An Artificial Intelligence-Based Decision Support System for Personalized Immunosuppressive Treatment in Kidney Transplant Recipients

Abstract

Kidney transplant is a life-saving procedure in the management of end-stage renal diseases. However, graft long-term survival entirely relies on the effective management of immunosuppressive medications. Inappropriate dosing of immunosuppressive drugs may result in either severe complications of graft rejection or drug-induced toxicity. This paper proposes an AI-driven framework to analyze immunosuppressive medication patterns in kidney transplant patients using their longitudinal clinical data. For patient risk level assessment, the proposed system integrates data preprocessing, temporal feature modeling, and deep learning-based classification to support clinical decision-making in adjusting immunosuppressive medications. Clinical parameters such as laboratory test results, medication dosage, and patient demographics are leveraged to represent the dynamic response of patients over time. Experimental results demonstrate that the proposed model yields a classification accuracy of 96.4%, with significant reduction in misclassifying high-risk cases. Results showed that the proposed approach is effective, reliable, and suitable for supporting immunosuppressive therapy optimization in post-kidney transplant care.

Ähnliche Arbeiten