Emerging role for radiomics in the management of intracranial arteriovenous and cavernous malformations

Intracranial arteriovenous malformations (AVMs) and cavernous malformations (CMs) pose substantial diagnostic, prognostic, and therapeutic challenges. Traditional imaging techniques used for AVM/CM diagnostic and treatment decision-making are limited by subjectivity and reliance on human interpretation. Radiomics, an artificial intelligence-driven technique that extracts quantitative imaging biomarkers, is a promising tool for improving detection, risk assessment, and treatment planning. The use of radiomic data for diagnosis, clinical course prediction, and outcome forecasting in patients with AVMs/CMs is reviewed, following a comprehensive search of the PubMed database for iterative terms of 'radiomics,' 'arteriovenous malformations,' and 'cavernous malformations.' Radiomic techniques demonstrate high diagnostic accuracy for differentiating AVM-related hemorrhages from other causes. Additionally, radiomic models have shown promise in predicting AVM rupture risk, epilepsy occurrence, and response to radiosurgery. In limited studies, radiomics have also shown utility in distinguishing CMs from other intracranial lesions and predicting CM hemorrhage risk. Radiomics may enhance personalized neurosurgical decision-making and patient outcomes for AVMs and CMs. Ongoing technological refinements, iterative testing, and addressing barriers to equitable access to this technology will be critical for widespread application.