International Journal of Image, Graphics and Signal Processing (IJIGSP)
IJIGSP Vol. 18, No. 4, 8 Aug. 2026
Cover page and Table of Contents: PDF (size: 1492KB)
Federated Transformer-Enhanced MobileNet V2 for Decentralized Polyp Detection in Colonoscopy Images
PDF (1492KB), PP.121-138
Views: 0 Downloads: 0
Author(s)
Kumar Janardan Patra
1
Jibitesh Mishra
1
Soumya Ranjan Nayak
2,*
Index Terms
Polyp Detection Federated Learning, MobileNetV2, Transformer, Privacy-Preserving AI, Medical Image Analysis, Deep Learning, Colonoscopy, Lightweight CNN, Decentralized Training
Abstract
Early polyp detection is vital in avoiding colorectal cancer, a top contributor to cancer-related deaths globally. Automated polyp detection has been greatly improved by deep learning, but extensive deployment tends to be hampered by patient data privacy fears. For this purpose, we suggest a federated learning (FL) paradigm that enables decentralized model training without sharing raw patient data while having high diagnostic accuracy. We first performed an extensive evaluation using DL models on a polyp dataset that we gathered. MobileNetV2 was the best performing model as per important metrics like accuracy, precision, recall, and F1 score. In order to further expand its representational power, we incorporated a transformer module into MobileNetV2 so that the model can better capture long-range dependencies and context information. Our new Transformer-Enhanced MobileNetV2 model was then implemented on several simulated clients in a federated learning scenario. This configuration enabled training over decentralized clinical data without violating patient privacy. We utilized standard FL algorithms for model averaging and evaluated the system in accuracy, precision, recall, F1 score, and convergence time. The devised approach performed extremely well with an accuracy of 98.37% and an F1 score of 0.971 while converging effectively in 27-33 rounds. These findings imply that a promising path for safe and effective medical image analysis is to combine transformer designs with lightweight models in a federated condition.
Cite This Paper
Kumar Janardan Patra, Jibitesh Mishra, Soumya Ranjan Nayak, "Federated Transformer-Enhanced MobileNet V2 for Decentralized Polyp Detection in Colonoscopy Images", International Journal of Image, Graphics and Signal Processing(IJIGSP), Vol.18, No.4, pp. 121-138, 2026. DOI:10.5815/ijigsp.2026.04.07
Reference
[1]Patra, K. J., Mishra, J., Dash, S. K., & Jayasingh, S. K. (2024, November). Enhancing Diagnostic Precision for Early Detection of Gastric Polyps with U-Net3+ Segmentation. In 2024 3rd Odisha International Conference on Electrical Power Engineering, Communication and Computing Technology (ODICON) (pp. 1-5). IEEE.
[2]Ribeiro, J., Nóbrega, S., & Cunha, A. (2022). Polyps detection in colonoscopies. Procedia Computer Science, 196, 477-484.
[3]Doo, F. X. (2025). Artificial Intelligence (AI) in Abdominal Imaging. In What Radiology Residents Need to Know: Abdominal Radiology (pp. 347-359). Cham: Springer Nature Switzerland.
[4]Hasan, M. M., & Rahman, M. M. (2025). Privacy-preserving polyp segmentation using federated learning with differential privacy. Smart Health, 36, 100551.
[5]Rauniyar, A., Hagos, D. H., Jha, D., Håkegård, J. E., Bagci, U., Rawat, D. B., & Vlassov, V. (2023). Federated learning for medical applications: A taxonomy, current trends, challenges, and future research directions. IEEE Internet of Things Journal, 11(5), 7374-7398.
[6]Devkota, A., Amireskandari, A., Palko, J., Thakkar, S., Adjeroh, D., Jiang, X., ... & Gyawali, P. K. (2025). Federated Foundation Model for GI Endoscopy Images. arXiv preprint arXiv:2505.24108.
[7]Yasmin, F., Hassan, M. M., Hasan, M., Zaman, S., Bairagi, A. K., El-Shafai, W., ... & Chun, Y. C. (2023). GastroNet: Gastrointestinal polyp and abnormal feature detection and classification with deep learning approach. IEEE Access, 11, 97605-97624.
[8]McMahan, B., Moore, E., Ramage, D., Hampson, S., & y Arcas, B. A. (2017, April). Communication-efficient learning of deep networks from decentralized data. In Artificial intelligence and statistics (pp. 1273-1282). PMLR.
[9]Ejiga Peter, O. O., Adeniran, O. T., John-Otumu, A. M., Khalifa, F., & Rahman, M. M. (2025). Text-Guided Synthesis in Medical Multimedia Retrieval: A Framework for Enhanced Colonoscopy Image Classification and Segmentation. Algorithms, 18(3), 155.
[10]de Moura Lima, A. C., de Paiva, L. F., Bráz, G., de Almeida, J. D. S., Silva, A. C., Coimbra, M. T., & de Paiva, A. C. (2023). A two-stage method for polyp detection in colonoscopy images based on saliency object extraction and transformers. IEEE Access, 11, 76108-76119.
[11]Ribeiro, E., Uhl, A., Wimmer, G., & Häfner, M. (2016). Exploring deep learning and transfer learning for colonic polyp classification. Computational and mathematical methods in medicine, 2016(1), 6584725.
[12]Klare, P., Sander, C., Prinzen, M., Haller, B., Nowack, S., Abdelhafez, M., ... & Wittenberg, T. (2019). Automated polyp detection in the colorectum: a prospective study (with videos). Gastrointestinal endoscopy, 89(3), 576-582.
[13]Gross, S., Trautwein, C., Behrens, A., Winograd, R., Palm, S., Lutz, H. H., ... & Tischendorf, J. J. (2011). Computer-based classification of small colorectal polyps by using narrow-band imaging with optical magnification. Gastrointestinal endoscopy, 74(6), 1354-1359.
