Skip to main content

Table 4 Results of the proposed model versus models reported for the 2017 ISBI Challenge [29]

From: Melanoma diagnosis using deep learning techniques on dermatoscopic images

Organization ACC_M AUC_M SE_M SP_M Balance accuracy Overall score
eVida (proposed method M6) 0.904 0.872 0.820 0.925 0.872 0.848
RECOD Titans 0.872 0.874 0.547 0.950 0.749 0.792
Popleyi 0.858 0.870 0.427 0.963 0.695 0.762
Kazuhisa Matsunaga 0.828 0.868 0.735 0.851 0.793 0.798
Monty python 0.823 0.856 0.103 0.998 0.551 0.687
T D 0.845 0.836 0.350 0.965 0.658 0.726
Xulei Yang 0.830 0.830 0.436 0.925 0.681 0.708
Rafael Sousa 0.827 0.805 0.521 0.901 0.711 0.727
x j 0.843 0.804 0.376 0.957 0.667 0.710
Cristina Vasconcelos 0.830 0.791 0.171 0.990 0.581 0.660
Cristina Vasconcelos 0.825 0.789 0.171 0.983 0.577 0.658
Euijoon Ahn 0.805 0.786 0.009 0.998 0.504 0.614
Balázs Harangi 0.828 0.783 0.470 0.915 0.693 0.701
Matt Berseth 0.822 0.782 0.222 0.967 0.595 0.652
INESC Tecnalia 0.480 0.765 0.906 0.377 0.642 0.601
Dylan Shen 0.832 0.759 0.308 0.959 0.634 0.663
Vic Lee 0.832 0.757 0.308 0.959 0.634 0.665
Masih Mahbod 0.732 0.715 0.402 0.812 0.607 0.610
Dennis Murphree 0.760 0.684 0.231 0.888 0.560 0.574
Hao Chang 0.770 0.636 0.103 0.932 0.518 0.541
Jaisakthi S.M 0.748 0.623 0.419 0.828 0.624 0.614
Wenhao Zhang 0.805 0.500 0.000 1.000 0.500 0.581
Wiselin Jiji 0.503 0.495 0.470 0.511 0.491 0.433
Yanzhi Song 0.723 0.475 0.068 0.882 0.475 0.467
  1. Bold values indicate the most relevant and representatives in this rearch work
  2. ACC accuracy, AUC area under the RC curve, SE_M sensitivity, SP_M specificity for melanoma detection