitrate dehydrogenase (IDH) mutation, O-6-methylguanine-DNA-methyltransferase (MGMT) promoter methylation, epidermal growth factor receptor vIII (EGFR) amplification, and Ki-67 expression, based on radiomic features from conventional and advanced magnetic resonance imaging (MRI). Our objective was to identify the best algorithm for each task. One hundred fifty-six adult patients with pathologic diagnosis of HGG were included. Three tumoral regions were manually segmented: contrast-enhancing tumor, necrosis, and non-enhancing tumor. Radiomic features were extracted with a custom version of Pyradiomics and selected through Boruta algorithm. A Grid Search algorithm was applied when computing ten times K-fold cross-validation (K=10) to get the highest mean and lowest spread of accuracy. Model performance was assessed as AUC-ROC curve mean values with 95% confidence intervals (CI). Extreme Gradient Boosting (xGB) obtained highest accuracy for OS (74,5%), Adaboost (AB) for IDH mutation (87.5%), MGMT methylation (70,8%), Ki-67 expression (86%), and EGFR amplification (81%). Ensemble classifiers showed the best performance across tasks. High-scoring radiomic features shed light on possible correlations between MRI and tumor histology.Luca PasquiniLuca PasquiniAntonio NapolitanoMartina LucignaniEmanuela TaglienteFrancesco DellepianeMaria Camilla Rossi-EspagnetMaria Camilla Rossi-EspagnetMatteo RitrovatoAntonello VidiriVeronica VillaniGiulio RanazziAntonella StoppacciaroAndrea RomanoAlberto Di NapoliAlberto Di NapoliAlessandro BozzaoFrontiers Media S.A.articleglioblastomamachine learningradiomicssurvivalhigh-grade glioma (HGG)geneticsNeoplasms. Tumors. Oncology. Including cancer and carcinogensRC254-282ENFrontiers in Oncology, Vol 11 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	glioblastoma machine learning radiomics survival high-grade glioma (HGG) genetics Neoplasms. Tumors. Oncology. Including cancer and carcinogens RC254-282
spellingShingle	glioblastoma machine learning radiomics survival high-grade glioma (HGG) genetics Neoplasms. Tumors. Oncology. Including cancer and carcinogens RC254-282 Luca Pasquini Luca Pasquini Antonio Napolitano Martina Lucignani Emanuela Tagliente Francesco Dellepiane Maria Camilla Rossi-Espagnet Maria Camilla Rossi-Espagnet Matteo Ritrovato Antonello Vidiri Veronica Villani Giulio Ranazzi Antonella Stoppacciaro Andrea Romano Alberto Di Napoli Alberto Di Napoli Alessandro Bozzao AI and High-Grade Glioma for Diagnosis and Outcome Prediction: Do All Machine Learning Models Perform Equally Well?
description	Radiomic models outperform clinical data for outcome prediction in high-grade gliomas (HGG). However, lack of parameter standardization limits clinical applications. Many machine learning (ML) radiomic models employ single classifiers rather than ensemble learning, which is known to boost performance, and comparative analyses are lacking in the literature. We aimed to compare ML classifiers to predict clinically relevant tasks for HGG: overall survival (OS), isocitrate dehydrogenase (IDH) mutation, O-6-methylguanine-DNA-methyltransferase (MGMT) promoter methylation, epidermal growth factor receptor vIII (EGFR) amplification, and Ki-67 expression, based on radiomic features from conventional and advanced magnetic resonance imaging (MRI). Our objective was to identify the best algorithm for each task. One hundred fifty-six adult patients with pathologic diagnosis of HGG were included. Three tumoral regions were manually segmented: contrast-enhancing tumor, necrosis, and non-enhancing tumor. Radiomic features were extracted with a custom version of Pyradiomics and selected through Boruta algorithm. A Grid Search algorithm was applied when computing ten times K-fold cross-validation (K=10) to get the highest mean and lowest spread of accuracy. Model performance was assessed as AUC-ROC curve mean values with 95% confidence intervals (CI). Extreme Gradient Boosting (xGB) obtained highest accuracy for OS (74,5%), Adaboost (AB) for IDH mutation (87.5%), MGMT methylation (70,8%), Ki-67 expression (86%), and EGFR amplification (81%). Ensemble classifiers showed the best performance across tasks. High-scoring radiomic features shed light on possible correlations between MRI and tumor histology.
