Classification of Digital Mammograms Using Nearest Neighbor Techniques
Endah Purwanti, Retna Apsari · International Journal of Computer Trends and Technology · 2014
The aim of our research is to classify digital mammograms into two classes, abnormal microcalcification and normal. Texture is one of the major mammographic characteristics. The statistical textural of Gray Level Coocurrence Matrix (GLCM) used in characterizing images are contrast, energy and entropy. K-Nearest Neighbor (K-NN) and Fuzzy K-Nearest Neighbor (FK-NN) was proposed for classifying images. The result of K-NN method shows that 77.78% accuracy , 50.00% sensitifity and 100% specifisity. The result of FK-NN method shows that 88.89% accuracy , 100% sensitifity and 80.00% specifisity. .