Brain MRI-based prognostication after cardiac arrest: qualitative assessment outperforms variable voxel-wise ADC...

Area of Science:

• Neurology
• Radiology
• Critical Care Medicine

Background:

• Quantitative apparent diffusion coefficient (ADC) analysis is explored for predicting neurological outcomes post-cardiac arrest.
• Existing studies show inconsistent optimal thresholds for ADC, hindering clinical application.

Purpose of the Study:

• To evaluate the prognostic value of voxel-wise ADC thresholds for neurological outcomes in cardiac arrest survivors.
• To compare quantitative ADC thresholds with qualitative MRI visual assessments for outcome prediction.

Main Methods:

• A cohort study of 261 comatose cardiac arrest patients treated with targeted temperature management.
• Analysis included subgroup evaluations based on arrest aetiology, MRI timing, and acquisition protocols.
• Primary outcome was poor neurological outcome (Cerebral Performance Category score 3-5).

Main Results:

• Percentage of brain voxels (PV) at 400 and 450 × 10^-6 mm²/s showed strong discriminative performance (AUC 0.86).
• Qualitative visual MRI assessment achieved higher AUC (0.91) with perfect specificity and superior sensitivity.
• Quantitative ADC thresholds exhibited significant variability across subgroups and MRI protocols.

Conclusions:

• Qualitative visual MRI assessment is a reliable independent and multimodal prognostic predictor.
• Quantitative ADC thresholds demonstrate limitations due to variability, challenging the use of a universal threshold.
• Further research may refine quantitative methods or emphasize multimodal approaches including qualitative assessment.