[Clinical and computer tomographic correlations in craniocerebral injuries: predicting outcome].
Statistic analysis of the results of examination of 114 patients with craniocerebral trauma disclosed clinico-computerized-tomographic signs which correlated with the outcomes of the disease. The possibilities of prognosticating the probable outcome of the trauma by mathematical statistic modeling were studied. Linear regressive models were constructed with consideration for only the clinical or for only the computerized-tomographic data or for both types. The informativeness of the model increased sharply (to 85%) when clinical and X-ray data were included in it. Linear discriminant models are marked by higher informative capacity (94%). It is shown that this trend in determining the severity of the trauma and planning adequate therapeutic measures is promising.