Endoscopic Deep Vision Intelligence: An EfficientNetB3-Based Diagnostic Framework for Multi-Class Gastrointestinal Structure and Lesion Recognition
Keywords:
Gastrointestinal endoscopy, Deep learning, EfficientNetB3, Multi-class classification, Endoscopic image analysis, Transfer learning, Computer-aided diagnosis, Medical image classification, Gastrointestinal lesion detection..Abstract
Accurate interpretation of Gastrointestinal (GI) endoscopic images is critical for early detection of diseases, along with making effective clinical decisions; however, manual diagnosis is time-consuming and subject to inter-observer variability. This work introduces Endoscopic Deep Vision Intelligence, a powerful deep learning-based diagnostic system based on the fine-tuned EfficientNetB3 deep neural network for multi-class gastrointestinal structure and lesion recognition. The proposed framework uses transfer learning and optimized feature scaling which maps out discriminative anatomical and pathological patterns from endoscopic images effectively. Experiments were performed on a publicly-available Kvasir endoscopy dataset of eight clinically significant classes. Comprehensive evaluation of using precision, recall, F1-score, confusion matrix, ROC--AUC analysis, and comparative model assessment proves the effectiveness of the proposed approach. The accurate fine-tuned EfficientNetB3 model is able to achieve an overall classification accuracy of 99% outperforming several existing state-of-the-art convolutional neural network architectures. The results verify great generalization, high discriminative power, and low misclassification among all classes. Owing to its high accuracy, efficiency, and reliability, the proposed framework shows great potential for its deployment into real world clinical decision support systems for automated gastrointestinal disease detection and anatomical structures recognition.
References
[1] Ahamed, M.F., Shafi, F.B., Nahiduzzaman, M., Ayari, M.A. and Khandakar, A., 2025. Interpretable deep learning architecture for gastrointestinal disease detection: A Tri-stage approach with PCA and XAI. Computers in Biology and Medicine, 185, p.109503.
[2] Waheed, Z., Gui, J., Amjad, K., Waheed, I. and Asif, S., 2025. An ensemble approach of deep CNN models with beta normalization aggregation for gastrointestinal disease detection. Biomedical Signal Processing and Control, 105, p.107567.
[3] Jagarajan, M. and Jayaraman, R., 2025. AI in gastrointestinal disease detection: overcoming segmentation challenges with Coati optimization strategy. Evolving Systems, 16(1), p.2.
[4] Jiang, Q., Yu, Y., Ren, Y., Li, S. and He, X., 2025. A review of deep learning methods for gastrointestinal diseases classification applied in computer-aided diagnosis system. Medical & Biological Engineering & Computing, 63(2), pp.293–320.
[5] Babu, P.L. and Jana, S., 2025. Gastrointestinal tract disease detection via deep learning based Duo-Feature Optimized Hexa-Classification model. Biomedical Signal Processing and Control, 100, p.106994.
[6] Fahad, M., Mobeen, N.E., Imran, A.S., Daudpota, S.M., Kastrati, Z., Cheikh, F.A. and Ullah, M., 2025. Deep insights into gastrointestinal health: A comprehensive analysis of GastroVision dataset using convolutional neural networks and explainable AI. Biomedical Signal Processing and Control, 102, p.107260.
[7] Park, J.Y., 2025. Image-enhanced endoscopy in upper gastrointestinal disease: focusing on texture and color enhancement imaging and red dichromatic imaging. Clinical Endoscopy, 58(2), pp.163–180.
[8] Khan, S.U.R., Asim, M.N., Vollmer, S. and Dengel, A., 2025. Robust & precise knowledge distillation-based novel context-aware predictor for disease detection in brain and gastrointestinal. arXiv preprint arXiv:2505.06381.
[9] Demirbaş, A.A., Üzen, H. and Fırat, H., 2024. Spatial-attention ConvMixer architecture for classification and detection of gastrointestinal diseases using the Kvasir dataset. Health Information Science and Systems, 12(1), p.32.
[10] Islam, M.S., Rony, M.A.T. and Sultan, T., 2024. GastroVRG: Enhancing early screening in gastrointestinal health via advanced transfer features. Intelligent Systems with Applications, 23, p.200399.
[11] Ahamed, M.F., Nahiduzzaman, M., Islam, M.R., Naznine, M., Ayari, M.A., Khandakar, A. and Haider, J., 2024. Detection of various gastrointestinal tract diseases through a deep learning method with ensemble ELM and explainable AI. Expert Systems with Applications, 256, p.124908.
[12] Asif, S., Zhao, M., Tang, F. and Zhu, Y., 2024. DCDS-Net: deep transfer network based on depth-wise separable convolution with residual connection for diagnosing gastrointestinal diseases. Biomedical Signal Processing and Control, 90, p.105866.
[13] Naz, J., Sharif, M., Yasmin, M., Raza, M. and Khan, M.A., 2021. Detection and classification of gastrointestinal diseases using machine learning. Current Medical Imaging Reviews, 17(4), pp.479–490.
[14] Hmoud Al-Adhaileh, M., Mohammed Senan, E., Alsaade, F.W., Aldhyani, T.H.H., Alsharif, N., Abdullah Alqarni, A., Uddin, M.I., Alzahrani, M.Y., Alzain, E.D. and Jadhav, M.E., 2021. Deep learning algorithms for detection and classification of gastrointestinal diseases. Complexity, 2021(1), p.6170416.
[15] Hussein, R.A., Al-Ouqaili, M.T. and Majeed, Y.H., 2021. Detection of Helicobacter Pylori infection by invasive and non-invasive techniques in patients with gastrointestinal diseases from Iraq: A validation study. PLOS ONE, 16(8), p.e0256393.
[16] Femi, P., Anestina, N., Anthony, O., Alade, A., Mustapha, A., Hamzah, F., Ebeye, C. and Obiageli, C., 2024. Advancements in Endoscopic Techniques for Early Detection and Minimally Invasive Treatment of Gastrointestinal Cancers: A Review of Diagnostic Accuracy, Clinical Outcomes, and Technological Innovations.