ABSTRACT
Different urinary tract conditions often exhibit overlapping symptoms such as a strong urge to urinate, burning sensations, abnormal urine output and fever, making diagnosis and treatment challenging and time-consuming. Traditional diagnostic methods are inefficient, time consuming, expensive to practice and involve the use of a domain expert and finally increase the rate of immortality for those affected. Additionally, a significant issue with AI models is their black-box nature making it difficult for a reliable and trustworthy AI system. Therefore, this study proposed a StackSHAP dependable AI machine learning models to predict two urinary system diseases: acute nephritis of the renal pelvis (ANRP) and inflammation of the urinary bladder (IUB). StackSHAP values are used to address the black-box nature for interpretability. StackSHAP tools are employed to demystify the model, providing transparency and identifying key features influencing predictions. Additionally, this paper handled the issue of the data imbalance by employing support vector machine synthetic minority over-sampling technique (SVMSMOTE) is employed. Additionally, our work is investigated with twelve different machine learning techniques and checkmated with various evaluation metrics, thereby achieving an excellent prediction score across the evaluation metrics. This proposed model ensures the ethical adoption in healthcare by addressing transparency, casuality and interpretability in urinary disease diagnosis, thus providing valuable insights and making our study clinically relevant.
References
- [1] Aleta, A., Džuho, A., & Hrvat, F. (2021). Detection of acute inflammation of urinary bladder and acute nephritis of renal pelvis origin using artificial neural network. In 8th European Medical and Biological Engineering Conference: Proceedings of the EMBEC 2020, November 29–December 3, 2020 Portorož, Slovenia(pp. 363-371). Springer International Publishing.
- [2] Tong, F., Shahid, M., Jin, P., Jung, S., Kim, W. H., & Kim, J. (2020). Classification of the urinary metabolome using machine learning and potential applications to diagnosing interstitial cystitis. Bladder, 7(2).
- [3] Jhang, J. F., Yu, W. R., Huang, W. T., & Kuo, H. C. (2024). Combination of urinary biomarkers and machine-learning models provided a higher predictive accuracy to predict long-term treatment outcomes of patients with interstitial cystitis/bladder pain syndrome. World Journal of Urology, 42(1), 173.
- [4] Kim, J., De Hoedt, A., Wiggins, E., Haywood, K., Jin, P., Greenwood, B., … & Anger, J. T. (2021). Diagnostic utility of serum and urinary metabolite analysis in patients with interstitial cystitis/painful bladder syndrome. Urology, 157, 85-92.
- [5] Xu, H., Tamrat, N. E., Gao, J., Xu, J., Zhou, Y., Zhang, S., … & Wei, Z. (2021). Combined signature of the urinary microbiome and metabolome in patients with interstitial cystitis. Frontiers in Cellular and Infection Microbiology, 11, 711746.
- [6] Kahramanlı, H. “Determining the Acute Inflammations using Back Propagation Algorithm with Adaptive Learning Coefficients.” RTET, ICABES, Pattaya, 2016.
- [7] Rosly, R., Makhtar, M., Awang, M., & Rahman, M. “Analyzing performance of classifiers for medical datasets.” International Journal of Engineering & Technology, 2018.
- [8] Lango, M., & Stefanowski, J. “Multi-class and feature selection extensions of Roughly Balanced Bagging for imbalanced data.” Journal of Intelligent Information Systems, 2017.
- [9] Linusson, H., Johansson, U., Boström, H., & Löfström, T. “Reliable Confidence Predictions Using Conformal Prediction.” PAKDD, 2016.
- Guo, H., Zhi, W., Liu, H., & Xu, M. “Imbalanced Learning Based on Logistic Discrimination.” Computational Intelligence and Neuroscience, 2016.
- Lochotinunt, S. Pechprasarn and T. Treebupachatsakul, “Classification model for predicting inflammation of the urinary bladder and acute nephritis of the renal pelvis,” 2022 14th Biomedical Engineering International Conference (BMEiCON), Songkhla, Thailand, 2022, pp. 1-5, doi: 10.1109/BMEiCON56653.2022.10012109.
- Ros, M. “Comparative study on the machine learning approaches for the prognosis of acute inflammations.” AIP Conference Proceedings, 2023.
- [13] Czerniak, Jacek. (2009). Acute Inflammations. UCI ML Repository.https://doi.org/10.24432/C5V59S.
Download all article in PDF
