Fig. 3From: Validation of machine learning models to detect amyloid pathologies across institutionsWhole Tissue CNN scores grouped by pathological diagnosis (Emory cohort). CNN scores generated from confidence heatmap processing are grouped into three distinct groups: control cases, pure AD (no secondary diagnosis), and all AD (pure AD + cases with secondary diagnosis of TDP and / or LBD). The all AD group is further divided into pure AD, AD+LBD, and AD+TDP ((c) and (f)). a-c Comparison for the cored CNN scores. d-f Comparisons for the diffuse CNN scores. AD: Alzhiemer’s disease, TDP: TDP-43 inclusions, LBD: Lewy body disease. For (a), (b), (d) and (e) a student’s 2-sided independent sample t-test was used to assess significance. For (c) and (f) an ANOVA with post-hoc analysis using Tukey’s test for multiple comparisons was used to assess significance. Alpha value of 0.05, significance is shown between groups with * for p-value less than 0.05, ** less than 0.01, *** less than 0.001, and **** less than 0.0001. Whiskers show the interquartile range of +/− 1.5*IQR. Outliers are shown as red + and medians are shown as horizontal red lines in the boxplots. Control group (n = 5), pure AD (n = 14), all AD (n = 30), AD+TDP (n = 8), and AD+LBD (n = 7)Back to article page