Alarm upon detection of impending sleep state

Details Alarm upon detection of impending sleep state Alarm upon detection of impending sleep state

2000-12-15

2002-12-24

Hindenburg, Max (Department: 3736)

Surgery

Diagnostic testing

Via monitoring a plurality of physiological data, e.g.,...

C600S544000, C600S546000, C600S300000

Reexamination Certificate

active

06497658

ABSTRACT:

BACKGROUND OF THE INVENTION
The present invention relates generally to the field of biofeedback. More particularly, the present invention relates to a device for detecting when a person is about to fall asleep and providing a feedback to the person indicating the impending sleep state.
Sleep deprivation is a common problem among airline pilots and other workers who are required to perform complex technical tasks over a prolonged period of time. Loss of sleep and fatigue in these settings can significantly impair cognitive function, and can lead to dangerous decrements in human performance. In particular, falling asleep at the controls of a commercial vehicle such as a truck, tanker, or airplane can lead to potentially catastrophic consequences. Besides the transportation context, drastic consequences may result by an operator falling asleep at the controls of a nuclear power plant, an electrical power grid, or other system that could impact many lives.
A number of proposals have been advanced to detect sleepiness by measuring various aspects of a subject's physiology or actions.
For example, headgear has been designed to measure eye and head movements. The resulting pattern of movements may then be correlated to sleep state, logged, and the log transmitted to a remote location. Unfortunately, detection of sleep onset based on changes in body movements is not a sufficiently timely method because the subject is likely quite inattentive by the time such sleep indications begin to occur. This method also requires sampling of the signal over a prolonged period of time, e.g., >30 seconds. This method. Is also subject to artifact if sleep onset is not associated with the anticipated eye and head movements. For further details, refer to U.S. Pat. No. 4,836,219, issued to Hobson, et al.
Automated monitoring of eye movement alone, as a surrogate for sleep, has also been proposed. It has the advantage of being a noninvasive monitor. However, it has a number of difficulties. In order to assess blink rate and eyelid speed, a normal baseline must be established. Thus, an extensive calibration time is required. Additionally, if a baseline is established during an already drowsy state, such a system may be unable to detect the relatively small change occurring with transition from drowsiness to sleep onset. Such a system is also susceptible to false positive measurements, because an increase in blink rate may not represent drowsiness, but rather dust in the eye or surprise. For further details, refer to U.S. Pat. No. 5,570,698, issued to Liang et al.
It has also been proposed to monitor a number of physiological changes at once and correlating them to drowsiness. Five parameters are described: pulse rate variability, vasomotor response, muscle tone, blood flow and reaction time. This method suffers from the disadvantage that there is no evidence firmly linking these physiological changes to sleep. For example, there may be several other explanations for change in vasomotor response. A high rate of false positive alarms is to be expected with this method. This method also requires baseline measurements that must be obtained over an extended period of time. This significantly limits the applicability of the method. As mentioned above, if the baseline is established during an already drowsy state, the small change from drowsiness to sleep may not be detectable. Additionally, physiological monitoring of sleep onset only detects the manifestations of sleep, and not the inattention that results from sleep onset. For further details, refer to U.S. Pat. No. 5,917,415, issued to Atlas.
A vehicle control device has been proposed that analyzes steering inputs and correlates these to the state of wakefulness of the operator. Specifically, measurements such as swerving movements (yaw rate), lateral movement, speed, and response time of the driver to these parameters are input to a computer, which sounds an alarm when values exceed certain preset parameters. For further details, refer to U.S. Pat. No. 5,815,070, issued to Yoshikawa. Similarly, it has been proposed to detect driver condition by comparison of driver inputs with previously defined baseline values. For further details, refer to U.S. Pat. No. 5,821,860 issued to Yokoyama et al. These more indirect methods suffer from most of the same problems note above with respect to methods that monitor head and eye movement and other physiological parameters. Of course, variations in both road and driver can lead to a highly variable baseline, and thus can only detect a case where the driver's condition is extremely degraded. These methods would require significant time to determine abnormal driving conditions, and may not detect degradation until too late to effectively arouse the driver.
Monitoring of EEG signals has also been proposed for monitoring wakefulness. The proposed device relies primarily on comparisons of driver response and road conditions. Such measurements are likely to be extremely variable from driver to driver and road to road, and false alarms are likely to be frequent. For further details, refer to Japanese Laid-Open Patent Publication (Kokai) no. 5-96971.
It has also been proposed to monitor EEG and use frequency analysis of the EEG to obtain “high frequency” (>30 Hz) components in order to predict sleep. This method suffers from a number of inaccuracies and disadvantages. Little data exists to firmly verify the hypothesis that EEG power in the high frequency range correlates reliably with drowsiness. No validation of this measure in explicitly sleep deprived humans has been provided. Moreover, it is unlikely that a continuous measure of drowsiness is available, as recent work suggests that drowsiness represents more an inability to maintain the waking state (i.e., propensity to enter the sleep state) than an ongoing alteration in waking brain function. Additionally, analysis of high frequency noise is technically difficult. Electrical noise, muscular movements (eyeblinks), etc. generate significant artifact. Accordingly, it is likely that false positives would be common. For further details, refer to U.S. Pat. No. 5,813,993, issued to Kaplan et al and Japanese Laid-Open Patent Publication (Kokai) no. 6-292661, by Mamoru et al.
Another proposal is to use a complex “neural network” to compare raw, unprocessed EEG signals to a known baseline signal to determine the probability of sleep. Such a device is proposed primarily for use as a portable sleep-scoring device, but may be adapted for use as a vigilance monitor. According to such a system, EEG signal is correlated to sleep stage, producing a “wakeogram.” Such a correlation is based on human observation of the EEG, which has not clearly been correlated to loss of alertness. The proposed system requires a period of extensive prior calibration to train the processor. Such a need for prior calibration significantly reduces the applicability of their device. For further details, refer to U.S. Pat. No. 5,999,846, issued to Pardey et al.
Another system proposes monitoring of EEG signals, the transmission of those signals to a remote location. It does not, however, propose logging brain signals to a third party for data analysis. For further details, refer to U.S. Pat. No. 6,011,991, issued to Mardirssian.
Sleep onset is characterized by specific changes in the human electroencephalogram (EEG) and electromyelogram (EMG). Skilled operator analysis of such changes in the raw EEG can readily prevent humans from falling asleep, and has been used in clinical medicine to study the physiology of sleep states. For example, refer to Leibenluft, E., Moul, D. E., Schwartz, P. J., Madden, P. A., and Wehr, T. A., “A Clinical Trial of Sleep Deprivation in Combination With Antidepressant Medication” Psychiatry Research, 46(3): 213-27, March 1993 (93262082).
Automated devices based on computer interpretation of the raw EEG also exist. These devices, however, are large, unwieldy, and do not process the EEG in real time. Hence, they cannot be used to prevent inadvertent sleep.
Thus, what is n

Affiliated with

Also associated with

Rating

Alarm upon detection of impending sleep state does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Alarm upon detection of impending sleep state, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Alarm upon detection of impending sleep state will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2951010