SDD LV segmentation for comparison with DL and CNN methods

LV segmented by SDD and localized by CHT, The disclosed method achieved DICE score 95.44% on the ACDC test case

%%%%%Automated LV segmentation by SDD threshold selection and CHT
%%%%%Run Demo.m to get the qualitative and quantitative results on the ACDC test set. For more details please refer to the following papers:
%[1]: ZiHao Wang and ZhenZhou Wang, Fully Automated Segmentation of the Left Ventricle in Magnetic Resonance Images,
%[2]: ZhenZhou Wang, "Automatic and optimal segmentation of the left ventricle in cardiac magnetic resonance images independent of the training sets," in IET Image Processing, vol. 13, no. 10, pp. 1725-1735, 22 8 2019, doi: 10.1049/iet-ipr.2018.5878.
%[3] ZhenZhou Wang, "Automatic Localization and Segmentation of the Ventricles in Magnetic Resonance Images," in IEEE Transactions on Circuits and Systems for Video Technology, doi: 10.1109/TCSVT.2020.2981530.
%%By the way, We tested two deep learning methods with disclosed trained
%%models and their DICE scores are only 90.28 % and 87.13 %, which are 5% less than ours.
We have tried to find all the ACDC competitors' trained models to reproduce their segmentation results, unfortunately, we found nothing. The ACDC organizers only checked their submitted results instead of checking their codes or software. Hence, their claimed accuracies are meaningless.