Table 3.

Distance metrics for human annotators and computational technique

Comparison	Case Number	Fraction=GT	Average Distance <GT	Average Distance >GT
RP1 versus GT, scheme: T, data: patients	48	0.56	−1.19±0.48	2±0 (1 case)
RP2 versus GT, scheme: T, data: patients	48	0.8	−1.4±0.52	2±0 (2 cases)
C versus GT, scheme: T, data: patients	54	0.5	−0.61±0.57	0.57±0.60
C versus GT, scheme: T, data: sections	121	0.52	−0.65±0.63	0.66±0.71
C versus GT, scheme: F, data: patients	33	0.65	−0.36±0.4	0.40±0.47
C versus GT, scheme: F, data: sections	85	0.98	−0.08±0.2	0.01±0.04
Baseline versus GT, scheme: T, data: patients	54	0.27	−1.79±0.88	1.9±0.96

Distance is defined as the difference of the assigned label minus the ground truth label. Negative distances indicate undercalling; positive distances indicate overcalling. T refers to classifications according to the Tervaert scheme; F refers to classifications according to the Fogo scheme. Values are reported as mean±SD taken over all of the cases. Data description identifies whether the experiment was performed using separate patients or separate sections as individual data. GT, ground truth; RP, renal pathologist; C, computer.