Fig 1.
Flow chart for the Delphi model for consensus generation.
Fig 2.
Geographic Distribution of Expert Participation in the Delphi Study.
This map illustrates the global representation of experts who contributed to the Delphi consensus on Acute Kidney Injury (AKI).
Table 1.
Summary of expert consensus statements on acute kidney injury in the ICU based on a modified Delphi process.
Fig 3.
Acute Kidney Injury (AKI) Delphi Model: Expert Consensus Framework.
This figure summarizes the major domains identified through the Delphi process, including etiologies, biomarker prioritization, biopsy considerations, and ICU management strategies. The biomarkers shown represent those most frequently ranked by experts for clinical decision-making, although not all reached the predefined consensus threshold.
Fig 4.
AKI Causal Pathways: A Simplified Directed Acyclic Graph (DAG) Model.
This diagram illustrates the key mechanisms of AKI, highlighting the roles of hemodynamics, nephrotoxicity, cardiac dysfunction, and urinary obstruction in the progression of kidney injury. It serves as a structured framework for AI-driven modeling and clinical decision-making. The edges (or arcs) represent relationships or transitions that connect the nodes (or vertices representing entities or objects within the graph with a specific direction, i.e., they point from one node to another, implying a flow from the starting node to the target node. There are no cycles in the graph, indicating that there is no closed loop within it, which allows them to represent systems where one element flows into another in a clear, non-recursive way. DAGs in healthcare are frequently used for workflow management, data flow analysis, and Bayesian networks, in which nodes represent random variables and directed edges represent conditional dependencies between them. They are also used to assess a process’s shortest path or detect cycles in the algorithms to model processes, dependencies, and data flows.