Wavelet Features Extraction for Medical Image Classification
Amir Rajaei, Lalitha Rangarajan · 2011
Abstratct: In this paper, we present a method for classification of medical images. Wavelet features of different modalities of medical images are extracted. Then mean and standard deviation of extracted wavelet features are computed. We utilize K-Nearest Neighbor classifier to classify medical imaging modalities as X-ray, MRI and CT. Experiments are conducted on medical database containing 4,500 images. We achieve 99.96 % classification accuracy which presents the efficiency of our proposed approach. Keywords: Wavelet, Central Moments, K-Nearest Neighbor, Medical Image Classification 1.