Enhanced Curvlet Transform based Artificial Neural Network for Brain Tumor Diagnosis
M. Venkata Ramana1, E. Sreenivasa Reddy2, CH. Satayanarayana3
1M.Venkata Ramana, Assistant professor, Department of CSE, GITAM Deemed to be University, Visakahapatanam (A.P), India.
2E.Sreenivasa Reddy, Department of CSE, Nagarjuna University, Guntur (A.P), India.
3CH. Satayanarayana, Department of CSE, JNTUK University, Kakinada (A.P), India.
Manuscript received on 13 December 2018 | Revised Manuscript received on 22 December 2018 | Manuscript Published on 30 December 2018 | PP: 245-252 | Volume-8 Issue-2S, December 2018 | Retrieval Number: 100.1/ijeat. B10581282S18/18©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Brain tumor is one of the health problems faced by human beings. It often leads to death of people. Detecting it early can help in taking treatment and improve quality of life. The detection has to be made with MRI brain tumor images. Fourier transform, wavelet transform, Ridgelet transform and Curvelet transform are the techniques exist for representing images. Fourier transform can represent signals with only frequency domain and information on temporal domain is missing. To overcome this drawback, wavelet transform is used which can represent signal using wavelets in both time and frequency domains. However, wavelets are not good for images with different angles and different scales. Ridgelets could handle images with line singularities but could not handle images with curves. Curvlet transform can overcome this problem besides representing images with different scales and different angles. Curvlet Transform (CT) with enhancements can support dynamic texture classification for detection of brain tumor. Thus in this paper Enhanced CT (ECT) is used to have better diagnosis of brain tumor. A framework with underlying algorithms based on ECT is designed and implemented. A prototype application is built using MATLAB to demonstrate proof of the concept. The empirical results revealed that the proposed method has significant performance improvement over state of the art approaches.
Keywords: Curvlet Transform, Enhanced Curvelet Transform, Brain Tumor Detection Framework.
Scope of the Article: Neural Information Processing