Automated Computer Aided Diagnosis Using Altered Multi-Phase Level Sets in Application to Categorize the Breast Cancer Biopsy Images

Abstract

As macroscopic Breast Cancer (BC) imaging diagnostics do not provide satisfied data, microscopic image analysis is used to get optimum cell-nuclei segmentation for image categorization. The paper presents a Novel Hybrid Segmentation Method which uses the energy detail coefficients attained from Non-Subsampled Contourlet Transform for cell-nuclei segmentation of BC histopathological images. NHSM solves the local inhomogeneities problem while preserving all necessary details. The system evaluation is done using training dataset of 98 BC histopathological images and testing dataset of 26 images. Total of 25 segmented features was extracted for classification using Decision Trees, k-Nearest Neighbor, multilayer perceptron, Multiclass-Support Vector Machine (M-SVM). The proposed NHSM segmentation results were undergone through both subjective and objective performance evaluation. Result evaluation proved that NHSM along with Multi Class-Support Vector Machine classifier provides highly precise and accurate classified result for BC tissue images compared to state-of-the-art segmentation methods. Also, the edge detection model initialization is highly robust and insensitive.

KEYWORDS:

• Active Contour Model
• Breast Cancer tissue images
• Classification
• Multi-phase level sets
• Non-subsampled Contourlet Transform

Acknowledgements

The authors thank Dr. Uppada Ramesh, Asst. Professor, Rangaraya Medical College for providing high resolution H&E stained breast cancer histopathological Images.

Additional information

Funding

The authors greatly acknowledge the Department of Science & Technology, Government of India for providing the financial support from July 2015 to June 2018 with their vide sanction reference No. SR/WOS-A/ET-1025/2014, under Women Scientist Scheme to carry out this work.

Notes on contributors

Rajyalakshmi Uppada

Rajyalakshmi Uppada received her PhD from JNTUK, Kakinada in 2018. She completed Master of Technology from Andhra University, Visakhapatnam in 2009. Currently, she is working as a professor of ECE Department in Aditya Engineering College (AEC), Surampalem, AP. Under her credit, she received funding for one major project from WOS-A, DST, New Delhi at JNTUK and one MODROB project from AICTE, India to carry out her research work at AEC. She also received one seminar grant from NCW, New Delhi. Her research interests include digital image processing, machine learning and communications. She has published 12 international research papers and presented 6 international conference papers.

Satya Prasad Kodati

Satya Prasad Kodati obtained PhD in the area of digital signal processing from IIT Madras, Chennai in 1989. He is currently working as Sr. Professor in Electronics & Communication Department & Rector, Vignan's Foundation for Science, Technology & Research, Vadlamudi, Guntur. He retired as a professor in Electronics & Communication Department and director-IST, JNTU Kakinada, India. He has 45 years of teaching experience, 25 years of research experience, guided 35 PhD scholars and guiding 20 students for PhD. He has held different positions as Head of ECE Department, vice principal, principal of UCEK, JNTUK, and as director of Evaluation and Rector at JNTUK, Kakinada. Email: [email protected]

Sanagapallela Koteswara Rao

Sanagapallela Koteswara Rao obtained his PhD in digital statistical signal processing from AU Visakhapatnam, India in 1998. He retired as a scientist ‘G’, associate director, DRDO-NSTL. He published 30 IEEE papers. With his thirty two years of design and development experience and expertise in the Anti Submarine Warfare (ASW) Fire Control Systems for torpedoes and rocket launchers, he created a strong edifice in the weapon technology in the country at NSTL. He is currently serving as professor, ECE Department, KL University, India. His research interests include statistical signal processing and adaptive signal processing. Email: [email protected]