[Effects of noise on sleep. Part 1. Development of an automatic analysing system for all-night sleep polygraphy by...

In order to assess the effects of noise on sleep, the authors have developed a microcomputer system which determines the sleep stage based on an all-night EEG (electroencephalogram), rapid eye movement and an EMG (electromyogram). All the polygraphic parameters for each epoch (including spindle, rapid eye movement, alpha and delta waves, and amount of muscle tension), which are necessary to determine the sleep stage, were determined by a microcomputer using a digital data processing program. Recognition of EEG waves is based on Fujimori's method with some modifications. The rules of Rechtschaffen et al. were adopted for judging sleep stage with a slight modification. Data were obtained from six healthy students of a university. Each student was polygraphed for five to six nights under various conditions of noise exposure. Judgements of the sleep stage were made by two medical doctors. Using randomly selected 10 nights' data, the agreement between judgements by the microcomputer system and by the doctor was 77%. The percentage of agreement increased to 84% for the epochs in which the two doctors agreed. It takes about one hour to determine all-night sleep stages by this system.