Introduction Polysomnography (psg), Consisting of Several Phys- Iological Measures Such as Electroencephalogra- Phy (eeg) and Electromyography (emg), Is Widely Considered as the Golden Standard to Differenti- Ate Sleep from Wakefulness in Sleep

POLYSOMNOGRAPHY (PSG), CONSISTING OF SEVERAL PHYSIOLOGICAL MEASURES SUCH AS ELECTROENCEPHALOGRAPHY (EEG) AND ELECTROMYOGRAPHY (EMG), IS WIDELY CONSIDERED AS THE GOLDEN STANDARD TO DIFFERENTIATE SLEEP FROM WAKEFULNESS IN SLEEP RESEARCH. However, it is not well suited for long-term and large-scale studies or for measurements outside of the laboratory. Therefore, unobtrusive, userfriendly and low-priced devices are called for measuring sleep/wake data. Several methods have been proposed to quantify sleep quality, such as sleep diaries, video-monitoring, transducers attached to the body or different mattresses measuring activity during sleep, and various active or passive switches detecting the sleep onset. Wrist actigraphy is the most widespread alternative to PSG to evaluate rest-activity cycles. An actigraph measures acceleration providing information on the activity of the user. The device can output either an acceleration measure or counts exceeding a given acceleration threshold, e.g. 0.1 g. In sleep studies, the latter choice is popular. The acceleration signal is sampled frequently and aggregated into epochs lasting e.g. 1 minute. Sleep/wake classification is made in epoch-by-epoch basis using mathematical methods.1-3 The ability of actigraphy to detect sleep/wake states is based on the fact that the wrist is moved more awake than asleep. Actigraphy is an unobtrusive and low-priced technique, but it is also less accurate than PSG, and it cannot separate different sleep states. As Sadeh et al4 listed results from several studies, the agreement rates, i.e., the agreement percents in epoch-by-epoch basis, were around 90% between PSG and actigraphy. The range was 85-96% for normal individuals and heterogeneous groups of patients. In addition, high correlation in total sleep time (TST) (r=0.90) was reported (range 0.81-0.98).4 The accuracy of actigraphy is considerably worse in insomniacs than in subjects suffering no sleep disorders.4 However, actigraphy is well suited to track trends in patients’ sleep cycles for prolonged periods.5 In addition to nighttime analysis, actigraphy can be used for 24-hour recording allowing nap analysis.6 Accelometers can also be applied, for example, to the problems related to the activity of the patient7 or to the evaluation of energy consumption.8,9 Potentially, actigraphy might provide valuable information in longterm monitoring of changes in health status, effects of therapy or medication, etc.10 However, a limiting factor with the currently available actigraphs is the monitoring period, which is typically limited to a few weeks. Furthermore, the data are not available online. The Vivago WristCare® (IST International Security Technology Oy, Helsinki, Finland; http:\\www.istsec.fi) is a wrist-worn online activity monitoring device, which is designed to be used as an automatic personal alarm system for the elderly and chronically ill (Figure 1). The system is essentially a social alarm system with a manual alarm button, but, in addition to this, it has advanced automatic features, which enable it to call help if the wearer is not capable of doing so, e.g., due to loss of consciousness. The wrist unit provides an activity signal, which is constructed from the measured force change at the unit’s movement sensor. The sensor is sensitive enough to record, e.g., muscle movements inside the wrist in addition to the hand movements. The automatic alarm functions are mainly based on the activity data, e.g., alarms are generated in case of unusually long period of passivity or total lack of movements. In addition the unit measures wrist temperature that may also be used for alarming. Skin conductivity measurement is used to automatically detect whether the unit is worn on the hand. The complete system consists of a wrist unit