MODELING OF TEMPERATURE WITHIN THE NEONATAL HEAD SUBJECTED TO TRANSCRANIAL CONDUCTIVE COOLING

J W Hand1, G M J Van Leeuwen1, , A D Edwards2, D Azzopardi2

1 Radiological Sciences Unit, Department of Imaging, and 2 Department of Paediatrics and Neonatal Medicine, Imperial College School of Medicine (Hammersmith Campus), London

It has recently been established that allowing or promoting cooling of the brain for a finite period beginning after an hypoxic-ischaemic insult can reduce histological injury, and improve functional outcome. Data acquired using magnetic resonance imaging (MRI) have shown that it is damage to deep brain structures such as the basal ganglia that predict the most severe neurodevelopmental impairment after birth asphyxia, while cortical injury has a less grave prognosis. Currently there is no practical means of measuring deep brain temperature in patients accurately and continuously over periods of 10s of hours.

We have therefore investigated cooling within an anatomically realistic model of an infant's head. The conventional bioheat transfer equation with heat sink is used to simulate head cooling by a chilled water cap alone, and in combination with whole body cooling (simulated by assuming a reduced core temperature). The effect of the perfusion value on the resulting temperature distribution is investigated, as is the effect of the thermal resistance of the skin and skull. Finally, in a second head anatomy, predictions of the heat sink model and a thermal model that accounts for heat transfer by discrete vessels are compared.

The results suggest that there is little or no change in temperature in deep brain structures associated with transcranial contact cooling (due to water at 10 oC) in the absence of whole body cooling. Significant changes to the modelled brain perfusion or the thermal conductivity of the skull did not result in significant changes to the deep brain temperatures. Only in the case in which the infant's core temperature was lowered to 34 oC, to simulate whole body cooling, was a significant change in deep brain temperature predicted. The two models used - each differing in the way they accounted for the effects of perfusion - gave essentially the same result.