Automated Evans index measurement using deep learning in acute subarachnoid hemorrhage: reliability…

AI / informaticsFrontiers in neurology· 2026

Automated Evans index measurement using deep learning in acute subarachnoid hemorrhage: reliability, agreement with experts, and association with external ventricular drainage.

Yanrui Cai, Huansong Wang, Huanhuan Yu … Feiyu Qiao

NeurotraumaAI / informaticsQ2IF 3.1

Automated Evans index remained independently linked to external ventricular drain placement

Automated Evans index measurement using deep learning in acute subarachnoid hemorrhage: reliability, agreement with experts, and association with external ventricular drainage.

Yanrui Cai … Feiyu Qiao

Frontiers in neurology · 2026 · Q2 in Neurology

Background

The reliability of automated evans index measurement in acute subarachnoid hemorrhage remains insufficiently validated.

Methods

In this retrospective cohort study, admission non-contrast CT scans from 364 patients with spontaneous subarachnoid hemorrhage were analyzed using TotalSegmentator (TotalSegmentator), an open-source nnU-Net-based deep learning pipeline, to generate automated evans index measurements.

n = 364 patients

Results

higher automated Evans index still predicted external ventricular drainage placement after adjustment

OR 1.09 (95% CI 1.03 to 1.17)

CI excludes the null - significant

More results

TotalSegmentator demonstrated excellent reproducibility between repeated inference runs (ICC = 0.996, 95% CI 0.996-0.997).

Agreement between expert readers was high (ICC = 0.983, 95% CI 0.978-0.988).

More results

Between-method agreement between TotalSegmentator and expert evans index measurements was good in the overall cohort (ICC = 0.76, 95% CI 0.73-0.81) and improved after exclusion of frontal horn hematoma (ICC = 0.87, 95% CI 0.85-0.89).

“

Conclusion · 1 of 3

Automated evans index measurement using TotalSegmentator provides reproducible and clinically informative assessment of ventricular enlargement on CT in acute subarachnoid hemorrhage.

Conclusion · 2 of 3

Although threshold-sensitive disagreement occurred near evans index = 0.30, automated evans index showed meaningful agreement with expert assessment and remained independently associated with contemporaneous external ventricular drainage decision-making.

Conclusion · 3 of 3

Further subarachnoid hemorrhage-specific refinement may improve robustness in hemorrhage-related ventricular distortion.

Automated Evans index measurement using deep learning in acute subarachnoid hemorrhage: reliability, agreement with experts, and association with external ventricular drainage.

Automated Evans index measurement using deep learning in acute subarachnoid hemorrhage: reliability, agreement with experts, and association with external ventricular drainage.

Related papers