AI Insight
Researchers developed CerebAI, an artificial intelligence system that classifies brain CT scans into three categories: no stroke, ischemic stroke, and hemorrhagic stroke. The system achieved 97.47% accuracy on a dataset of 6,774 CT scans and uses Integrated Gradients to provide visual explanations showing which brain regions influenced its decisions. CerebAI outperformed several established AI models and includes features designed for potential clinical deployment, including the ability to process medical imaging files directly.
Why it matters
Rapid and accurate stroke classification is critical because ischemic and hemorrhagic strokes require opposite treatments—giving the wrong treatment can be fatal. An AI system that not only classifies strokes accurately but also shows clinicians exactly which brain regions it identified as abnormal could improve emergency decision-making and build physician trust in AI-assisted diagnosis.
Understand the Science
⚠️ Preprint – Noch nicht peer-reviewed
Dieser Artikel wurde noch nicht von unabhängigen Experten begutachtet. Die Ergebnisse sind vorläufig und sollten mit Vorsicht interpretiert werden.
Stroke is a leading cause of death and long-term disability worldwide, affecting approximately 15 million individuals annually. Prompt and accurate subtype differentiation between ischemic and hemorrhagic stroke is clinically critical, as the two conditions demand diametrically opposite interventions – thrombolytic therapy versus surgical decompression. Yet the majority of existing deep learning approaches reduce this problem to binary detection, and virtually none address the opacity of their decision-making in a clinically actionable manner. We present CerebAI, an explainable, deployment-oriented three-class CT stroke classification system built on a fine-tuned ConvNeXt-Base backbone with Integrated Gradients (IG) attribution. Trained on 6,774 non-contrast CT scans stratified across No Stroke, Ischemic Stroke, and Hemorrhagic Stroke, CerebAI achieves a weighted F1-score of 0.9746 (95% CI: [0.9625, 0.9851]), accuracy of 97.47%, macro-averaged AUC of 0.9921, mean Intersection-over-Union (mIoU) of 0.9276, Expected Calibration Error (ECE) of 0.0115, mean Brier Score of 0.0150, and Cohen’s {kappa} of 0.9483 – surpassing ResNet-50, EfficientNet-B4, and Vision Transformer (ViT-B/16) baselines across all reported metrics. Integrated Gradients produce pixel-precise saliency maps that localize pathological regions with greater anatomical fidelity than Gradient-weighted Class Activation Mapping (Grad-CAM), a finding we support with side-by-side qualitative comparison. CerebAI additionally incorporates a native DICOM processing pipeline to facilitate future clinical translation. Code and model weights are publicly available to support reproducibility and further research.
Source: CerebAI: Explainable Three-Class Stroke CT Classification via ConvNeXt and Integrated Gradients