Stroke outcome and evolution prediction from CT brain using a spatiotemporal diffusion autoencoder

Researchers developed a spatiotemporal diffusion autoencoder using CT brain images to predict stroke outcomes and evolution. The AI model achieved best-in-class performance for predicting next-day severity and functional outcomes using a dataset of 5,824 CT images from 3,573 patients across two medical centers.

• →New AI model uses diffusion probabilistic models to create self-supervised stroke representations from CT images.
• →Method extends to accommodate longitudinal images and time from stroke onset for improved predictions.
• →Model achieved best performance for predicting next-day severity and functional outcome at discharge.
• →Research utilized minimal labels across 5,824 CT images from 3,573 patients at two medical centers.
• →Approach could revolutionize stroke care by enabling individualized clinical decision-making.