A Design Science Roadmap for Auditable Ocular Disease Classification: Evidence Mapping of AI Governance Gaps

Rizal Rachman; Eddy Soeryanto Soegoto; Irawan Afrianto; Irfan Dwiguna sumitra; Zainal Arifin Hasibuan

doi:10.66472/ismat.v1i2.442

Authors

Rizal Rachman Universitas Komputer Indonesia
Eddy Soeryanto Soegoto Universitas Komputer Indonesia
Irawan Afrianto Universitas Komputer Indonesia
Irfan Dwiguna sumitra Universitas Komputer Indonesia
Zainal Arifin Hasibuan Universitas Komputer Indonesia

DOI:

https://doi.org/10.66472/ismat.v1i2.442

Keywords:

Design Science Research, Evidence Mapping, Explainable AI, Model Auditability, Ocular Disease Classification

Abstract

The integration of Artificial Intelligence (AI) into ocular diagnostics has led to substantial improvements in predictive accuracy. However, a persistent gap remains between technical performance and clinical accountability. The present study addresses the "accuracy trap" and the lack of transparency in current deep learning models for ocular disease classification. The objective of the research is twofold: firstly, to identify methodological deficiencies in extant literature and, secondly, to propose a standardised evaluative framework to ensure model auditability. A systematic evidence mapping (SEM) approach, combined with design science research methodology (DSRM), was utilised to scrutinise 10 high-impact Scopus-indexed studies published between 2023 and 2026. The findings reveal a critical "predictive validity gap," where 80% of the evidence base relies on aggregate accuracy while 90% remains "black box" without functional Explainable AI (XAI) layers. The synthesis of these gaps resulted in the formulation of a conceptual roadmap that mandated multi-metric evaluation, incorporating Cohen's Kappa, and pathophysiological traceability. In conclusion, this research establishes that clinical deployment of AI must transition from model-centric success to a governance-oriented paradigm that prioritises decision utility and auditable audit trails. This roadmap provides a rigorous blueprint for the future implementation of transparent and accountable medical AI systems.

References

Adeleke, O., Olugbogi, J. A., & Abimbade, O. (2024). Transforming healthcare leadership decision-making through AI-driven predictive analytics: A new era of financial governance. https://doi.org/10.54660/IJMOR.2024.3.6.159-175

Albashayreh, A., Mosleh, H., & Yousef, M. (2025). Ocular disease recognition using deep learning models. In 2025 International Conference on New Trends in Computing Sciences (ICTCS) (pp. 276–283). IEEE. https://doi.org/10.1109/ICTCS65341.2025.10989463

Al-Fahdawi, S., et al. (2024). Fundus-DeepNet: Multi-label deep learning classification system for enhanced detection of multiple ocular diseases through data fusion of fundus images. Information Fusion, 102, Article 102059. https://doi.org/10.1016/j.inffus.2023.102059

Ali, G., Dastgir, A., Iqbal, M. W., Anwar, M., & Faheem, M. (2023). A hybrid convolutional neural network model for automatic diabetic retinopathy classification from fundus images. IEEE Journal of Translational Engineering in Health and Medicine, 11, 341–350. https://doi.org/10.1109/JTEHM.2023.3282104

Asia, A.-O., et al. (2022). Detection of diabetic retinopathy in retinal fundus images using CNN classification models. Electronics, 11(17), Article 2740. https://doi.org/10.3390/electronics11172740

Birkstedt, T., Minkkinen, M., Tandon, A., & Mäntymäki, A. (2023). AI governance: Themes, knowledge gaps and future agendas. Internet Research, 33(7), 133–167. https://doi.org/10.1108/INTR-01-2022-0042

Buettner, R. (2025). A novel transformer-CNN hybrid deep learning architecture for robust broad-coverage diagnosis of eye diseases on color fundus images.

de Figueiredo, L. A., Dias, J. V. P., Polati, M., Carricondo, P. C., & Debert, I. (2021). Strabismus and artificial intelligence app: Optimizing diagnostic and accuracy. Translational Vision Science & Technology, 10(7), Article 22. https://doi.org/10.1167/tvst.10.7.22

Duvvuri, K., Chethana, S., Charan, S. S., Srihitha, V., Ramesh, T. K., & Srikanth, K. S. (2022). Grad-CAM for visualizing diabetic retinopathy. In 2022 3rd International Conference for Emerging Technology (INCET) (pp. 1–4). IEEE. https://doi.org/10.1109/INCET54531.2022.9824598

El Arab, R. A., et al. (2025). Bridging the gap: From AI success in clinical trials to real-world healthcare implementation—A narrative review. Healthcare, 13(7), Article 701. https://doi.org/10.3390/healthcare13070701

Goktas, P., & Grzybowski, A. (2025). Shaping the future of healthcare: Ethical clinical challenges and pathways to trustworthy AI. Journal of Clinical Medicine, 14(5), Article 1605. https://doi.org/10.3390/jcm14051605

Hassan, M. ul, et al. (2024). A transfer learning enabled approach for ocular disease detection and classification. Health Information Science and Systems, 12(1), Article 36. https://doi.org/10.1007/s13755-024-00293-8

Herrera-Chavez, A. I., et al. (2024). Multi-label image classification for ocular disease diagnosis using K-fold cross-validation on the ODIR-5K dataset. In 2024 IEEE 33rd International Symposium on Industrial Electronics (ISIE) (pp. 1–6). IEEE. https://doi.org/10.1109/ISIE54533.2024.10595740

Jawad, M. A., Khursheed, F., Nawaz, S., & Mir, A. H. (2024). Towards improved fundus disease detection using Swin transformers. Multimedia Tools and Applications, 83(32), 78125–78159. https://doi.org/10.1007/s11042-024-18627-9

Khayyam, Y. T., Shah, S. M. A., Uddin, H. W., Farhan, M., & Mahmood, M. (n.d.). Multiclass eye disease recognition and classification using an ensemble deep transfer learning framework.

