Abstract
Since medical equipment has limited resources and little computational capacity, large segmentation models cannot be installed on them. Colonoscopy equipment, which has little computer power for deep learning models, is one such example where large segmentation models cannot be embedded. The solution to this problem is to perform model compression on cutting-edge models that have demonstrated outstanding diagnostic and prediction capabilities while performing segmentation. Knowledge distillation of large medical image segmentation models is the main focus of our study. However, unlike other knowledge distillation papers, we find the best student and teacher pair with varied network architecture. Even though the network architectures of our student and teacher pairs are different, the pairs still achieve better results in terms of segmentation score, pixel accuracy intersection-over-union (IoU, Jaccard Index), and dice coefficient (F1 Score). Overfitting is one of the main issues with this methodology, which we were able to minimize through depth-wise model pruning and hyperparameter tuning. Finally, we achieved the Xception-Efficientb0 pair as knowledge distillation which can outperform state-of-the-art models’ values of performance metrics on Kvasir-SEG and CVC-ClinicDB datasets.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Wang, R., et al.: Medical image segmentation using deep learning: a survey. IET Image Process. 16(5), 1243–1267 (2022)
Roth, H.R., Oda, H., Zhou, X., Shimizu, N., Yang, Y., Hayashi, Y., et al.: An application of cascaded 3D fully convolutional networks for medical image segmentation. Comput. Med. Imag. Graph. 66, 90–99 (2018)
Masud, M., et al.: A machine learning approach to diagnosing lung and colon cancer using a deep learning-based classification framework. Sensors 21(3), 748 (2021)
Brenner, H., Stock, C., Hoffmeister, M.: Effect of screening sigmoidoscopy and screening colonoscopy on colorectal cancer incidence and mortality: systematic review and meta-analysis of randomised controlled trials and observational studies. BMJ 128, 348 (2014)
Asano, T.K., McLeod, R.S.: Dietary fibre for the prevention of colorectal adenomas and carcinomas. Cochrane Database Syst. Rev. 1, CD003430 (2002)
Sivananthan, A., Glover, B., Ayaru, L., Patel, K., Darzi, A., Patel, N.: The evolution of lower gastrointestinal endoscopy: where are we now? Therap. Adv. Gastrointest. Endosc. 13, 2631774520979591 (2020)
Eu, C.Y., Tang, T.B., Lin, C.H., Lee, L.H., Lu, C.K.: Automatic polyp segmentation in colonoscopy images using a modified deep convolutional encoder-decoder architecture. Sensors 21(16), 5630 (2021)
Brandao, P.: Enhancing endoscopic navigation and polyp detection using artificial intelligence (Doctoral dissertation, UCL (University College London)) (2021)
Zhang, J., Tao, D.: Empowering things with intelligence: a survey of the progress, challenges, and opportunities in artificial intelligence of things. IEEE Internet Things J. 8(10), 7789–7817 (2020)
Feng, J., Li, S., Li, X., Wu, F., Tian, Q., Yang, M.H., Ling, H.: Taplab: a fast framework for semantic video segmentation tapping into compressed-domain knowledge. IEEE Trans. Pattern Anal. Mach. Intell. 44(3), 1591–1603 (2020)
Wang, W., Zhou, T., Porikli, F., Crandall, D., Van Gool, L.: A survey on deep learning techniques for video segmentation. arXiv preprint arXiv:2107.01153 (2021)
Xie, J., Shuai, B., Hu, J.F., Lin, J., Zheng, W.S.: Improving fast segmentation with teacher-student learning. arXiv preprint arXiv:1810.08476 (2018)
Li, G., Yun, I., Kim, J., Kim, J.: Dabnet: depth-wise asymmetric bottleneck for real-time semantic segmentation. arXiv preprint arXiv:1907.11357 (2019)
Sanchez-Peralta, L.F., Bote-Curiel, L., Picon, A., Sanchez-Margallo, F.M., Pagador, J.B.: Deep learning to find colorectal polyps in colonoscopy: a systematic literature review. Artif. Intell. Med. 108, 101923 (2020)
Kayes, M.I.: A lightweight and robust convolutional neural network for carcinogenic polyp identification (Doctoral dissertation, University of Science and Technology) (2021)
Chavarrias-Solano, P.E., Teevno, M.A., Ochoa-Ruiz, G., Ali, S.: Knowledge distillation with a class-aware loss for endoscopic disease detection. In: MICCAI Workshop on Cancer Prevention Through Early Detection, pp. 67–76. Springer, Cham (2022)
Sivaprakasam, M.: XP-Net: An Attention Segmentation Network by Dual Teacher Hierarchical Knowledge Distillation for Polyp Generalization (2022)
Kang, J., Gwak, J.: KD-ResUNet++: automatic polyp segmentation via self-knowledge distillation. In: MediaEval (2020)
Jha, D., Smedsrud, P.H., Riegler, M.A., Halvorsen, P., Lange, T.D., Johansen, D., Johansen, H.D.: Kvasir-seg: a segmented polyp dataset. In: International Conference on Multimedia Modeling, pp. 451–462. Springer, Cham (2020)
Chlap, P., et al.: A review of medical image data augmentation techniques for deep learning applications. J. Med. Imag. Radiat. Oncol. 65(5), 545–563 (2021)
Shorten, C., Khoshgoftaar, T.M.: A survey on image data augmentation for deep learning. J. Big Data 6(1), 1–48 (2019)
Patel, K., Bur, A.M., Wang, G.: Enhanced U-Net: a feature enhancement network for polyp segmentation. In: Proceedings of the 2021 18th Conference on Robots and Vision (CRV), pp. 181–188. IEEE (2021)
Zhou, Z., Rahman Siddiquee, M.M., Tajbakhsh, N., Liang, J.: Unet++: a nested U-Net architecture for medical image segmentation. In: Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support, pp. 3–11. Springer, Cham (2018)
Cahall, D.E., Rasool, G., Bouaynaya, N.C., Fathallah-Shaykh, H.M.: Inception modules enhance brain tumor segmentation. Front. Comput. Neurosci. 13, 44 (2019)
Chahal, E.S., Patel, A., Gupta, A., Purwar, A.: Unet based exception model for prostate cancer segmentation from MRI images. Multimedia Tools Appl. 81(26), 37333–37349 (2022)
Tan, M., Le, Q.: Efficient net: rethinking model scaling for convolutional neural networks. In: International Conference on Machine Learning, pp. 6105–6114. PMLR (2019)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
About this paper
Cite this paper
Kar, I., Mukhopadhyay, S., Balaiwar, R., Khule, T. (2023). A Novel Knowledge Distillation Technique for Colonoscopy and Medical Image Segmentation. In: Bhateja, V., Yang, XS., Ferreira, M.C., Sengar, S.S., Travieso-Gonzalez, C.M. (eds) Evolution in Computational Intelligence. FICTA 2023. Smart Innovation, Systems and Technologies, vol 370. Springer, Singapore. https://doi.org/10.1007/978-981-99-6702-5_7
Download citation
DOI: https://doi.org/10.1007/978-981-99-6702-5_7
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-99-6701-8
Online ISBN: 978-981-99-6702-5
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)