Abstract
An effective and early polypectomy of adenomatous and serrated polyps may dramatically reduce the incidence of colorectal cancer. Recently, Narrow-band imaging (NBI) sequences have emerged to enhance the description of these polyp types from the description of their microvascular and surface textural patterns. Despite the major observation capabilities, the in-situ analysis from the colonoscopy procedure remains challenging due to dependence on the expertise of gastroenterologists to localize and characterize polyps. This work introduces a robust frame-level strategy that achieves a full characterization of polyp patterns to differentiate among serrated, adenoma, and hyperplastic samples. The proposed strategy learns a deep convolutional representation that supports classification but also retrieves attention maps to localize main regions associated with the lesion. From this deep representation, it was also possible to build a low-dimensional representation space that allows visualizing a particular frame-video sample w.r.t to other diagnosed samples. From a total of 76 public available colonoscopies, the proposed strategy achieves an average classification accuracy of 90.79%. Besides, the proposed approach achieves a remarkable classification of polyps to be resected and also the masses diagnosed as serrated, a task with major diagnostic variability among experts.
Graphic Abstract
![](http://media.springernature.com/lw685/springer-static/image/art%3A10.1007%2Fs12553-021-00633-8/MediaObjects/12553_2021_633_Figa_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs12553-021-00633-8/MediaObjects/12553_2021_633_Fig1_HTML.jpg)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs12553-021-00633-8/MediaObjects/12553_2021_633_Fig2_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs12553-021-00633-8/MediaObjects/12553_2021_633_Fig3_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs12553-021-00633-8/MediaObjects/12553_2021_633_Fig4_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs12553-021-00633-8/MediaObjects/12553_2021_633_Fig5_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs12553-021-00633-8/MediaObjects/12553_2021_633_Fig6_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs12553-021-00633-8/MediaObjects/12553_2021_633_Fig7_HTML.png)
Similar content being viewed by others
Notes
http://www.image-net.org/
References
Hyuna S, et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2021;71(3):209–49.
Ortega-Morán JF, et al. Medical needs related to the endoscopic technology and colonoscopy for colorectal cancer diagnosis. BMC Cancer. 2021;21(1):1–12.
Kudo S, et al. Colorectal tumours and pit pattern. J Clin Pathol. 1994;47(10):880–5.
El Hajjar A, Jean-François R. Artificial intelligence in gastrointestinal endoscopy: general overview. Chinese Med J. 2020;133(3):326.
Van D, Sascha C, et al. Polyp morphology: an interobserver evaluation for the Paris classification among international experts. Official J Am College of Gastroenterology ACG. 2015;110(1):180–7.
Stehle T, et al. Classification of colon polyps in NBI endoscopy using vascularization features. Medical Imaging 2009: Computer-Aided Diagnosis. Vol. 7260. International Society for Optics and Photonics, 2009.
Mesejo P, et al. Computer-aided classification of gastrointestinal lesions in regular colonoscopy. IEEE Trans Med Imaging. 2016;35(9):2051–63.
Rogart JN, et al. Narrow-band imaging without high magnification to differentiate polyps during real-time colonoscopy: improvement with experience. Gastrointest Endosc. 2008;68(6):1136–45.
Viovan II, et al. The role of narrow band imaging in colorectal polyp detection. Bosn J Basic Med Sci. 2017;17(2):152.
Kazuhiro G, et al. Appearance of enhanced tissue features in narrow-band endoscopic imaging. J Biomed Opt. 2004;9(3):568–77.
Yoko K, et al. Computer-aided diagnosis of colorectal polyp histology by using a real-time image recognition system and narrow-band imaging magnifying colonoscopy. Gastrointest Endosc. 2016;83(3):643–9.
Koichi O, et al. Clinicopathological characteristics of serrated polyps as precursors to colorectal cancer: current status and management. J Gastroenterol Hepatol. 2017;32(2):358–67.
Yuichi M, et al. Computer-aided diagnosis for colonoscopy. Endoscopy. 2017;49(08):813–9.
Yusuke H, et al. Convolutional neural network for differentiating gastric cancer from gastritis using magnified endoscopy with narrow band imaging. Dig Dis Sci. 2020;65(5):1355–63.
