Deep learning based enhanced tumor segmentation approach for MR brain images

• Published in: Applied soft computing Vol. 78; pp. 346 - 354
• Main Authors: Mittal, Mamta, Goyal, Lalit Mohan, Kaur, Sumit, Kaur, Iqbaldeep, Verma, Amit, Jude Hemanth, D.
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.05.2019
• Subjects: Brain tumor MRI; Growing Convolution Neural Network (GCNN); Random forest; Segmentation; Stationary Wavelet Transform (SWT); Growing Convolution Neural Network (GCNN); Random forest; Brain tumor MRI; Segmentation; Stationary Wavelet Transform (SWT)
• DOI: 10.1016/j.asoc.2019.02.036
• ISSN: 1568-4946

Automation in medical industry has become one of the necessities in today’s medical scenario. Radiologists/physicians need such automation techniques for accurate diagnosis and treatment planning. Automatic segmentation of tumor portion from Magnetic Resonance (MR) brain images is a challenging task. Several methodologies have been developed with an objective to enhance the segmentation efficiency of the automated system. However, there is always scope for improvement in the segmentation process of medical image analysis. In this work, deep learning-based approach is proposed for brain tumor image segmentation. The proposed method includes the concept of Stationary Wavelet Transform (SWT) and new Growing Convolution Neural Network (GCNN). The significant objective of this work is to enhance the accuracy of the conventional system. A comparative analysis with Support Vector Machine (SVM) and Convolution Neural Network (CNN) is carried out in this work. The experimental results prove that the proposed technique has outperformed SVM and CNN in terms of accuracy, PSNR, MSE and other performance parameters. •A Growing Deep Convolutional Network (GCNN) is proposed for Magnetic Resonance brain image segmentation.•The proposed approach is different from the conventional deep learning approach.•Conventional CNN is also implemented for comparative analysis.•An extensive quantitative analysis is performed in terms of accuracy measures to validate the superior nature of the proposed approach.