Ways of Improving of Active Contour Methods in Colonoscopy Image Segmentation

Raneem Ismail; Szilvia Nagy

doi:10.5566/ias.2604

Authors

Raneem Ismail Széchenyi István University Department of Telecommunications https://orcid.org/0000-0002-3333-4205
Szilvia Nagy Széchenyi István University Department of Telecommunications https://orcid.org/0000-0001-9556-5095

DOI:

https://doi.org/10.5566/ias.2604

Keywords:

active contour segmentation methods, Chan–Vese method, colonoscopy image, geodesic method, Sørensen–Dice Similarity Coefficient

Abstract

As colonoscopy is the standard screening approach for colorectal polyps, and the first step of the correct classification and the efficient automatic diagnostics is the accurate detection and segmentation of the existing polyps, it is worth researching systematically, how colonoscopy databases are responding to two of the most influential variational segmentation methods, the geodesic and Chan–Vese active contour methods. Due to the quality variation of the colonoscopy databases, pre-processing steps are made. Then, 14 various filtered images are evaluated as different inputs for the active contour methods using the Sørensen–Dice Similarity Coefficient as a performance measurement metric. The effects of the initial mask shape and its size together with the number of iterations, contraction bias and smoothness factor are studied. In general, the Chan–Vese method showed more efficiency to match the actual contour of the polyp than the geodesic one with an initial mask possibly located within the polyp area. Preprocessing such as reflection removal, background subtraction and mean or median filtering can improve the Sørensen–Dice coefficient by up to 0.5.

References

Adams R, Bischof L (1994). Seeded region growing. IEEE Transactions on Pattern Analysis and Ma-chine Intelligence 16 (6): 641–7.

Bernal J, Sanchez F J, Vilarino F (2012). Towards Automatic Polyp Detection with a Polyp Appear-ance Model. Pattern Recognition 45: 3166–82.

Bernal J, Sánchez FJ, Fernández-Esparrach G, Gil D, Rodríguez C, Vilariño F (2015). WM-DOVA maps for accurate polyp highlighting in colonoscopy: Validation vs. saliency maps from physi-cians. Comput Med Imaging Graph. 43: 99-111.

Bhat S H, Kumar P (2019). Segmentation of Optic Disc by Localized Active Contour Model in Retinal Fundus Image. In M. C. Trivedi, et al. (Eds.), Smart Innovations in Communication and Computa-tional Sciences; Springer Verlag 851: 35–44.

Brice Claude R, Fennema Claude L (1970). Scene Analysis Using Regions. Technical Note 17. AI Center, SRI International, 333 Ravenswood Ave, Menlo Park, CA 94025, April.

Canny J (1986). A Computational Approach to Edge Detection. IEEE Transactions on Pattern Analysis and Machine Intelligence; 8(6):679–98.

Carass A, Roy S, Gherman A. et al (2020). Evaluating White Matter Lesion Segmentations with Re-fined Sørensen-Dice Analysis. Sci Rep 10: 8242.

Caselles V, Kimmel R, Sapiro G (1997). Geodesic Active Contours. International Journal of Computer Vision 22: 61–79.

Chan T, Vese L (2000). Image segmentation using level sets and the piecewise-constant Mumford-Shah model. Tech. Rep. 0014, Computational Applied Math Group.

Chan T, Vese L (2001). Active contours without edges. IEEE Transactions on Image Processing; 10 :266–277.

Dervieux A, Thomasset F (1979). A finite element method for the simulation of Rayleigh–Taylor in-stability. Rautman, R. (ed.) Approximation Methods for Navier–Stokes Problems. Lecture Notes in Mathematics Berlin; Springer, 771 : 145–158.

Dutta S, Sasmal P, Bhuyan M K, Iwahori Y (2018). Automatic Segmentation of Polyps in Endoscopic Image Using Level-Set Formulation. 2018 International Conference on Wireless Communications, Signal Processing and Networking (WiSPNET), Chennai; 1–5.

Fang L, Pan X, Yao Y.et al (2020). A hybrid active contour model for ultrasound image segmentation. Soft Comput; 24 : 18611–25.

