Detection of ischemic changes on baseline multimodal computed tomography: expert reading vs. Brainomix and RAPID...

Area of Science:

• Neurology
• Radiology
• Medical Imaging

Background:

• Early detection of ischemic changes in acute ischemic stroke is crucial for timely treatment and improved patient outcomes.
• Non-contrast CT (NCCT) and CT perfusion (CTP) are key imaging modalities for evaluating ischemic stroke.
• Assessing early ischemic changes aids in predicting final infarct volume and guiding therapeutic decisions.

Purpose of the Study:

• To compare the accuracy of expert interpretation and automated software (e-ASPECTS) for assessing ischemic changes on NCCT and CTP.
• To evaluate the diagnostic performance of different CTP parameters (CBF<30%, Tmax>10s) in predicting final infarcts.
• To determine the accuracy of these methods for final infarct prediction in anterior circulation ischemic stroke.

Main Methods:

• ASPECTS scoring was applied to baseline neuroimaging of 81 anterior circulation ischemic stroke patients.
• Ischemic changes were assessed by two expert raters on NCCT, by e-ASPECTS software on NCCT, and visually on CTP maps.
• Accuracy metrics (sensitivity, specificity, PPV, NPV) were calculated by comparing baseline findings with 24-hour follow-up infarcts.

Main Results:

• Median ASPECTS was 9 across all modalities.
• Overall accuracy for infarct prediction was 0.76 (e-ASPECTS), 0.79 (consensus), 0.82 (CBF<30%), and 0.80 (Tmax>10s).
• CTP parameters (CBF<30%, Tmax>10s) demonstrated the highest sensitivity and negative predictive value compared to NCCT-based assessments.

Conclusions:

• CT perfusion parameters, particularly CBF<30% and Tmax>10s, offer high accuracy in assessing ischemic changes and predicting final infarcts.
• Automated software (e-ASPECTS) and expert consensus on NCCT provide valuable but slightly less accurate assessments.
• These findings suggest that CTP, especially with automated analysis, has significant potential to enhance the detection and prediction of ischemic changes in stroke.