Table 8.

Results for renal cell carcinoma baseline and KD approaches as percentages with corresponding 95% CIs. Baseline models were trained from scratch until convergence on the corresponding magnification. Boldface text indicates the best-performing model for each magnification and metric

	Renal Cell Carcinoma
	Baseline	KD
mag = 0.625x (16 μm/pixel)
Accuracy	82.39 (79.80–85.02)	85.11 (82.45–87.83)
F1-Score	62.21 (55.38–68.97)	66.41 (59.35–73.31)
Precision	61.38 (54.99–67.89)	69.66 (63.31–75.99)
Recall	64.81 (57.33–72.37)	68.68 (61.32–75.66)
mag = 1.25x (8 μm/pixel)
Accuracy	87.44 (84.77–90.06)	89.11 (86.54–91.61)
F1-Score	73.73 (66.99–80.17)	77.10 (70.91–82.99)
Precision	72.38 (65.91–78.88)	75.66 (69.99–81.28)
Recall	76.73 (69.62–83.27)	82.64 (76.48–88.01)
mag = 2.5x (4 μm/pixel)
Accuracy	88.24 (85.72–90.76)	88.39 (85.74–90.92)
F1-Score	73.94 (67.87–79.84)	75.84 (69.57–81.71)
Precision	73.94 (67.87–79.84)	75.34 (69.70–80.85)
Recall	79.78 (73.25–85.74)	81.72 (76.06–86.76)