[Automated diagnosis of vibration disease].

Based on a multidimensional discriminant analysis of the clinical and physiological characteristics derived as a result of examination of 593 miners, classification functions were plotted, allowing the diagnosis of vibratory disease in the automatic mode by means of mathematic computations. Functional indicators of skin sensitivity, particularly vibratory, dynamometry, rheovasography and thermography of the hand and fingers, a "white spot" symptom, and characteristics derived on a mathematic analysis of heart rhythm were established to be informative for automated diagnosis. The diagnosis was found to be more accurate in case of using a two-stage variant of the discriminant analysis where the differentiation is first made between healthy workers and workers with a pathology, followed by the discrimination of the preclinical and clinical phase of the disease. The automated diagnostic system is demonstrated to compare very favourably with the conventional method of diagnosis.