DERMA: A Melanoma Diagnosis Platform Based on Collaborative Multilabel Analog Reasoning
Table 1
Sensitivity, specificity, and accuracy results obtained in melanocytic, melanoma, and BCC classification through the different DERMA challenges: the noncollaborative CBR classification which only uses dermoscopy data, the noncollaborative CBR classification which only uses confocal data, the plain collaborative system, the collaborative system that enhances the collaboration with preprocessing rules, and the collaborative system with a DML organized case memory. Each result shows the -test comparison between the result obtained on this DERMA configuration in comparison with the previous one using 95% of confidence level. This is presented with an (↑) if it is significantly better and (—) if there is no significant difference.
