AI and high-grade glioma for diagnosis and outcome prediction: Do all machine learning models perform equally well? Journal Article


Authors: Pasquini, L.; Napolitano, A.; Lucignani, M.; Tagliente, E.; Dellepiane, F.; Rossi-Espagnet, M. C.; Ritrovato, M.; Vidiri, A.; Villani, V.; Ranazzi, G.; Stoppacciaro, A.; Romano, A.; Di Napoli, A.; Bozzao, A.
Article Title: AI and high-grade glioma for diagnosis and outcome prediction: Do all machine learning models perform equally well?
Abstract: 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. Copyright © 2021 Pasquini, Napolitano, Lucignani, Tagliente, Dellepiane, Rossi-Espagnet, Ritrovato, Vidiri, Villani, Ranazzi, Stoppacciaro, Romano, Di Napoli and Bozzao.
Keywords: immunohistochemistry; survival; cancer survival; controlled study; human tissue; protein expression; gene mutation; major clinical study; overall survival; genetics; histopathology; radiation dose; nuclear magnetic resonance imaging; brain tumor; glioma; sensitivity and specificity; ki 67 antigen; gene amplification; epidermal growth factor receptor; immunoreactivity; retrospective study; dna methylation; histology; glioblastoma; methylated dna protein cysteine methyltransferase; observational study; genetic screening; dna extraction; receiver operating characteristic; human experiment; isocitrate dehydrogenase; head and neck squamous cell carcinoma; tumor necrosis; diagnostic test accuracy study; apparent diffusion coefficient; machine learning; human; article; radiomics; high-grade glioma (hgg); isocitrate dehydrogenase (nad)
Journal Title: Frontiers in Oncology
Volume: 11
ISSN: 2234-943X
Publisher: Frontiers Media S.A.  
Date Published: 2021-11-01
Start Page: 601425
Language: English
DOI: 10.3389/fonc.2021.601425
PROVIDER: scopus
PMCID: PMC8649764
PUBMED: 34888226
DOI/URL:
Notes: Article -- Export Date: 3 January 2022 -- Source: Scopus
Altmetric
Citation Impact
BMJ Impact Analytics
MSK Authors