Automatic Skin Tumor Detection Using Online Tiger Claw Region Based Segmentation – A Novel Comparative Technique

Abstract

Skin tumor acts as a premier factor for high death rate throughout the world. It is difficult for the radiologists to segment the skin tumour cells. Various research work focus on accurate segmentation but not on the time of processing. The intention of this research work is to provide an efficient enhancement method and tumor detection from other unaltered regions. This work relies mainly on computed tomography (CT) tumor images of the skin, benign or malignant, that has been implemented efficiently. In this research paper, a novel methodology called Online Tiger Claw Region Based Segmentation (OTCRBS) is proposed which is used mainly to detect the boundary of unaffected Skin Cell, similar to tiger which uses its claws to tear off the skin of its prey during the search for its food. By using metric for the region, various properties can be formulated for the detection of anomalous skin cells. 98.68 and 97.71% accuracy is produced for procurement of benign and malignant nodule in MATLAB 2018a, respectively. Computation time was only 7.65 s. Comparative analysis is made with different segmentation methods. Experimental results establish that the proposed flow outperforms all the existing segmentation methods for the proper detection of tumor cells.

KEYWORDS:

• Computed tomography
• Proposed online tiger claw region based segmentation algorithm
• Skin tumor

Acknowledgements

This file contains the input datasets that are obtained from the ISIC – 2017 dataset of Skin with tumor and dataset from International Cancer Centre (ICC), Neyyoor, Tamil Nadu, India. The authors thank Arunachala College of Engineering for Women, for their support at this work. The authors also acknowledge for the creation of the free public available ISIC skin tumor database and International Cancer Centre (ICC), Neyyoor, Tamil Nadu for providing skin tumor database used in this research work. Finally, we thank anonymous reviewers for helping to strengthen this paper. In this paper, AA as investigator, analyzed the key concepts of the proposed research work with the coding of the Mapping concepts and segmentation of Tumor cells in Skin tumor images. This paper was designed, directed and coordinated by the co-authors.

Additional information

Notes on contributors

A. Ashwini

Ashwini A, ME, PhD received her BE degree in ECE and ME degree in communication and networking from Anna University, Chennai. She is University rank holder in UG and PG. She is currently pursuing her PhD degree in Anna University Chennai, India. She has published many papers and participated in some international conferences. Her research interests include medical image processing, nanotechnology, and image segmentation. Corresponding author. Email: [email protected]

V. Kavitha

Kavitha V obtained her BE degree in CSE (1996) from MS University and ME degree in CSE (2000) from Madurai Kamaraj University. She is University rank holder in UG and gold medalist in PG. She received her PhD in computer science and engineering from Anna University Chennai in 2009. Presently, she is working as professor in the Department of CSE at the University College of Engineering, Kancheepuram. Her research interests are wireless networks,wireless sensor networks, image processing; cloud computing. She has published many papers in national and international journals in areas such as network security, mobile computing, wireless network security and cloud computing. Email: [email protected]