The role of artificial intelligence in predicting the clinical outcomes associated with different therapeutic approaches for vestibular schwannoma: A systematic review and meta-analysis.

Introduction: Vestibular schwannoma is the most common neoplasm located at the skull base. The therapeutic strategy for managing vestibular schwannoma is formulated based on individual patient characteristics and specific imaging findings. Recently, there has been a growing interest in applying artificial intelligence (AI) to predict treatment outcomes in the field of neurosurgical oncology.

Aim: This systematic review and meta-analysis aims to assess the efficacy of AI algorithms in predicting outcomes associated with various therapeutic strategies for vestibular schwannoma.

Method and material: The study was conducted under PRISMA guidelines, involving comprehensive data extraction from multiple databases, specifically PubMed, Scopus, Embase, Web of Science, and the Cochrane Library, until January 31, 2025. Statistical analyses were performed using Comprehensive Meta-analysis (CMA) software version 3.0.

Results: This systematic review and meta-analysis included data from 21 studies. AI algorithms achieved an area under the curve of 0.80 in predicting microsurgery outcomes, with an accuracy (positive predictions regardless of whether they are positive or negative) of 81.5% and sensitivity (true positive rate) of 83%. In subgroup analysis, AI showed better accuracy for forecasting facial function than for hearing preservation following microsurgery. For tumor control after radiosurgery, the AUC was 0.722 with an accuracy of 58.5%, while predicting tumor progression after conservative management yielded an AUC of 0.912 and 87.5% accuracy.

Conclusion: AI algorithms can be valuable prognostic tools for evaluating outcomes across therapeutic interventions. Nonetheless, further prospective studies are essential to establish the optimal model for clinical application.