Automated Tissue Classification Framework for Reproducible Chronic Wound Assessment

Table 3

Performance evaluation of various classifiers for wound tissue classification.

Statistical learning schemes	Tissue-wise accuracy (%)			Overall accuracy (%)	Kappa statistic
	Granulation	Slough	Necrotic
Bayesian classifier	86.48	78.27	78.72	81.15	0.704
SVM with linear kernel	82.88	86.47	48.93	72.76	0.653
SVM with 2nd polynomial kernel	81.98	88.69	73.40	81.35	0.718
SVM with 3rd polynomial kernel	87.84	90.90	79.78	86.13	0.793
SVM with RBF kernel	82.88	88.91	74.46	80.08	0.697