RADIOMICS AND ARTIFICIAL INTELLIGENCE IN GLIOMA GRADING: A PREDICTIVE MODELING LITERATURE REVIEW

Anastasia, Monica Cherlady
Astawa, Gregorius Adista Enrico

International Journal of Multidisciplinary Research and Literature (IJOMRAL)
Vol. 4 No. 2 (2025): INTERNATIONAL JOURNAL OF MULTIDISCIPLINARY RESEARCH AND LITERATURE

The integration of radiomics and artificial intelligence (AI) has revolutionized glioma grading by enhancing diagnostic accuracy through the analysis of complex imaging patterns. These techniques leverage radiomic features such as texture, shape, and intensity, analyzed by machine learning and deep learning models, to differentiate low- and high-grade gliomas with over 90% accuracy. However, challenges like the lack of standardized imaging protocols, model generalizability, and interpretability hinder clinical implementation. Potential solutions include multicentric collaborations, external validation, and explainable AI approaches. Future directions focus on combining radiomics with multi-omics data and developing hybrid CNN-Transformer architectures to enable more personalized therapies.