Prognostication of cardiac arrest survivors using low apparent diffusion coefficient cluster volume.

Area of Science:

• Neuroimaging
• Neurology
• Radiology

Background:

• Cardiac arrest (CA) survivors require accurate neuroprognostication for optimal care.
• Current methods for predicting neurological outcomes post-CA have limitations.
• Diffusion-weighted MRI offers potential for quantitative assessment of brain injury.

Purpose of the Study:

• To develop and validate a novel neuroprognostication method for out-of-hospital cardiac arrest (CA) patients.
• To assess the performance of a new quantitative MRI metric, DC-LADCV, in predicting long-term neurological outcomes.
• To evaluate the method's efficacy in a multicenter setting.

Main Methods:

• Retrospective analysis of adult CA patients from three centers undergoing MRI within 12 hours of resuscitation.
• Extraction of quantitative MRI parameters including average ADCs, LADCV, and DC-LADCV across various ADC thresholds.
• Evaluation of diagnostic performance using Area Under the Receiver Operating Characteristic Curve (AUROC) and sensitivity at 100% specificity for poor neurological outcome (CPC score >2 at 6 months).

Main Results:

• The novel DC-LADCV metric demonstrated superior performance compared to average ADCs and LADCV, achieving a maximum AUROC of 0.955.
• DC-LADCV showed high sensitivity (>80%) for predicting poor neurological outcomes across a broad range of ADC thresholds (400-580 × 10(-6) mm(2) s(-1)).
• The method was validated across multiple centers, indicating robust generalizability.

Conclusions:

• Quantitative analysis of the relative volume of the most dominant cluster of low ADC voxels (DC-LADCV) is a highly effective neuroprognostication tool for out-of-hospital CA patients.
• The DC-LADCV method offers accurate and reliable prediction of neurological outcomes in a multicenter setting.
• This approach represents a significant advancement in post-CA neuroprognosis, aiding clinical decision-making.