Detection of Specular Reflection from Smart Colposcopy Image using RGB Color Space and Convolutional Neural Network

Detection of Specular Reflection from Smart Colposcopy Image using RGB Color Space and Convolutional Neural Network
  IJETT-book-cover           
  
© 2023 by IJETT Journal
Volume-71 Issue-10
Year of Publication : 2023
Author : Jennyfer Susan M B, Subashini P, Krishnaveni M, Indhumathi T
DOI : 10.14445/22315381/IJETT-V71I10P204

How to Cite?

Jennyfer Susan M B, Subashini P, Krishnaveni M, Indhumathi T, "Detection of Specular Reflection from Smart Colposcopy Image using RGB Color Space and Convolutional Neural Network," International Journal of Engineering Trends and Technology, vol. 71, no. 10, pp. 29-38, 2023. Crossref, https://doi.org/10.14445/22315381/IJETT-V71I10P204

Abstract
Cervical cancer is a prevalent malignancy among women, particularly in developing nations with high fatality rates. Screening for cervical cancer often involves the use of smart colposcopy, which captures images of the cervix. However, specular reflection frequently affects these images, resulting in bright pixels that obscure neoplasm on the cervical images. To address this issue, proposed an intensity-based threshold method on RGB color space to identify specular reflection in smart colposcopy images. The binary masks are generated using this threshold, with glare regions represented as 1 and non-glare regions as 0, allowing for automated identification of specular reflection using convolutional neural networks such as the fully convolutional network and the U-Net Model. The proposed approach achieved an accuracy of 89.26% in identifying specular reflection using the threshold method alone. However, the combination of novel binary masking along with U-Net improved the accuracy to 99.71%. This study demonstrates the potential for the proposed method to improve the accuracy of identifying specular reflection in smart colposcopy images, which could ultimately enhance screening for cervical cancer in resource-limited settings.

Keywords
Cervical cancer, Specular reflection, Smart colposcopy image, Convolution Neural Network, RGB color space.

