Lung Nodule Detection: Image Enhancement using Fuzzy Rule Based Contrast Limited Adaptive Histogram Equalization and Entropy Weighted Residual Convolution Neural Network Method in CT
K.S. Gowri laksshmi1, R.Umagandhi2
1K.S. Gowri laksshmi MSc, M.Phil, , Research scholar, Department of Computer Science, Kongunadu Arts and Science College, Coimbatore, Tamil Nadu, India.
2Dr.R.Umagandhi MCA, M.Phil, PhD, Associate Professor and Head Department of Computer Technology, Kongunadu Arts and Science College, Coimbatore, Tamil Nadu, India.
Manuscript received on November 20, 2019. | Revised Manuscript received on November 28, 2019. | Manuscript published on 30 November, 2019. | PP: 7496-7502 | Volume-8 Issue-4, November 2019. | Retrieval Number: D5328118419/2019©BEIESP | DOI: 10.35940/ijrte.D5328.118419
Open Access | Ethics and 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: Computed Tomography (CT) images are read by several lung nodule detection methods. The early step of contrast enhancement is mandatory because of low contrast in original image and further techniques of image processing are with unsatisfactory results. Hence this process are resulted an enhanced image of clearly discrete lung area from background. Image enhancement, feature extraction, and classification are three primary steps. In this work, Rule based Contrast Limited Adaptive Histogram Equalization (FRCLAHE) perform image enhancement step followed by feature extraction and Fuzzy Rule (FR) determines the contrast value. From rules upper contrast value are determined then image is enhancement from CLAHE. In the second, the feature extraction is conducted using the Fuzzy Continuous Wavelet Transform (FCWT) and Gray Level Feature Extraction (GLCM). After this step, the classification is completed using the Entropy Weighted Residual Convolution Neural Network (EWRCNN). Finally, the results are evaluated between the samples, compared to FP reduction with Faster R-CNN alone, the inclusion of rule‐based classification lead to an improvement in detection accuracy for the CAD system. These preliminary results demonstrate the feasibility of the proposed EWRCNN approach to lung nodule detection and FP reduction on CT images.
Keywords: lung Nodule Detection, T2-weighted MR images, Faster R-CNN, FP reduction, Fuzzy Rule based Contrast Limited Adaptive Histogram Equalization Fuzzy Continuous Wavelet Transform, Gray Level Feature Extraction, Entropy Weighted Residual Convolution Neural Network.
Scope of the Article: Fuzzy Logics.