Effect of Geometry and Blood Flow in Cooling of Upper Leg

Abstract

Systemic hypothermia has been shown to reduce neurological damage in postcardiac arrest patients. Most doctors conductively cool patients with liquid-circulating devices or ice. However, there is a lack of knowledge about the thermal response of tissue to localized cooling. Current thermal models are designed for determining human comfort and have not been evaluated for the targeted low temperatures required for inducing hypothermia. Metabolic heat generation and tissue perfusion rate can significantly change under low temperature, which in turn affects the overall heat flux and cooling rates.This study evaluates the effects of geometry and blood flow in the upper leg by comparing numerical simulation results with previous experimental data.A geometrically accurate 3D model of the Visible Human Male upper leg was simulated using COMSOL Multiphysics. The segmented tissue included skin, fat, muscle, and bone. Large blood vessels, femoral vein and artery, great saphenous vein, and anterior femoral vein, are modeled, as they can be significant heat sources.Penne's bioheat equation was used to solve the temperature field. Variation of metabolic heat generation with temperature is modeled using the Q10 factor of 2, as widely used in literature [1]. Cutaneous vasoconstriction, tissue blood flow, and shivering are modeled using correlations provided by Fiala [1]. The numerical model results were compared with cold immersion studies [2–4].As seen in Fig. 1, the simulation results agree well with experimental data during whole-body cooling at 18 °C, but there is significant discrepancy with experimental data in localized cooling at 4 °C. The most likely reason for this discrepancy is hypothesized to be incorrect modeling of cutaneous blood flow at low temperatures. Data from immersion at 8 °C shows that low temperatures result is cold-induced vasodilation [4], which is not accounted for in the current model. When cutaneous blood flow is set constant at initial value, the error is significantly reduced. With vasodilation, local heat flux increases by 40% at 40 minutes, showing that proper modeling of vasoconstriction can significantly improve the model.Analysis of muscle temperature profile during cooling at 8 °C shows disagreement between experimental data and the current model. Fig. 2 shows the temperature profiles at 70 min of cooling. The most likely cause of this discrepancy was hypothesized to be inaccurate modeling of the muscle blood flow.Gergson et al. observed that femoral artery blood flow dropped to nearly half the initial value during immersion at 8 °C [4]. Combined with the fact that skin blood flow did not decrease (due to cold-induced vasodilation), they were able to conclude that the muscle blood flow must drop. In the current model, when the muscle perfusion is held constant at half the initial value, the results are significantly closer to experimental data.Results from the current simulations show that cutaneous and muscle perfusion both significantly influence the response of the model. Current perfusion models are not accurate at low temperature cooling.