CT-based deep learning radiogenomics for predicting key glioma genotypes (IDH, ATRX, EGFR, TP53)
Neuroradiology · 2026
DOI: 10.1007/s00234-026-04043-3 ↗Kurz gefasst
Fachartikel in der Zeitschrift Neuroradiology, 2026. Verfasst von Zohal Alnour Ahmed Emam und 5 weiteren Autor:innen. Beteiligte Einrichtung: Department of Medical Physics, Dokuz Eylul University, İzmir, Turkey. zohalalnourahmed.emam@ogr.deu.edu.tr. Hauptthemen laut MEDLINE: Brain Neoplasms, Deep Learning, Tomography, X-Ray Computed, Glioma. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 57 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Dokuz Eylül University.
Abstract (Originalsprache)
Purpose
Molecular subtyping guides diagnosis and targeted therapy for gliomas. Although MRI—the current imaging standard—can be time-consuming and is sometimes contraindicated, computed tomography (CT) is faster, more widely available, and often preferable in emergency and resource-limited settings. We evaluated whether CT-based radiogenomic signatures combined with machine learning could accurately predict clinically relevant glioma molecular markers.
Methods
In this retrospective study, we characterised non-contrast CT (NCCT) scans from 197 adults with histologically confirmed gliomas. Models were developed to predict mutations in four molecular markers: ATRX ( n =81), EGFR ( n =17), TP53 ( n =71), and IDH ( n =183). We extracted 208 quantitative radiomic features and added basic demographic variables. Feature selection used LASSO-RFE and Gradient Boosting–RFE with cross-validation. Six classical machine-learning classifiers and deep-learning approaches — including custom fully connected neural networks (FCNN) and TabNet — were trained and compared using ROC–AUC as the primary performance metric.
Results
Deep-learning methods outperformed conventional classifiers for all targets. TabNet achieved ROC–AUCs of 0.900 (95% CI: 0.717–0.989; ATRX), 0.955 (95% CI: 0.661–0.978; TP53), and 0.917 (95% CI: 0.858–0.975; EGFR). A custom FCNN obtained a ROC–AUC of 0.971 (95% CI: 0.876–0.995) for IDH. Cross-validation coefficients of variation were 2.0% for ATRX and 3.0% for TP53 and EGFR, and 16.0% for IDH. Deep-learning approaches yielded statistically significant improvements over conventional methods ( p -values ranging from
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Substanzen
Autor:innen und Institutionen
- Department of Medical Physics, Dokuz Eylul University, İzmir, Turkey. zohalalnourahmed.emam@ogr.deu.edu.tr.
- Department of Radiology, Dokuz Eylül University Hospital, İzmir, Turkey. emel.ada@deu.edu.tr.
- Department of Pathology, Dokuz Eylül University Hospital, İzmir, Turkey.
- Ataturk Vocational School of Health Services, Afyonkarahisar Health Sciences University, Afyon, Turkey.
- Department of Atomic and Molecular Physics, Dokuz Eylul University, İzmir, Turkey.
- Department of Electrical and Electronics Engineering, Dokuz Eylul University, İzmir, Turkey.
Förderung
- Dokuz Eylül University
Zitieren
APA
Emam, Z. A. A., Ada, E., Pehlivanoğlu, B., Çavuşoğlu, B., Akgüngör, K., & Selver, M. A. (2026). CT-based deep learning radiogenomics for predicting key glioma genotypes (IDH, ATRX, EGFR, TP53). Neuroradiology. https://doi.org/10.1007/s00234-026-04043-3
Vancouver
Emam ZAA, Ada E, Pehlivanoğlu B, Çavuşoğlu B, Akgüngör K, Selver MA. CT-based deep learning radiogenomics for predicting key glioma genotypes (IDH, ATRX, EGFR, TP53). Neuroradiology. 2026. doi:10.1007/s00234-026-04043-3
BibTeX
@article{Emam2026,
author = {Emam, Zohal Alnour Ahmed and Ada, Emel and Pehlivanoğlu, Burçin and Çavuşoğlu, Berrin and Akgüngör, Kadir and Selver, M Alper},
title = {CT-based deep learning radiogenomics for predicting key glioma genotypes (IDH, ATRX, EGFR, TP53)},
journal = {Neuroradiology},
year = {2026},
doi = {10.1007/s00234-026-04043-3}
}
EUVIMED