Santhi, Ramalingam (2018) DISCRIMINATION OF PAPILLARY AND MEDULLARY CARCINOMAS OF THYROID NODULES USING STATISTICAL FEATURES DRIVEN SUPPORT VECTOR MACHINE CLASSIFIER. International Journal of Advance Research in Science and Engineering, 7 (3). ISSN 2319-8354

[thumbnail of 30.pdf] Text
30.pdf - Published Version
Download (200kB)
Official URL: http://www.ijarse.com

Abstract

In this work, discrimination of two major types of carcinomas of thyroid nodule, namely, papillary and
medullary carcinomas is addressed. The region based watershed segmentation technique is initially utilized to
segment the medullary and papillary carcinoma cell regions in multi-stained Thyroid Fine Needle Aspiration
Biopsy (FNAB) cytological images. It removes the background stain information from the images and retains
the required foreground malignant cell information. Second order statistical features are, then, extracted from
the segmented images using two-level Discrete Wavelet Transform (DWT) decomposition. The significant
statistical features are used as meaningful descriptors by the Support Vector Machine (SVM) classifier for
discriminating papillary and medullary carcinomas of thyroid nodules. The support vector machine classifier
results highest promising discrimination accuracy of 95%. Also, the developed automated thyroid cancer
diagnostic system produces 90% of sensitivity and 100% of specificity with respect to papillary and medullary
thyroid carcinoma images.

Item Type: Article
Uncontrolled Keywords: Carcinoma, Discrimination, Malignancy, Segmentation, Thyroid, Watershed
Depositing User: Mr Team Mosys
Date Deposited: 18 Aug 2022 03:35
Last Modified: 18 Aug 2022 03:35
URI: http://ir.psgcas.ac.in/id/eprint/1487

Actions (login required)

View Item
View Item
Altmetric