Related Experiment Video
Updated: Apr 15, 2026

TBase - an Integrated Electronic Health Record and Research Database for Kidney Transplant Recipients
Published on: April 13, 2021
Clinical Artificial Intelligence Agents in Nephrology: From Prediction to Action Through Workflow-Native
Charat Thongprayoon1, Francesco Pesce2,3, Wisit Cheungpasitporn1
1Division of Nephrology and Hypertension, Department of Medicine, Mayo Clinic, Rochester, MN 55905, USA.
Artificial intelligence agents offer a new paradigm for kidney care, moving beyond prediction to real-time clinical action and coordinated care in nephrology. This approach enhances patient management for conditions like chronic kidney disease and acute kidney injury.
Area of Science:
- Nephrology
- Artificial Intelligence
- Health Informatics
Background:
- Current AI in nephrology primarily focuses on predictive models for outcomes like acute kidney injury (AKI) and chronic kidney disease (CKD) progression.
- Limited clinical impact of existing AI models stems from prediction alone not translating to coordinated clinical action.
- Clinical artificial intelligence agents are proposed as a new paradigm for seamless integration into nephrology workflows.
Purpose of the Study:
- To propose clinical artificial intelligence agents as a novel framework for integrating AI into nephrology workflows.
- To define the architecture, governance, and evaluation principles for these agents.
- To shift AI's role from isolated prediction to longitudinal clinical orchestration.
Main Methods:
- Conducted a narrative synthesis of literature on AI systems, agentic architectures, clinical decision support, and digital health in kidney care.
- Developed a conceptual framework for clinical artificial intelligence agents in nephrology.
- Drew from interdisciplinary sources in medicine, health informatics, and AI research.
Main Results:
- Clinical artificial intelligence agents are workflow-integrated systems capable of perception, reasoning, planning, and action to support coordinated clinical care.
- A layered architecture (perception, cognition, planning/control, action, learning) is described.
- Potential applications include CKD management, AKI monitoring, dialysis optimization, transplant care, and patient-facing systems.
Conclusions:
- Clinical artificial intelligence agents enable a shift from isolated prediction to longitudinal clinical orchestration in nephrology.
- Future evaluations should prioritize workflow integration, time-to-action, clinician oversight, safety, and patient outcomes.
- This framework supports the responsible development and integration of agentic AI in kidney care.
More Related Videos
Related Concept Videos
Non-equilibrium in the Cell
Chronic Kidney Disease III: Interprofessional Care
Acute Kidney Injury IV: Diagnostic Studies and Prevention
Acute Kidney Injury V: Interprofessional Care
Chronic Kidney Disease IV: Nursing Management
Chronic Kidney Disease I: Introduction

