. Author manuscript; available in PMC: 2023 Apr 1.

Published in final edited form as: Comput Methods Programs Biomed. 2022 Aug 4;225:107058. doi: 10.1016/j.cmpb.2022.107058

Table 4.

Overall bolus localization performance. Best results for each metric are highlighted in bold. For the RMSE we reported the median and the inter-quartile range (in parentheses).

	CNN backbone	r_y	RMSE
CNN Models	VGG16	0.432	0.606 (0.400–0.917)
	InceptionV3	0.878	0.255 (0.146–0.430)
	ResNet50	0.934	0.306 (0.193–0.460)
	CNN3	0.837	0.189 (0.099–0.320)
	CNN4	0.886	0.230 (0.114–0.390)
Friedman’s test			X²(4) = 583.87, p < .001