A Retrospective Cohort Study
This groundbreaking project explores the use of artificial intelligence for early detection and classification of ischemic strokes from MRI images, with a focus on improving healthcare access and outcomes in low- and middle-income countries (LMIC).
Understanding the critical need for accessible stroke detection technology
Advanced machine learning algorithms analyze MRI scans to identify and classify ischemic stroke patterns with high accuracy, supporting clinical decision-making.
Stroke is a leading cause of death and disability worldwide. In LMIC settings, limited access to specialists makes early detection challenging—our AI aims to bridge this gap.
This study brings together leading institutions and researchers in neurology, radiology, and artificial intelligence to advance stroke care globally.
Ischemic stroke accounts for approximately 87% of all strokes and requires rapid diagnosis for effective treatment. However, many LMIC healthcare systems face shortages of trained radiologists and neurologists, leading to delayed or missed diagnoses. This retrospective cohort study investigates whether AI-powered computer-aided detection systems can accurately identify and classify ischemic strokes from MRI images in resource-limited settings.
This study is a collaborative effort between leading medical institutions, including university hospitals, research centers in neurology and radiology, and AI research labs. Our multidisciplinary team includes neurologists, radiologists, data scientists, and public health experts committed to improving stroke outcomes worldwide.
Explore our published research, conference presentations, and ongoing findings
A comprehensive analysis of convolutional neural network architectures for automated detection of acute ischemic stroke from diffusion-weighted MRI sequences.
Preliminary results from our retrospective cohort study demonstrating the feasibility and accuracy of AI-powered stroke detection in LMIC healthcare facilities.
Evaluation of transfer learning techniques for improving stroke subtype classification accuracy with limited training datasets.
Multi-center validation comparing AI system performance against expert radiologist interpretations across diverse LMIC healthcare settings.
Have questions about our research? Interested in collaboration? Get in touch with our team.