Screening of patients with Huntington's chorea by pneumoencephalography: criteria for decision and probability of error.

The problem of making a decision in case of two clinical hypotheses is treated by the application of the Neyman-Pearson criterion. As an example, a pneumoencephalographic index (frontal horn width divided by septum-caudate distance) of two different diseases, Huntington's chorea and Parkinsonism, is analyzed. As the aim of the investigator is to detect the maximum number of cases of Huntington's chorea, an appropriately high alpha, (the probability of "false alarm") should be chosen, and maximization of 1-beta (the probability of detection of Huntington's chorea) carried out.