Dana-Farber Cancer Institute has developed an AI model, MANTIS, that uses DNA methylation profiling to individualize prognosis and guide therapy for children with brain tumors.
Key Details
- 1MANTIS leverages DNA methylation data and neural networks for survival prediction in pediatric medulloblastoma and ependymoma.
- 2The model outperformed traditional risk prediction methods across several large patient cohorts.
- 3MANTIS enables time-dependent, personalized risk estimates rather than broad grouping.
- 4Analysis suggests over half of children with average-risk medulloblastoma could safely receive lower-dose craniospinal irradiation, potentially preserving cognitive function.
- 5DNA methylation assays are already routine in tumor testing, increasing the model’s clinical feasibility.
- 6Study published in Nature Genetics on November 3, 2026; led by Dana-Farber researchers and supported by prominent pediatric cancer foundations.
Why It Matters
This work demonstrates how integrating AI with established molecular diagnostics can refine prognostication and increase treatment personalization for pediatric brain tumor patients, potentially minimizing long-term side effects. It may serve as a model for extending AI-based risk stratification to other cancers using routinely acquired clinical data.
Source
EurekAlert
Related News
•EurekAlert
Mayo Clinic Unveils AI-Enabled Clinical Decision Support Platform
Mayo Clinic Platform launches an AI-enabled Clinical Decision Support solution to streamline clinician workflows and provide evidence-based, individualized patient care.
•EurekAlert
Quantum Machine Learning Shows Promise for Lung Cancer Prediction with Limited Data
FAU researchers demonstrate quantum machine learning can accurately predict lung cancer from sparse clinical data sets.
•EurekAlert
AI Outperforms Clinicians in Predicting Survival from Heart MRI
AI analysis of cardiac MRI is better linked to patient survival than clinician measurements, suggesting a key role for AI in heart function assessment.