Related Experiment Video
Updated: Jul 17, 2025
05:21
Computerized Adaptive Testing System of Functional Assessment of Stroke
Published on: January 7, 2019
5.8K
Performance evaluation of two different software programs for automated ASPECTS scoring in patients with suspected
Carolina Río Bártulos1, Lukas Pirl1, Dennis Lier1
1Insitut für Röntgendiagnostik und Nuklearmedizin, Städtisches Klinikum Braunschweig gGmbH, Braunschweig, Germany.
Clinical Hemorheology and Microcirculation
|August 28, 2023
Summary
AI software for stroke assessment shows varying performance. eASPECTS demonstrated better agreement and correlation with human raters than VIA_ASPECTS, though VIA_ASPECTS had higher sensitivity for detecting ASPECTS scores.
Area of Science:
- Radiology
- Artificial Intelligence
- Stroke Imaging
Background:
- Advancements in Artificial Intelligence (AI) have significantly expanded the software options available to radiologists.
- AI-powered software for assessing the ASPECTS score in stroke patients is a key development.
Purpose of the Study:
- To compare the clinical performance of two AI software packages, eASPECTS and VIA_ASPECTS, in a real-world setting.
- To evaluate the software's utility in a 24/7 stroke unit at a maximum care facility.
Main Methods:
- Analysis of 637 noncontrast CT scans from patients with suspected stroke.
- Comparison of software performance against human raters, assessing agreement, correlation, sensitivity, specificity, and accuracy.
Main Results:
- eASPECTS exhibited good agreement and strong correlation with human raters.
- VIA_ASPECTS demonstrated lower accuracy and specificity compared to eASPECTS, but higher sensitivity.
Conclusions:
- Both eASPECTS and VIA_ASPECTS show potential as decision support tools for radiologists in stroke assessment.
- Differences exist in the intended use and performance characteristics of the two software packages.
