[Diagnosis and prognosis of cerebral ischemic disturbances course using a method of artificial neuronal networks]

Based on the data of examination of 224 patients with different stages of cerebral ischemic disturbances (CID) and 84 age-matched controls, an artificial neuronal network was constructed and tried in differential diagnosis of CID stages according to the data of transcranial ultrasonic dopplerography. Diagnostic efficacy of the network was 80% for sensitivity, 100% for specificity and 82.7% for reliability. A modeling of the influence of the main risk factors for cerebral ischemia and of the reserve state of cerebral hemodynamics for establishing the stage of CID was performed.