AI Evaluates Molecular Markers In Glioma Histopathology
A systematic review published in J Med Internet Res (2026) evaluated AI models from 2015–2025 that classify IDH mutation and 1p/19q codeletion using histopathology images. From 22 studies, pooled accuracy was 85.46%, sensitivity 84.55%, specificity 86.03%, and AUC 86.53%; hybrid and multimodal models performed best (up to 92.8% accuracy), and IDH detection outperformed 1p/19q classification. Findings recommend larger multimodal datasets and clinical integration with expert judgment.
Key Points
- 1Pooled evidence shows AI achieves 85.5% accuracy and 86.5% AUC for molecular classification.
- 2Hybrid and multimodal models outperform unimodal ones, reaching 92.8% accuracy and higher sensitivity.
- 3Practitioners should prioritize multimodal, hybrid AI frameworks and larger datasets for reliable IDH detection.
Scoring Rationale
High-quality systematic evidence with pooled metrics and PROSPERO registration, limited by small study count and no formal meta-analysis.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