Georgieva V, Petrov P (2017). An Approach for Colorectal Polyp Segmentation. Conference on Communication, Electromagnetics and Medical Application (CEMA'2017), Sofia, Bulgaria, Octo-ber.

Georgieva V, Petrov P, Nagy S, Sziová, B (2018). Detecting contours of pathological forms in colon-oscopy images using a hybrid method. Proceedings of 13th International Conference on Commu-nication, Electromagnetics and Medical Application (CEMA'2018) October 2018, Sofia, Bulgaria.

Ismail R, Nagy S (2021). On Metrics Used in Colonoscopy Image Processing for Detection of Colorec-tal Polyps. In: New Approaches for Multidimensional Signal Processing, Smart Innovation, Sys-tems and Technologies; Springer Singapore, Chapter 10, 216:137–51.

Juan S S, Aymeric H, Olivier R, Xavier D, Bertrand G (2014). Towards Embedded Detection of Polyps in WCE Images for Early Diagnosis of Colorectal Cancer. International Journal of Comput-er Assisted Radiology and Surgery, Springer Verlag (Germany) 9 (2): 283–93.

Kass M, Witkin A, Terzopoulos D (1988). Snakes: Active contour models. Int J Comput Vision; 1, 321–31.

Kichenassamy S, Kumar A, Olver P, Tannenbaum A, Yezzi A (1995). Gradient flows and geometric active contour models. Anon (Ed.), IEEE International Conference on Computer Vision; 810–15.

Mumford D, Shah J (1989). Optimal approximations by piecewise smooth functions and associated variational problems. Communications on Pure and Applied Mathematics 42(5): 577–685.

Nagy S, Sziová B, Pipek J (2019). On Structural Entropy and Spatial Filling Factor Analysis of Colonoscopy Pictures. Entropy 21(3): 256.

Nagy S, Sziová B, Solecki L (2020). The effect of background and outlier subtraction on the structural entropy of two-dimensional measured data. International Journal of Reasoning-based Intelligent Systems 12 (3): 200–9.

Nagy Sz, Lilik F, Kóczy L T (2017). Entropy based fuzzy classification and detection aid for colorec-tal polyps. IEEE Africon 2017, Cape Town, South Africa, September.

Nguyen N, Vo D M, Lee S (2020). Contour-Aware Polyp Segmentation in Colonoscopy Images Using Detailed Upsamling Encoder-Decoder Networks. IEEE Access 8: 99495–508.

Osher S, Sethian J A (1988). Fronts Propagation with Curvature Dependent Speed: Algorithms Based on Hamilton Jacobi Formulations. Journal of Computational Physics November 79(1): 12–49.

Pipek J, Varga I (1992). Universal classification scheme for the spatial-localization properties of one-particle states in finite, d-dimensional systems. Phys Rev A , 46: 3148–63.

Rényi A (1960). On measures of information and entropy. Proceedings of the fourth Berkeley Symposium on Mathematics, Statistics and Probability, Berkeley, CA, USA, 20 June–30 July 547–61.

Sasmal P, Iwahori Y, Bhuyan M K, Kasugai K (2018). Active contour segmentation of polyps in cap-sule endoscopic images. 2018 International Conference on Signals and Systems (ICSigSys) Bali 2018 201–4.

Shannon C E (1948). A mathematic theory of communication. Bell Syst. Tech. J. 27: 379–423.

Taha A A, Hanbury A (2015). Metrics for evaluating 3D medical image segmentation: analysis, selec-tion, and tool. BMC Med Imaging 15:29.

Yang X, Jiang X (2020). A Hybrid Active Contour Model based on New Edge-Stop Functions for Im-age Segmentation. International Journal of Ambient Computing and Intelligence; 11: 87–98.

Yuji I, Akira H, Yoshinori A, Bhuyan M, Robert J, Kunio K (2015). Automatic Detection of Polyp Us-ing Hessian Filter and HOG Features. 19th International Conference in Knowledge Based and In-telligent Information and Engineering Systems - KES2015, Procedia Computer Science 60: 730– 9.