Skip to main content
Fig. 4 | Acta Neuropathologica Communications

Fig. 4

From: Deep learning reveals disease-specific signatures of white matter pathology in tauopathies

Fig. 4

Aggregates from different diseases show distinct shapes and sizes. Individual aggregates were identified in the WM of WSI and were characterized by features describing their size and shape. a Example image patch (left) from a WSI with individual aggregates colored based on sample features: area, eccentricity, and minor axis length. b Individual WSI (rows) were characterized based on median feature values (columns) of aggregates in their WM and were ordered based on hierarchical clustering. Values within each feature were z-score normalized to allow comparison across features. Colors on top (Red/blue/green) indicate disease associated with each WSI. c Boxplots of median feature values for area, eccentricity, and minor axis length shown in WM regions of WSI (gray dots) compared across each disease. Mann–Whitney test, with Bonferroni-based multiple-hypothesis testing correction (across diseases) was used for statistical comparison

Back to article page