Skip to main content
. 2017 Dec 7;12(12):e0186425. doi: 10.1371/journal.pone.0186425

Fig 2. Comparison for Kmeans and XHC for different number of clusters.

Fig 2

With increasing number of clusters, Kmeans resulted in more clusters around the wound as this area represented the major part of the tissue. XHC further highlighted the middle part of the images, which was the result of a hierarchical decomposition of the signatures. This allowed a better investigation and comparison of the wound tissue and healing progress.