A new study abstract from the University of Louisville, led by Adam E Gaweda (University of Louisville Health Sciences Center, Louisville, USA), was presented at the recent American Society of Nephrology (ASN) Kidney Week meeting (1–5 November 2023, Philadelphia, USA), which suggests that the use of artificial intelligence (AI) could help predict complications during dialysis treatments. This could mean that any intervention that might be needed to prevent venous stenosis can be done before it occurs, therefore removing any interruptions to dialysis treatment.
Gaweda et al evaluated an AI approach to predicting venous stenosis in different locations before they occurred, obtaining the data they used from the University of Louisville dialysis electronic medical record system and the interventional nephrology programme for the years 2018 to 2020. Information was also routinely obtained, and events were determined by direct examination within the interventional facility and segregated into inflow, outflow, central, and other stenosis. Resistance was defined as the ratio of pressure to blood flow rate.
With the use of the data collected, the AI method was able to predict events one week in advance. This shows that, even with no additional data collection, there is “sufficient information [that] exists during the dialysis procedure that can be leveraged using advanced mathematical techniques to adequately predict dialysis access complications,” as the authors determined. They concluded that “the methodology applied also allows for the explanation of the observations seen and can provide predictions through a mathematical model or an expert system in the form of a decision tree.”
Considering that, according to Gaweda et al, “the single most important factor determining quality of dialysis is adequate access,” being able to accurately predict issues that can occur regarding access could be both time and life-saving. AI-based approaches can be a helpful adjunct in the clinical care of patients and help with early detection of vascular access complications and associated morbidity and costs.