Neuroanesthesia & neurocritical care
Explainable Machine Learning for Coma Outcome Prediction Based on Structural and Functional Brain MRI
Authors' conclusion
A new ensemble of brain MRI-derived metrics was specifically related to coma state, its etiology, and the patient's potential for recovery at 3 months.
Study context
- Population
- Patients with coma (GCS at admission ≤ 9) of traumatic or anoxic origin, 3 ICUs, Toulouse (France)64 coma patients · 55 controls
- Participants
- 26 / 38 / 55
- Exposure
- Patients with coma vs controls
- Traumaticn = 26
- Anoxicn = 38
- Controlsn = 55reference
- Outcome
- ML model performance
Methodological detailsmulticentre, cross-validated explainable machine learning (ML)
- Multicentre
- Cross-validated explainable ML
Main finding
As reported in the abstract
Coma diagnosis (mean accuracy)
0.934%
Primary brain injury discrimination (mean accuracy)
0.762 %
Neurologic outcome prediction (mean accuracy)
0.824 %
Observational data: associations, not effects.
Further findings
Unfavorable outcome · 3 months
- Patients with unfavorable outcome (32/64)
- 50%
Delay between ICU admission and MRI scan
- Median delay, days (IQR 6-16 d)
- 9
Parts of this Brief are withheld: the automatic checks could not confirm them against the abstract.
Original publication
Explainable Machine Learning for Coma Outcome Prediction Based on Structural and Functional Brain MRI. Crit Care Med. Published 27 February 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →