Predicted thermophysiological responses of humans to MRI fields

A relatively simple two-node model of human thermoregulation was developed to predict response changes during MRI procedures. Subsequent modifications of the model simulated impairments in cardiovascular function in terms of altered skin blood flow. In the present work, the model was programmed to predict the consequences of certain procedures used in the clinic, namely, precooling of the patient to the prevailing environment and covering the patient with a light blanket. Some of the fundamental predictions of the model during 20-min MRI scans at a low SAR were tested on two male subjects in the clinical setting. The following conclusions may be drawn: (1) Precooling of the patient for 20 min to the prevailing ambient conditions, whether inadvertent or deliberate, has little value in terms of preventing a rise in body temperatures. At the conclusion of a subsequent 20-min MRI scan, even at SARs as low as 2 W/kg, the modest effects of precooling are all but eliminated. Thus, inadvertent precooling should be no cause for concern; deliberate precooling carries little advantage for the patient and wastes valuable time. (2) Use of a blanket during an MRI scan should be discouraged in the normal clinical setting except when the SAR is 2 W/kg or less. At higher SARs, this added insulation impedes convective and radiative heat loss through evaporation of sweat. The result is an increase of heat storage in the body and a greater rise in core temperature than would occur otherwise. (3) Clinical tests on two normal male subjects have provided limited confirmation of the predictions of the two-node model. During 20-min MRI scans at a whole-body SAR of 1.2 W/kg, core and skin temperatures, sweat rate, and judgments of thermal sensation and discomfort were very similar to predicted values. Unexpected findings of an incremental increase in core temperature with successive scans and a sweating rebound following each scan may be important for future investigation. (4) Although pleased with the limited confirmation of our predictions, we are constantly aware of the limitations of the two-node model to accurately predict thermoregulatory responses of patients undergoing clinical MRI of various body parts. It is essential to keep in mind that the simulations are based on RF exposure of the whole body; thus, the predicted increase in core temperature will be proportionately higher than would be the case if only a portion of the body were exposed within the MRI device.(ABSTRACT TRUNCATED AT 400 WORDS)