Boosted Capsule Network for Gastrointestinal Disease Recognition

  • Authors

    • Henrietta Adjei Pokuaa Department of Computer Science and Informatics, University of Energy and Natural Resources and Department of Computer Science, Sunyani Technical University
    • Adebayo Felix Adekoya Faculty of Computing Engineering and Mathematical Science, Catholic University of Ghana https://orcid.org/0000-0002-5029-2393
    • Benjamin Asubam Weyori Department of Computer and Electrical Sciences, University of Energy and Natural Resources https://orcid.org/0000-0001-5422-4251
    • Mighty Abra Ayidzoe Department of Computer Science and Informatics, University of Energy and Natural Resources https://orcid.org/0000-0001-5105-9625
    https://doi.org/10.14419/rcjf9t51

    Received date: January 1, 2026

    Accepted date: February 6, 2026

    Published date: February 23, 2026

  • Capsule Networks; Convolutional Neural Networks; Feature Boosting; Gastrointestinal Diseases; Class Capsule Boosting

  • Abstract

    Gastrointestinal diseases affect over 40% of the global population and rank among the leading causes of cancer-related deaths worldwide. ‎Wireless capsule endoscopy (WCE), though minimally invasive, generates 45,000-50,000 images per procedure, with expert endoscopists ‎missing 22-28% of pathological cases due to examination complexity and varied disease presentations. Deep learning approaches like Con-‎volutional Neural Networks require large annotated datasets, which are rarely available in medical imaging. Capsule Networks (CapsNets) ‎excel in limited-data scenarios through spatial-hierarchy inference but struggle with subtle features in complex medical images. We present a Boosted Capsule Network that incorporates dual enhancement mechanisms: modified feature boosting via intensity inversion to expose diverse low-level patterns, reducing the generalization gap by 45%, and class capsule amplification to sharpen decision boundaries and improve inter-class separation. Ablation studies on the Kvasir-V2 dataset show individual components achieve 89.3% (feature boosting) and 91.7% ‎‎(class boosting), while their synergistic combination reaches 97.90% accuracy, a 15.8% improvement over baseline CapsNet (82.10%) and ‎‎1.1% over previous state-of-the-art (96.80%). The lightweight architecture adds zero trainable parameters for class boosting and minimal ‎overhead for feature enhancement, enabling real-time clinical deployment. We experimentally deployed the model as a web-based prototype ‎with human-in-the-loop uncertainty handling for confidence scores below 0.5‎.

