Computer-aided diagnosis system for classifying benign and malignant thyroid nodules in multi-stained FNAB cytological images

摘要: An automated computer-aided diagnosis system is developed to classify benign and malignant thyroid nodules using multi-stained fine needle aspiration biopsy (FNAB) cytological images. In the first phase, image segmentation performed remove background staining information retain appropriate foreground cell objects in images mathematical morphology watershed transform methods. Subsequently, statistical features are extracted two-level discrete wavelet (DWT) decomposition, gray level co-occurrence matrix (GLCM) Gabor filter based The classifiers k-nearest neighbor (k-NN), Elman neural network (ENN) support vector machine (SVM) tested for classifying nodules. combination of segmentation, GLCM k-NN classifier results a lowest diagnostic accuracy 60 %. highest 93.33 % achieved by ENN trained with bank from segmented It also observed that SVM its 90 % DWT along experimental suggest FNAB would be useful identifying cancer irrespective protocol used.