Interests in Human Stroke Evaluation and Monitoring

Background Stroke is a major health problem, affecting approximately 700,000 persons annually in the US alone, with more expected as the average age and life span increase. When a patient enters an emergency room and stroke is suspected, standard procedure is to first use x-ray CT to rule out brain hemorrhage. If the CT is negative, but stroke is concluded based upon behavioral and cognitive measures, then the stroke is assumed to be ischemic, and thrombolytic drugs, e.g., t-PA, may be used to restore blood flow within the ischemic penumbra. The development of tPA gives new hope for many stroke patients, but since it carries a 6% risk of causing brain hemorrhage, it must is used conservatively. Current guidelines suggest that t-PA be used only within three to six hours of stroke onset. Since many victims arrive for treatment beyond this window and diagnosis takes some time, it is likely that many patients are excluded from treatment unnecessarily. Others may be exposed to risk unnecessarily if the stroke is small or has already reached completion. Stroke treatment could be much better directed if quantitative measures of the physiological state of ischemic tissue were available. X-ray CT and conventional MRI do not show ischemic tissue which is still viable for rescue, but only dead tissue in the infracted core. Diffusion tensor-weighted magnetic resonance imaging (MRI) is now being developed to detect ischemic tissue, but MRI machines are expensive, not portable, and not practical for continuous monitoring. Thus MRI technology would be well complemented by a practical, affordable imaging methodology quantifying the physiological state of ischemic tissue. Commercial EEG/EIT acquisition systems which meet these criteria are now being developed by Electrical Geodesics, for example, but much work remain to be done to understand how to analyze these data effectively. The following three topics address the spatial, temporal and theoretical aspects of this problem.