Thyroid Tumor Diagnosis System using Spatial Fuzzy C-Means (SFCM) Classification Approach

This methodology consists of preprocessing of ultra sound thyroid images, feature computations and diagnosis stage. The preprocessing stage of the proposed method detects and reduces the noise contents in the source ultra sound thyroid images. Then, the texture features are computed from the preprocessed ultra sound thyroid image. Finally, these features are diagnosed into normal, mild and severe case using spatial Fuzzy-C-Mean classification (SFCM) approach. This method is tested on the ultra sound thyroid images in both DDTI and Open-CAS dataset with respect to accuracy, precision, recall and diagnosis rate