[14]Li, H., Zhen, G., Chu, C., Ma, Y., & Zhao, Y. (2024). Noise-Perception Multi-Frame Collaborative Network for Enhanced Polyp Detection in Endoscopic Videos. Electronics, 14(1), 62.
[15]Ali, S., Zhou, F., Braden, B., Bailey, A., Yang, S., Cheng, G., ... & Rittscher, J. (2020). An objective comparison of detection and segmentation algorithms for artefacts in clinical endoscopy. Scientific reports, 10(1), 2748.
[16]Krenzer, A., Heil, S., Fitting, D., Matti, S., Zoller, W. G., Hann, A., & Puppe, F. (2023). Automated classification of polyps using deep learning architectures and few-shot learning. BMC Medical Imaging, 23(1), 59.
[17]Younas, F., Usman, M., & Yan, W. Q. (2023). A deep ensemble learning method for colorectal polyp classification with optimized network parameters. Applied Intelligence, 53(2), 2410-2433.
[18]Hsu, C.-M., Hsu, C.-C., Hsu, Z.-M., Shih, F.-Y., Chang, M.-L., & Chen, T.-H. (2021). Colorectal Polyp Image Detection and Classification through Grayscale Images and Deep Learning. Sensors, 21(18), 5995. https://doi.org/10.3390/s21185995
[19]Korbar, B., Olofson, A. M., Miraflor, A. P., Nicka, C. M., Suriawinata, M. A., Torresani, L., ... & Hassanpour, S. (2017). Deep learning for classification of colorectal polyps on whole-slide images. Journal of pathology informatics, 8(1), 30.
[20]Eriyanti, N. A., Sigit, R., & Harsono, T. (2021, September). Classification of colon polyp on endoscopic image using support vector machine. In 2021 International Electronics Symposium (IES) (pp. 244-250). IEEE.
[21]Lo, C.-M., Yeh, Y.-H., Tang, J.-H., Chang, C.-C., & Yeh, H.-J. (2022). Rapid Polyp Classification in Colonoscopy Using Textural and Convolutional Features. Healthcare, 10(8), 1494. https://doi.org/10.3390/healthcare10081494
[22]Gao, H., Wu, Q., Zhao, X., Zhu, J., & Zhang, M. (2023). FedADT: An Adaptive Method Based on Derivative Term for Federated Learning. Sensors, 23(13), 6034. https://doi.org/10.3390/s23136034
[23]Jha, D., Smedsrud, P. H., Riegler, M. A., Halvorsen, P., De Lange, T., Johansen, D., & Johansen, H. D. (2019, December). Kvasir-seg: A segmented polyp dataset. In International conference on multimedia modeling (pp. 451-462). Cham: Springer International Publishing.
[24]Jha, D., Smedsrud, P. H., Riegler, M. A., Halvorsen, P., De Lange, T., Johansen, D., & Johansen, H. D. (2019, December). Kvasir-seg: A segmented polyp dataset. In International conference on multimedia modeling (pp. 451-462). Cham: Springer International Publishing.
[25]Khan, M. A., Sarfraz, M. S., Alhaisoni, M., Albesher, A. A., Wang, S., & Ashraf, I. (2020). StomachNet: Optimal deep learning features fusion for stomach abnormalities classification. IEEE Access, 8, 197969-197981.
[26]Mohammad, F., & Al-Razgan, M. (2022). Deep feature fusion and optimization-based approach for stomach disease classification. Sensors, 22(7), 2801.
[27]Shorten, C., & Khoshgoftaar, T. M. (2019). A survey on image data augmentation for deep learning. Journal of big data, 6(1), 1-48.
[28]Mikołajczyk, A., & Grochowski, M. (2018, May). Data augmentation for improving deep learning in image classification problem. In 2018 international interdisciplinary PhD workshop (IIPhDW) (pp. 117-122). IEEE.
[29]Viazovetskyi, Y., Ivashkin, V., & Kashin, E. (2020). Stylegan2 distillation for feed-forward image manipulation. In Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XXII 16 (pp. 170-186). Springer International Publishing.
[30]Zeng, P., Song, R., Chen, S., Li, X., Li, H., Chen, Y., ... & Chen, Z. (2025). Expert-guided StyleGAN2 image generation elevates AI diagnostic accuracy for maxillary sinus lesions. Communications Medicine, 5(1), 1-10.
[31]Tajbakhsh, N., Gurudu, S. R., & Liang, J. (2015). Automated polyp detection in colonoscopy videos using shape and context information. IEEE transactions on medical imaging, 35(2), 630-644.
[32]Elkarazle, K., Raman, V., Then, P., & Chua, C. (2023). Improved colorectal polyp segmentation using enhanced MA-NET and modified Mix-ViT transformer. IEEE Access, 11, 69295-69309.
[33]Fu, B., Zhang, M., He, J., Cao, Y., Guo, Y., & Wang, R. (2022). StoHisNet: A hybrid multi-classification model with CNN and Transformer for gastric pathology images. Computer Methods and Programs in Biomedicine, 221, 106924.
[34]Bae, S. H., & Yoon, K. J. (2015). Polyp detection via imbalanced learning and discriminative feature learning. IEEE transactions on medical imaging, 34(11), 2379-2393.
[35]Cao, X., Fan, K., & Ma, H. (2024). Federal learning-based a dual-branch deep learning model for colon polyp segmentation. Multimedia Tools and Applications, 1-22.
[36]Li, X., Huang, K., Yang, W., Wang, S., & Zhang, Z. (2019). On the convergence of fedavg on non-iid data. arXiv preprint arXiv:1907.02189.