Skip to main content
Fig. 2 | Acta Neuropathologica Communications

Fig. 2

From: Interpretable deep learning of myelin histopathology in age-related cognitive impairment

Fig. 2

Weakly supervised classification predicts cognitive impairment based on whole slide image data from the hippocampus and frontal cortex. a Venn diagram showing the overlap of the measures used for defining the presence of cognitive impairment in brain donors. b Average receiver operating characteristic curves across tenfold cross-validation. Error envelopes show ± 1 standard deviation. Horizontal dotted lines show chance-level predictions. c Summary statistics for test evaluation of model performance across tenfold cross-validation in the frontal cortex and hippocampus. Balanced accuracy refers to the accuracy of predictions weighted by the proportion of labels in both groups in the test split. Horizontal lines are shown at the arithmetic mean values. d, e Probability estimates of cognitive impairment from the top-performing model by each measure of cognitive impairment in the hippocampus (d) and frontal cortex (e). CDR = Clinical Dementia Rating; MMSE = Mini-Mental State Examination; AUC = Area Under the Curve

Back to article page