Neurometrics in cerebral ischemia and uremic encephalopathy.

The neurometric method as introduced by John was used to study three groups of patients with cerebral ischemia, three groups of patients with renal disease and an additional normal control group. The traditional neurometric approach was slightly modified: relative band power values were not expressed as a percentage of the total power per derivation but as a percentage of the "global power"; frequency matrices were used in addition to power matrices. From the study of the three groups of patients with one-sided supratentorial ischemia it appeared that sensitivity and specificity are completely satisfactory when using neurometrics in patients with severe ischemia in the middle cerebral artery territory studied within 48 hours of the onset of the stroke. However, in ischemia patients with less pronounced clinical signs and especially in patients without persistent neurological deficit the sensitivity is much lower. In studying dialysed and non-dialysed renal patients signs of an (often subclinical) encephalopathy could be detected in approximately 37% of all patients. Follow-up studies of the ischemia patients and the renal patients over a period of several years revealed a parallelism between clinical scores and qEEG scores in the ischemia patients; almost all qEEG improvement occurred in the first three months after the stroke. The qEEG profile of the groups of dialysed patients tended to be more or less stable over a period of several years.