Non-invasive measurement of brain temperature using radiometric thermometry: experimental validation and clinical observations in asphyxiated newborns.

Background: Therapeutic hypothermia (HT) has been shown to decrease death and severe disability in infants with hypoxic-ischemic encephalopathy (HIE). Rectal temperature (RT) is used to determine the temperature set-points for treatment with HT, however experimental studies have shown significant differences between RT and brain temperature during HT. Knowledge of actual brain temperature during HT might allow better determination of optimal degree of cooling and improve outcomes.

Objectives: To compare measurements of brain temperature obtained by non-invasive radiometric thermometry (RadT) to direct tissue measurements in an experimental model of HT, and to use RadT in newborn infants with HIE undergoing HT.

Study Design: RadT measurements of brain temperature were compared to fiber optic (Luxtron) thermometry measurements placed at a depth of 1.5 centimeters into the brain of cooled miniswine. Following validation studies, brain RadT and RT measurements were continuously recorded in thirty infants with HIE during HT and rewarming.

Results: RadT and Luxtron probe temperatures were comparable in miniswine throughout a temperature range similar to therapeutic HT. RadT measurements of brain temperature were higher than RT in 60% of infants with HIE undergoing HT. Higher RadT measurements compared to RT were associated with cerebral white matter abnormalities (p = 0.01).

Conclusions: RadT provides a safe, passive and non-invasive way to measure brain temperature that can be used in the clinical setting. RadT may be helpful in determining the optimal degree of cooling and identifying infants at highest risk of brain injury.