AI-Powered CT Perfusion Analysis for Stroke

Abstract

Time is brain. In acute ischemic stroke, every minute without treatment means the loss of 1.9 million neurons. CT Perfusion (CTP) imaging provides the critical information clinicians need to make life-saving treatment decisions. This project presents an advanced AI-powered pipeline that automates CTP analysis in under 90 seconds.

The Clinical Challenge

Acute ischemic stroke occurs when a blood clot blocks an artery supplying the brain. Modern treatments like mechanical thrombectomy can remove these clots, but success depends on answering: Is there still salvageable brain tissue?

CT Perfusion imaging reveals two critical tissue types:

Ischemic Core: Irreversibly damaged tissue

Penumbra: Tissue with compromised perfusion but maintained viability—the therapeutic target

Pipeline Architecture: Seven Stages

1. Intelligent DICOM Processing

Automatic 4D volume reconstruction from potentially out-of-order slices

2. Brain Tissue Segmentation

Combination of density thresholding, connected component analysis, and morphological operations

3. Motion Correction with Deep Learning

Custom neural network predicts motion parameters in 15-25 seconds

4. Arterial and Venous Input Function Detection

Multi-strategy approach with U-Net segmentation for anatomical guidance

5. FFT-Based Deconvolution

Frequency-domain processing with Wiener filtering for perfusion extraction

6. Perfusion Parameter Extraction

CBF, CBV, MTT, and Tmax computation for clinical decision-making

7. Tissue Classification

Automated core and penumbra segmentation with clinical thresholds

Deep Learning Components

Motion Correction Network: Encoder-decoder architecture with keypoint representations

MCA Segmentation U-Net: Restricts AIF selection to anatomically appropriate regions

Sinus Segmentation U-Net: ResNet50 backbone for venous structure identification

Clinical Impact

The system delivers accurate tissue classification in under 90 seconds—fast enough to guide emergency thrombectomy decisions, potentially saving countless lives through faster, more accurate stroke treatment.