Tischendorf JJW, et al. Computer-aided classification of colorectal polyps based on vascular patterns: a pilot study. Endoscopy. 2010;42(03):203–7.
Sebastian G, et al. Computer-based classification of small colorectal polyps by using narrow-band imaging with optical magnification. Gastrointest Endosc. 2011;74(6):1354–9.
Lan L, et al. Convolutional neural network for the diagnosis of early gastric cancer based on magnifying narrow band imaging. Gastric Cancer. 2020;23(1):126–32.
Hiroya U, et al. Application of artificial intelligence using a convolutional neural network for diagnosis of early gastric cancer based on magnifying endoscopy with narrow-band imaging. J Gastroenterol Hepatol. 2021;36(2):482–9.
Peng-Jen C, et al. Accurate classification of diminutive colorectal polyps using computer-aided analysis. Gastroenterology. 2018;154(3):568–75.
Henrik T, et al. Serrated polyps-a concealed but prevalent precursor of colorectal cancer. Scand J Gastroenterol. 2017;52(6–7):654–61.
Ruikai Z, et al. Automatic detection and classification of colorectal polyps by transferring low-level CNN features from nonmedical domain. IEEE J Biomed Health Inform. 2016;21(1):41–7.
Ribeiro E, Andreas U, Michael H. Colonic polyp classification with convolutional neural networks. 2016 IEEE 29th International Symposium on Computer-Based Medical Systems (CBMS). IEEE, 2016.
Haumaier F, Sterlacci W, Vieth M. Histological and molecular classification of gastrointestinal polyps. Best Pract Res Clin Gastroenterol. 2017;31(4):369–79.
Gregor U, et al. Deep learning localizes and identifies polyps in real time with 96% accuracy in screening colonoscopy. Gastroenterology. 2018;155(4):1069–78.
Kim DH, et al. Serrated polyps at CT colonography: prevalence and characteristics of the serrated polyp spectrum. Radiology. 2016;280(2):455–63.
Longacre TA, Cecilia MFP. Mixed hyperplastic adenomatous polyps/serrated adenomas A distinct form of colorectal neoplasia. Am J Surg Pathol. 1990;14(6):524–37.
Van der ML, Geoffrey H. Visualizing data using t-SNE. J Mach Learn Res. 2008;9(11).
Selvaraju RR, et al. Grad-cam: Visual explanations from deep networks via gradient-based localization. Proceedings of the IEEE international conference on computer vision. 2017.
Simonyan K, Andrew Z. Very deep convolutional networks for large-scale image recognition. 2014. https://arxiv.org/abs/1409.1556.
Howard AG, et al. Mobilenets: Efficient convolutional neural networks for mobile vision applications. 2017.https://arxiv.org/abs/1704.04861.
Chollet F. keras. 2015.
Abadi M, et al. Tensorflow: A system for large-scale machine learning. 12th USENIX symposium on operating systems design and implementation (OSDI 16). 2016.
Hiroyasu U, et al. Colorectal Polyp Classification Based On Latent Sharing Features Domain from Multiple Endoscopy Images. Procedia Computer Science. 2020;176:2507–14.
Robin Z, et al. Prediction of Polyp Pathology Using Convolutional Neural Networks Achieves Resect and Discard Thresholds. Am J Gastroenterol. 2020;115(1):138.
Jin EH, et al. Improved accuracy in optical diagnosis of colorectal polyps using convolutional neural networks with visual explanations. Gastroenterology. 2020;158(8):2169–79.
Pu LZCT, et al. Computer-aided diagnosis for characterization of colorectal lesions: comprehensive software that includes differentiation of serrated lesions. Gastrointest Endosc. 2020;92(4):891–9.
Eladio RD, et al. Real-time artificial intelligence-based histologic classification of colorectal polyps with augmented visualization. Gastrointest Endosc. 2021;93(3):662–70.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflicts of interest
The authors declare that they have no conflict of interest.
Additional information
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Sierra-Jerez, F., Martínez, F. A deep representation to fully characterize hyperplastic, adenoma, and serrated polyps on narrow band imaging sequences. Health Technol. 12, 401–413 (2022). https://doi.org/10.1007/s12553-021-00633-8
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s12553-021-00633-8