Kadhim, Mustafa
Persson, Emilia
Haraldsson, André
Gustafsson, Christian Jamtheim
Nilsson, Mikael
Kügele, Malin
Bäck, Sven
Ceberg, Sofie
Funding for this research was provided by:
Lund University
Article History
Received: 25 June 2025
Accepted: 8 October 2025
First Online: 15 October 2025
Declarations
:
: The authors declare no competing interests.
: To compare the impact of different loss functions on each image similarity metric in Fig. 5c, we first confirmed non-normality using the Shapiro–Wilk test (p < 0.05). We then applied the nonparametric Kruskal–Wallis test as an omnibus assessment of differences among groups, followed by Dunn’s post-hoc pairwise comparisons with Bonferroni correction to identify which pairs differed significantly. All analyses were performed in Python using SciPy (v.1.13.1) and Scikit-posthocs (v.0.11.4.).