Kinger, S., & Kulkarni, V. (2024). A review of explainable AI in medical imaging: Implications and applications. International Journal of Computers and Applications, 46(11), 983–997. https://doi.org/10.1080/1206212X.2024.2404082

Kothinti, R. R. (2024). Artificial intelligence in healthcare: Revolutionizing precision medicine, predictive analytics, and ethical considerations in autonomous diagnostics. World Journal of Advanced Research and Reviews, 19(3), 3395–3406.

Li, D. (2025). AI-driven healthcare analytics and predictive medicine: Advanced machine learning frameworks for clinical decision support and patient outcome optimization. Spectrum of Research, 5(1).

Mayya, V., S. K., Kulkarni, U., Surya, D. K., & Acharya, U. R. (2023). An empirical study of preprocessing techniques with convolutional neural networks for accurate detection of chronic ocular diseases using fundus images. Applied Intelligence, 53(2), 1548–1566. https://doi.org/10.1007/s10489-022-03490-8

Mewada, S., et al. (2022). Smart diagnostic expert system for defect in forging process by using machine learning process. Journal of Nanomaterials, 2022(1), Article 2567194. https://doi.org/10.1155/2022/2567194

Muhamediyeva, D. T., Safarova, L. U., & Ruzikulov, R. F. (2023). Expert systems for diagnostics of infectious diseases in cattle. In BIO Web of Conferences (Vol. 71, Article 01073). EDP Sciences. https://doi.org/10.1051/bioconf/20237101073

Oguz, C., Aydin, T., & Yaganoglu, M. (2024). A CNN-based hybrid model to detect glaucoma disease. Multimedia Tools and Applications, 83(6), 17921–17939. https://doi.org/10.1007/s11042-023-16129-8

Ouda, O., AbdelMaksoud, E., Abd El-Aziz, A. A., & Elmogy, M. (2022). Multiple ocular disease diagnosis using fundus images based on multi-label deep learning classification. Electronics, 11(13), Article 1966. https://doi.org/10.3390/electronics11131966

Poly, T. N., Islam, M. M., Walther, B. A., Lin, M. C., & Li, Y.-C. J. (2023). Artificial intelligence in diabetic retinopathy: Bibliometric analysis. Computer Methods and Programs in Biomedicine, 231, Article 107358. https://doi.org/10.1016/j.cmpb.2023.107358

Sah, U. K., Chatterjee, J. M., & Sujatha, R. (2026). Multi-class eye disease classification using deep learning EfficientNetB0 fusion techniques. Scientific Reports, 16(1), Article 6368. https://doi.org/10.1038/s41598-026-35357-0

Santone, A., Cesarelli, M., Colasuonno, E., Bevilacqua, V., & Mercaldo, F. (2024). A method for ocular disease diagnosis through visual prediction explainability. Electronics, 13(14), Article 2706. https://doi.org/10.3390/electronics13142706

Tan, T.-E., & Wong, T. Y. (2023). Diabetic retinopathy: Looking forward to 2030. Frontiers in Endocrinology, 13, Article 1077669. https://doi.org/10.3389/fendo.2022.1077669

Thulasi Vel, D. V., & Durgaraju, S. (2025). Deep multi-label classification for automated ocular disease detection using enhanced CNN architectures. In International Conference on Computational Intelligence and Network Security (pp. 115–127). Springer. https://doi.org/10.1007/978-3-032-09572-5_10

Vach, W., & Gerke, O. (2023). Gwet’s AC1 is not a substitute for Cohen’s kappa: A comparison of basic properties. MethodsX, 10, Article 102212. https://doi.org/10.1016/j.mex.2023.102212

Wahab Sait, A. R. (2023). Artificial intelligence-driven eye disease classification model. Applied Sciences, 13(20), Article 11437. https://doi.org/10.3390/app132011437

Więckowska, B., Kubiak, K. B., Jóźwiak, P., Moryson, W., & Stawińska-Witoszyńska, B. (2022). Cohen’s kappa coefficient as a measure to assess classification improvement following the addition of a new marker to a regression model. International Journal of Environmental Research and Public Health, 19(16), Article 10213. https://doi.org/10.3390/ijerph191610213

Yuan, H., Kang, L., & Li, Y. (2025). Opening the black box of deep learning: Validating the statistical association between explainable artificial intelligence (XAI) and clinical domain knowledge in fundus image-based glaucoma diagnosis (arXiv Preprint No. arXiv:2504.04549). arXiv.