format	article
author	Luca Pasquini Luca Pasquini Antonio Napolitano Martina Lucignani Emanuela Tagliente Francesco Dellepiane Maria Camilla Rossi-Espagnet Maria Camilla Rossi-Espagnet Matteo Ritrovato Antonello Vidiri Veronica Villani Giulio Ranazzi Antonella Stoppacciaro Andrea Romano Alberto Di Napoli Alberto Di Napoli Alessandro Bozzao
author_facet	Luca Pasquini Luca Pasquini Antonio Napolitano Martina Lucignani Emanuela Tagliente Francesco Dellepiane Maria Camilla Rossi-Espagnet Maria Camilla Rossi-Espagnet Matteo Ritrovato Antonello Vidiri Veronica Villani Giulio Ranazzi Antonella Stoppacciaro Andrea Romano Alberto Di Napoli Alberto Di Napoli Alessandro Bozzao
author_sort	Luca Pasquini
title	AI and High-Grade Glioma for Diagnosis and Outcome Prediction: Do All Machine Learning Models Perform Equally Well?
title_short	AI and High-Grade Glioma for Diagnosis and Outcome Prediction: Do All Machine Learning Models Perform Equally Well?
title_full	AI and High-Grade Glioma for Diagnosis and Outcome Prediction: Do All Machine Learning Models Perform Equally Well?
title_fullStr	AI and High-Grade Glioma for Diagnosis and Outcome Prediction: Do All Machine Learning Models Perform Equally Well?
title_full_unstemmed	AI and High-Grade Glioma for Diagnosis and Outcome Prediction: Do All Machine Learning Models Perform Equally Well?
title_sort	ai and high-grade glioma for diagnosis and outcome prediction: do all machine learning models perform equally well?
publisher	Frontiers Media S.A.
publishDate	2021
url	https://doaj.org/article/41f565c7ad4b436b98ef650d7c372b72
work_keys_str_mv	AT lucapasquini aiandhighgradegliomafordiagnosisandoutcomepredictiondoallmachinelearningmodelsperformequallywell AT lucapasquini aiandhighgradegliomafordiagnosisandoutcomepredictiondoallmachinelearningmodelsperformequallywell AT antonionapolitano aiandhighgradegliomafordiagnosisandoutcomepredictiondoallmachinelearningmodelsperformequallywell AT martinalucignani aiandhighgradegliomafordiagnosisandoutcomepredictiondoallmachinelearningmodelsperformequallywell AT emanuelatagliente aiandhighgradegliomafordiagnosisandoutcomepredictiondoallmachinelearningmodelsperformequallywell AT francescodellepiane aiandhighgradegliomafordiagnosisandoutcomepredictiondoallmachinelearningmodelsperformequallywell AT mariacamillarossiespagnet aiandhighgradegliomafordiagnosisandoutcomepredictiondoallmachinelearningmodelsperformequallywell AT mariacamillarossiespagnet aiandhighgradegliomafordiagnosisandoutcomepredictiondoallmachinelearningmodelsperformequallywell AT matteoritrovato aiandhighgradegliomafordiagnosisandoutcomepredictiondoallmachinelearningmodelsperformequallywell AT antonellovidiri aiandhighgradegliomafordiagnosisandoutcomepredictiondoallmachinelearningmodelsperformequallywell AT veronicavillani aiandhighgradegliomafordiagnosisandoutcomepredictiondoallmachinelearningmodelsperformequallywell AT giulioranazzi aiandhighgradegliomafordiagnosisandoutcomepredictiondoallmachinelearningmodelsperformequallywell AT antonellastoppacciaro aiandhighgradegliomafordiagnosisandoutcomepredictiondoallmachinelearningmodelsperformequallywell AT andrearomano aiandhighgradegliomafordiagnosisandoutcomepredictiondoallmachinelearningmodelsperformequallywell AT albertodinapoli aiandhighgradegliomafordiagnosisandoutcomepredictiondoallmachinelearningmodelsperformequallywell AT albertodinapoli aiandhighgradegliomafordiagnosisandoutcomepredictiondoallmachinelearningmodelsperformequallywell AT alessandrobozzao aiandhighgradegliomafordiagnosisandoutcomepredictiondoallmachinelearningmodelsperformequallywell
_version_	1718406617038323712

AI and High-Grade Glioma for Diagnosis and Outcome Prediction: Do All Machine Learning Models Perform Equally Well?

Ejemplares similares