Related Experiment Video
Updated: Mar 15, 2026
A Fibrin-Enriched and tPA-Sensitive Photothrombotic Stroke Model
Published on: June 4, 2021
Automated CTA-Derived Collateral Grading and Morphologic Metrics for Enhanced Prediction of Post-Stroke Outcomes
Aditi Deshpande1, Jing Wang1, Krzysztof M Bochenek1
1From the University of California (A.D., K.L.), Riverside, CA; Division of Vascular Neurology and Neurocritical Care (J.W., K.M.B., L.A., S.Y., Z.B., D.M., G.K.H., P.T.-F.), Inova Neuroscience and Spine Institute, Inova Fairfax Medical Campus (IFMC), Falls Church, VA; Department of Medical Education (J.W., K.M.B., P.O., G.K.H., P.T.-F.), University of Virginia, IFMC Campus, Falls Chruch, VA; and Division of Neuroradiology (K.M.B., P.O.), IFMC, Falls Church, VA.
An automated collateral index (qCI) from CT angiography (CTA) accurately assesses collateral circulation in acute ischemic stroke (AIS) patients. This qCI tool improves stroke outcome prediction, especially where CT perfusion (CTP) is unavailable.
Area of Science:
- Neuroimaging
- Artificial Intelligence in Medicine
- Cerebrovascular Diseases
Background:
- Collateral circulation is crucial for acute ischemic stroke (AIS) treatment response but is subjectively assessed.
- CT perfusion (CTP) is limited by availability and artifacts, hindering practical clinical use.
- Accessible alternatives for collateral assessment are needed.
Purpose of the Study:
- To develop and validate an automated quantitative collateral index (qCI) using CT angiography (CTA).
- To evaluate CTA-derived features for predicting post-stroke recovery and functional outcomes.
- To compare CTA-based assessment with CT perfusion (CTP) metrics.
Main Methods:
- Retrospective analysis of 230 AIS patients undergoing endovascular thrombectomy (EVT).
- Deep learning U-Net segmentation for 3D vessel network generation and morphology metric computation from CTA.
- Automated qCI derivation via hemispheric comparison; gradient-boosted decision trees for outcome prediction using CTA, CTP, and combined data.
Main Results:
- Automated qCI demonstrated strong concordance with expert grading (accuracy 0.863, κ=0.786).
- CTA-only models outperformed CTP-only models in predicting 90-day functional outcomes (mRS).
- Combined CTA+CTP models yielded the best overall predictive performance for stroke outcomes.
Conclusions:
- Automated CTA-derived qCI offers rapid, objective collateral assessment, agreeing well with expert evaluation.
- CTA-based features are superior to CTP metrics for certain predictive tasks and enhance prognostic accuracy when combined with CTP.
- qCI is a scalable, practical tool for improving stroke outcome prediction, particularly in CTP-limited settings.