  • References

    1. Afriyie Y, Weyori BA & Opoku AA (2022), Gastrointestinal tract disease recognition based on denoising capsule network. Cogent Engineering 9(1), https://doi.org/10.1080/23311916.2022.2142072.
    2. Ahmed A (2022), Classification of Gastrointestinal Images Based on Transfer Learning and Denoising Convolutional Neural Networks. Lecture Notes in Networks and Systems 288, 631–639, https://doi.org/10.1007/978-981-16-5120-5_48.
    3. Akoto-Adjepong V, Appiah O, Mensah PK & Appiahene P (2024), TTDCapsNet: Tri Texton-Dense Capsule Network for complex and medical image recognition. Plos One 19(3), e0300133, https://doi.org/10.1371/journal.pone.0300133.
    4. Ayidzoe MA, Yu Y, Mensah PK, Cai J, Adu K & Yifan T (2021), Gabor Capsule Network with Preprocessing Blocks for the Recognition of Complex Images. Machine Vision and Applications 32, https://doi.org/10.1007/s00138-021-01221-6.
    5. Bajhaiya D & Unni SN (2024), Deep learning-enabled detection and localization of gastrointestinal diseases using wireless-capsule endoscopic im-ages. Biomedical Signal Processing and Control 93, 106125, https://doi.org/10.1016/j.bspc.2024.106125.
    6. Freedman D, Blau Y, Katzir L, Aides A, Shimshoni I, Veikherman D, Golany T, Gordon A, Corrado G & Matias Y (2020), Detecting deficient coverage in colonoscopies. IEEE Transactions on Medical Imaging 39(11), 3451–3462, https://doi.org/10.1109/TMI.2020.2994221.
    7. Hasan MM, Hossain MM, Mia S, Ahammad MS & Rahman MM (2022), A combined approach of non-subsampled contourlet transform and con-volutional neural network to detect gastrointestinal polyp. Multimedia Tools and Applications 81(7), 9949–9968, https://doi.org/10.1007/s11042-022-12250-2.
    8. Lonseko ZM, Adjei PE, Du W, Luo C, Hu D, Zhu L, Gan T & Rao N (2021), Gastrointestinal disease classification in endoscopic images using at-tention-guided convolutional neural networks. Applied Sciences (Switzerland) 11(23), https://doi.org/10.3390/app112311136.
    9. Mohapatra S, Nayak J, Mishra M, Pati GK, Naik B & Swarnkar T (2021), Wavelet Transform and Deep Convolutional Neural Network-Based Smart Healthcare System for Gastrointestinal Disease Detection. Interdisciplinary Sciences – Computational Life Sciences 13(2), 212–228, https://doi.org/10.1007/s12539-021-00417-8.
    10. Oh CK, Kim T, Cho YK, Cheung DY, Lee BI, Cho YS, Kim JI, Choi MG, Lee HH & Lee S (2021), Convolutional neural network-based object detection model to identify gastrointestinal stromal tumors in endoscopic ultrasound images. Journal of Gastroenterology and Hepatology (Austral-ia) 36(12), 3387–3394, https://doi.org/10.1111/jgh.15653.
    11. Pogorelov K, Randel KR, Griwodz C, Eskeland SL, de Lange T, Johansen D, Spampinato C, Dang-Nguyen DT, Lux M, Schmidt PT, Riegler M & Halvorsen P (2017), Kvasir: A Multi-Class Image Dataset for Computer Aided Gastrointestinal Disease Detection. Proceedings of the 8th ACM on Multimedia Systems Conference, 164–169, https://doi.org/10.1017/CBO9781107415324.004.
    12. Pokuaa HA, Adekoya AF, Weyori BA & Nyarko-Boateng O (2024), DPafy-GCaps: Denoising patch-and-amplify Gabor capsule network for the recognition of gastrointestinal diseases. Turkish Journal of Electrical Engineering and Computer Sciences 32(3), 452–464, https://doi.org/10.55730/1300-0632.4080.
    13. Pokuaa HA, Adekoya AF, Weyori BA & Nyarko-Boateng O (2024), Patch-and-amplify Capsule Network for the recognition of gastrointestinal diseases. Scientific African 25, e02277, https://doi.org/10.1016/j.sciaf.2024.e02277.
    14. Raut V, Gunjan R, Shete VV & Eknath UD (2023), Gastrointestinal tract disease segmentation and classification in wireless capsule endoscopy using intelligent deep learning model. Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization 11(3), 606–622, https://doi.org/10.1080/21681163.2022.2099298.
    15. Sabour S, Nicholas F & Hinton GE (2017), Dynamic Routing Between Capsules. Advances in Neural Information Processing Systems, 3856–3866.
    16. Sarsengeldin M, Imatayeva S, Abeuov N, Naukhanov M, Erdogan AS, Jha D & Bagci U (2023), Gastrointestinal Disease Diagnosis with Hybrid Model of Capsules and CNNs. IEEE International Conference on Electro Information Technology, 2023-May, 143–146, https://doi.org/10.1109/eIT57321.2023.10187250.
    17. Sengul SB & Ozkan IA (2024), HMedCaps: a new hybrid capsule network architecture for complex medical images. Neural Computing and Appli-cations 36(33), 20589–20606, https://doi.org/10.1007/s00521-024-10147-9.
    18. Wang W, Yang X, Li X & Tang J (2022), Convolutional-capsule network for gastrointestinal endoscopy image classification. International Journal of Intelligent Systems 37(9), 5796–5815, https://doi.org/10.1002/int.22815.
    19. Yadav S & Dhage S (2024), TE-CapsNet: time efficient capsule network for automatic disease classification from medical images. Multimedia Tools and Applications 83(16), 49389–49418, https://doi.org/10.1007/s11042-023-17458-4.
    20. Zhang K, Zuo W, Chen Y, Meng D & Zhang L (2017), Beyond a Gaussian denoiser: Residual learning of deep CNN for image denoising. IEEE Transactions on Image Processing 26(7), 3142–3155, https://doi.org/10.1109/TIP.2017.2662206.

  • Downloads

  • How to Cite

    Pokuaa, H. A., Adekoya, A. F., Weyori, B. A. ., & Ayidzoe, M. A. (2026). Boosted Capsule Network for Gastrointestinal Disease Recognition. International Journal of Basic and Applied Sciences, 15(2), 22-30. https://doi.org/10.14419/rcjf9t51