References
[1] Ganesh Balasubramaniam et al., “Survival Rate of Cervical Cancer from A Study Conducted in India,” Indian Journal of Medical Sciences, vol. 72, no. 1, pp. 203-211, 2020.
[CrossRef] [Google Scholar] [Publisher Link]
[2] Cervical Cancer, 2022. [Online]. Available: https://www.who.int/health-topics/cervical-cancer#tab=tab_1
[3] C.O. Iloghalu et al., “An Intelligent Model for Improved Breast Cancer Prognosis,” SSRG International Journal of Electronics and Communication Engineering, vol. 10, no. 8, pp. 36-47, 2023.
[CrossRef] [Publisher Link]
[4] Rengaswamy Sankaranarayanan et al., “Visual Inspection of the Uterine Cervix After the Application of Acetic Acid in the Detection of Cervical Carcinoma and Its Precursors,” Cancer, vol. 83, no. 10, pp. 2150-2156, 1998.
[CrossRef] [Google Scholar] [Publisher Link]
[5] Steven A. Shafer, “Using Color to Separate Reflection Components,” Color Research and Application, vol. 10, no. 4, pp. 210-218, 1985.
[CrossRef] [Google Scholar] [Publisher Link]
[6] T.Gevers, and H.M. Stokman, “Classifying Color Transitions Into Shadow-Geometry, Illumination, Highlight or Material Edges,” International Conference on Image Processing, vol. 1, pp. 521-524, 2000.
[CrossRef] [Google Scholar] [Publisher Link]
[7] V. Kudva, K. Prasad, and S. Guruvare, “Detection of Specular Reflection and Segmentation of Cervix Region in Uterine Cervix Images for Cervical Cancer Screening,” Intermediate-Range Ballistic Missile, vol. 38, no. 5, pp. 281-291, 2017.
[CrossRef] [Google Scholar] [Publisher Link]
[8] Othmane El Meslouhi et al., “Automatic Detection and Inpainting of Specular Reflections for Colposcopic Images,” Open Computer Science, vol. 1, pp. 341-354, 2011.
[CrossRef] [Google Scholar] [Publisher Link]
[9] Xiaoxia Wang et al., “Detection and Inpainting of Specular Reflection in Colposcopic Images with Exemplar-Based Method,” 13th International Conference on Anti-Counterfeiting, Security, and Identification, pp. 90-94, 2019.
[CrossRef] [Google Scholar] [Publisher Link]
[10] Abhishek Das, Avijit Kar, and Debasis Bhattacharyya, “Elimination of Specular Reflection and Identification of ROI: The First Step in Automated Detection of Cervical Cancer Using Digital Colposcopy,” IEEE International Conference on Imaging Systems and Techniques, pp. 237-241, 2011.
[CrossRef] [Google Scholar] [Publisher Link]
[11] Holger Lange, “Automatic Glare Removal in Reflectance Imagery of the Uterine Cervix,” Medical Imaging 2005: Image Processing, pp. 2183-2192, 2005.
[CrossRef] [Google Scholar] [Publisher Link]
[12] V. Pallavi, and K. Payal, “Automated Analysis of Cervix Images to Grade the Severity of Cancer,” Annual International Conference of the IEEE Engineering in Medicine and Biology Society, pp. 3439-3442, 2011.
[CrossRef] [Google Scholar] [Publisher Link]
[13] Muhammad Asif et al., “Intrinsic Layer Based Automatic Specular Reflection Detection in Endoscopic Images,” Computers in Biology and Medicine, vol. 128, no. 1, 2020.
[CrossRef] [Google Scholar] [Publisher Link]
[14] Mojtaba Akbari et al., “Adaptive Specular Reflection Detection and Inpainting in Colonoscopy Video Frames,” 25th IEEE International Conference on Image Processing, pp. 3134-3138, 2018.
[CrossRef] [Google Scholar] [Publisher Link]
[15] Leanne Attard et al., “Specular Highlights Detection Using a U-Net Based Deep Learning Architecture,” Fourth International Conference on Multimedia Computing, Networking and Applications, pp. 4-6, 2020.
[CrossRef] [Google Scholar] [Publisher Link]
[16] Yodit Abebe Ayalew, Kinde Anlay Fante, and Mohammed Aliy Mohammed, “Modified U-Net for Liver Cancer Segmentation from Computed Tomography Images with a New Class Balancing Method,” BMC Biomedical Engineering, vol. 3, no. 4, pp. 1-13, 2021.
[CrossRef] [Google Scholar] [Publisher Link]
[17] Juš Lozej et al., “End-to-End Iris Segmentation using U-Net,” IEEE International Work Conference on Bioinspired Intelligence, pp. 1-6, 2018.
[CrossRef] [Google Scholar] [Publisher Link]
[18] Bambang Krismono Triwijoyo, and Ahmat Adil, “Analysis of Medical Image Resizing using Bicubic Interpolation Algorithm,” Jurnal Ilmu Komputer, vol. 14, no. 2, pp. 20-29, 2021.
[Google Scholar] [Publisher Link]
[19] Olaf Ronneberger, Philipp Fischer, and Thomas Brox, “U-Net: Convolutional Networks for Biomedical Image Segmentation,” Medical Image Computing and Computer-Assisted Intervention-MICCAI 2015: 18th International Conference, pp. 234-241, 2015.
[CrossRef] [Google Scholar] [Publisher Link]
[20] Jonathan Long, Evan Shelhamer, and Trevor Darrell, “Fully Convolutional Networks for Semantic Segmentation,” IEEE Conference on Computer Vision and Pattern Recognition, pp. 3431-3440, 2015.
[Google Scholar] [Publisher Link]
[21] Baoxian Yu et al., “Specular Highlight Detection Based on Color Distribution for Endoscopic Images,” Frontier in Physics, vol. 8, 2021.
[CrossRef] [Google Scholar] [Publisher Link]
[22] Jian Yang Yefei Gao et al., “Dynamic Searching and Classification for Highlight Removal on Endoscopic Image,” Procedia of Computer Science, vol. 107, no. 2017, pp. 762-767.
[CrossRef] [Google Scholar] [Publisher Link]
[23] Kaggle, 2017. [Online]. Available: https://www.kaggle.com/competitions/intel-mobileodt-cervical-cancer-screening/overview/