Clinical Report: Deep Learning-Based Automation of Evans Index Assessment

Overview

This study evaluates the reliability of automated Evans Index (EI) measurement using deep learning in patients with acute subarachnoid hemorrhage (SAH). The findings indicate that automated EI measurements demonstrate high reproducibility and meaningful agreement with expert assessments.

Background

The Evans index is a critical tool for assessing ventricular enlargement in SAH. Manual EI measurement is often time-consuming and subject to variability.

Data Highlights

Key Findings

• Automated EI measurement using TotalSegmentator (TS) shows excellent reproducibility (ICC = 0.996).
• High agreement between expert readers for manual EI measurements (ICC = 0.983).
• Good agreement between TS and expert EI measurements (ICC = 0.76), improving to 0.87 after excluding frontal horn hematoma.
• TS identified a higher percentage of EI > 0.30 cases compared to expert assessment (29% vs. 17%).
• TS-derived EI demonstrated significant discrimination for EVD placement (AUC = 0.75).
• TS-derived EI remained independently associated with EVD placement after adjusting for clinical covariates (adjusted OR = 1.09).

Clinical Implications

The findings indicate that automated EI measurement may improve the consistency of ventricular enlargement assessments in acute SAH.

Conclusion

Automated EI measurement using deep learning provides a reproducible assessment tool for ventricular enlargement in acute SAH.

Related Resources & Content

1. Frontiers in Neurology, 2026 -- Automated Evans Index Measurement Using Deep Learning in Acute Subarachnoid Hemorrhage: Reliability, Agreement with Experts, and Association with External Ventricular Drainage
2. npj Digital Medicine — Automated real-time assessment of intracranial hemorrhage detection AI using an ensembled monitoring model (EMM)
3. Factors Influencing Shunt Placement in Patients with Aneurysmal Subarachnoid Hemorrhage: A Retrospective Study from a Single Institution
4. Frontiers in Neurology — External ventricular drainage in modern neurosurgical practice: optimization, standardization, and emerging guidance technologies
5. npj Digital Medicine — An end-to-end deep learning pipeline for hematoma expansion prediction in spontaneous intracerebral hemorrhage based on non-contrast computed tomography
6. 2023 Guideline for the Management of Patients With Aneurysmal Subarachnoid Hemorrhage
7. External ventricular drainage in modern neurosurgical practice: optimization, standardization, and emerging guidance technologies
8. Acute care of aneurysmal subarachnoid hemorrhage: practical consensus statement from a multidisciplinary group of German-speaking neurointensivists and neuroradiologists on behalf of the DIVI neurology section | Neurological Research and Practice | Springer Nature Link
9. Effectiveness of Lumbar Cerebrospinal Fluid Drain Among Patients With Aneurysmal Subarachnoid Hemorrhage: A Randomized Clinical Trial | Trials | JAMA Neurology | JAMA Network
10. Effectiveness of Cerebrospinal Fluid Lumbar Drainage Among Patients with Aneurysmal Subarachnoid Hemorrhage: An Updated Systematic Review and Meta-Analysis - ScienceDirect
11. Idiopathic Normal Pressure Hydrocephalus - StatPearls - NCBI Bookshelf
12. Interobserver reliability of the DESH score in idiopathic chronic hydrocephalus - ScienceDirect
13. Deep Learning-Based Models for Ventricular Segmentation in Hydrocephalus: A Systematic Review and Meta-Analysis - ScienceDirect
14. Frontiers | Automated Evans Index Measurement Using Deep Learning in Acute Subarachnoid Hemorrhage: Reliability, Agreement with Experts, and Association with External Ventricular Drainage