The current method for measurement of local cerebral metabolic rate of glucose (LCMRG) has relatively small errors when applied to normal healthy tissue. When applied to ischemic regions in brains of stroke patients, the method gives estimates which average about 50% too low. Therefore, the authors introduce a modified Bayes regression (BR) to compute the LCMRG, and use computer simulation studies to demonstrate that BR has relatively small errors for either ischemic or normal tissue. These improved estimates are within the clinical constraints of a one-hour data collection period.
|Original language||English (US)|
|Title of host publication||Proceedings - Annual Symposium on Computer Applications in Medical Care|
|Editors||Gerald S. Cohen|
|Number of pages||4|
|Publication status||Published - 1984|
ASJC Scopus subject areas