Fig. 3
From: Deep learning reveals disease-specific signatures of white matter pathology in tauopathies
Tau burden is disease specific. Scatter plot of tau burden in cortex (x-axis) vs WM (y-axis) regions across multiple WSI (individual data points) from different diseases (point colors). Tau burden in a region was estimated by the ratio of area covered by tau aggregates (from aggregate classifier) to total area of the region (WM/cortex from the region classifier). Data from each disease are fit using a least squares straight-line fit and the best-fit slope is shown next to the fitted line