Publication:

Combined thermal imaging and colour duplex ultrasound assessments of renal fistula patients. (2004)

Author(s): Murray A; Chishti A; Allen J; Talbot D; Oates CP; Ahmed I

    Abstract: Vascular and clinical assessments of fistula function are important in patients undergoing or preparing to undergo renal dialysis. Objective assessments at Freeman Hospital now include combined colour duplex ultrasound and medicalthermography. For example, these modalities can help study problems relating tohigh fistula flow and vascular steal, where digital blood flow (and hence skintemperature) can be impaired. The aims of this pilot study were a) to determine iffistula region skin temperature was related to fistula region blood flow and b) tocompare simple differences in mean hand temperature against the clinical stealgrading.Renal patients were clinically assessed for vascular steal by the transplantsurgeon (either steal or no steal). Patients also underwent objective vascularmeasurements which comprised thermal imaging of the hands and fistula regionfollowed by fistula blood flow estimation using colour duplex ultrasound at thebrachial artery. Differences in hand temperature and mean fistula regiontemperature were determined using dedicated image processing software (FLIRSC300 thermal imaging system with ThermaCam Researcher image processingsoftware, skin emissivity 0.97). These temperatures were then compared withfistula flow and steal grading.Twelve patients were studied (mean age 59 years), with five classed as havingsome degree of steal. Ultrasound measurements also identified the presence ofstenosis in three patients. Estimated fistula flows ranged from 30 - 1950 ml min-1(mean [standard deviation] of 1100 [640] ml min-1) and were correlated withmean fistula region skin temperature (R = +0.6, p<0.05). Thermography usuallyclearly highlighted the warmer superficial blood vessels in the region of the fistula(33.7 [1]oC). Hand (non-fistula – fistula side) temperature differences with a cut-off of +1oC were found to separate steal from non-steal patients with anaccuracy of 92% (specificity 100%, sensitivity 80%). In this study the maximumdifference between mean hand temperatures for a patient with steal was close to5oC.We have demonstrated an association between fistula region skin temperatureand estimated fistula blood flow. We have also shown that a bilateral handtemperature difference cut-off of +1oC separates steal from non-steal patientswith an accuracy of greater than 90%. Further work is now needed to explore theclinical utility of these findings, to identify which patients subsequently neededsurgery, and also to examine the detailed characteristics of the fistula thermal profiles.

      • Date: 3rd November 2004
      • Conference Name: 2004 UK Symposium on Medical Infrared Thermography.
      • Volume: 14(4)
      • Pages: 144
      • Publication type: Conference Proceedings (inc. abstract)
      • Bibliographic status: Published
        Staff

        Professor Alan Murray
        Professor of Cardiovascular Physics