Periodic Reporting for period 1 - One-EG (Wearable Brain Monitoring Technology for Quick Diagnosis of Sleep Disorders)
Reporting period: 2016-01-01 to 2017-06-30
Summary of the context and overall objectives of the project
The focus of this proof of concept project was to create a novel, ultra-low power, tiny, and user friendly, Electroencephalography (EEG)-based wearable technology that could facilitate automatic monitoring of sleep. Such a technology would be useful for neurological research purposes and potentially to assist with the diagnosis of sleep disorders.
The technology that has been created, as a result, is unique because of its following aspects.
1. Size: Including batteries, the prototype system weights only 6g with a diameter of 23mm. This represents a massive reduction in both weight and size with respect to any other commercial technology. This also results in a device that is extremely comfortable to sleep with or to use for extended periods of time.
2. Power Consumption: The system can provide continuous monitoring of EEG signals for up to 36 hours. An average night of sleep is generally 8 hours. Hence, on a single charge, the system can monitor 4 nights of sleep without the need for the battery to be recharged.
3. Comfort: The system has been designed such that the user only has to peel an adhesive and click on the sensor on to their forehead. It takes less than 30 seconds to put the sensor on and start brain monitoring.
4. Ease of Use: Brain monitoring can be started and stopped with just a single tap using the accompanying user-friendly smartphone app. The app itself provides visual feedback about the status of the sensor. Additionally, data being recorded is streamed wirelessly to the smartphone and can be displayed in real-time on this app or recorded for offline access and analysis later.
5. Reliability: The system, once attached, stays in place for the whole night resulting in data being acquired with high integrity. This is made possible by designing the system enclosure in which the weight is uniformly distributed. This, together with the adhesive, results in a highly reliable electrode attachment.
6. Intelligent Signal Analysis: A novel algorithm for sleep staging has been implemented on the smartphone to provide sleep scores in real-time while the EEG signals are being acquired. The result can be visualised using an interactive bar plot showing different sleep stages. Additionally, other sleep parameters such as the total sleep time, time in bed, sleep latency, REM latency, wake up time, etc. are also computed that can assist with the diagnosis of several sleep disorders.
7. High Quality Data: Its ability to provide high quality raw one-channel EEG signals makes it a very useful tool for data collection in various neurological research applications.
It is estimated that 30% of the population in Europe, 20% in Japan and 50% in the USA have one or more sleep related problem. The social and financial impacts of these are huge. Only in Europe the overall annual costs of sleep disorders are over €35 billion. However, the direct costs of sleep disorders account for only 2% of their total costs since the major financial impact results from the indirect part they play in work related injuries, loss of productivity and motor vehicle accidents. As a representative example, twenty percent of all fatal road accidents are caused by someone falling sleep, with an alarming 37% of drivers having reported to have fallen asleep whilst driving at least once. Furthermore, poor quality sleep is associated with physical as well as psychological impairments, with sleep disorders representing a twofold risk factor for later occurrence of depression.
Diagnosis of sleep disorders is an expensive procedure that requires performing a sleep study to monitor multiple parameters including neural activity (EEG), eye movements (EOG), and muscle activity (EMG), amongst others. Together, these signals are used to perform sleep scoring to detect the presence of various sleep disorders. The main problem associated with diagnosing sleep disorders is the lack of diagnostic sleep clinics, where patients can be monitored using polysomnography (PSG). PSG is very time-consuming as it requires at least one night of sleep recording which is subsequently scored by a sleep specialist. Due to increasing healthcare costs and limited resources, access to PSG is severely limited. In cases where it is available, PSG is very expensive and uncomfortable for the patient, requires long time from medical specialists in order to both attach the sensors and interpret the signals, and has got very long waiting times. In the UK alone it is estimated that the demand exceeds the supply by a factor of 50 and this number is increasing. Although portable PSG systems are being used to screen patients at home and try to increase the diagnostic yield, these systems still require long time from qualified medical specialists to later on interpret the signals, and also, do not normally include the EEG channels which makes them only of some benefit for those disorders that involve respiratory disturbances. Furthermore, even in these cases, the lack of neurophysiological channels makes impossible sleep staging (hypnogram) and the use of actigraphy results in under or over estimation of many of the diagnostic indices.
The sleep monitoring system that has been created as part of this proof-of-concept grant is able to achieve unprecedentedly low levels of power consumption, which will effectively have a massive impact on the size and overall usability of a wearable sleep diagnostic system. As a result of its low power consumption, small size, and ease of use, it could be used to alleviate some of the aforementioned problems with the existing sleep monitoring systems. For example, it could be used on its own to collect and analyse sleep EEG signals providing the full patient’s hypnogram and potentially additional sleep diagnostic features (such as spindles and micro arousals). Together with other already existing cardio-respiratory sensors it can also provide additional diagnostic information for respiratory related sleep disorders.
In addition, a number of sleep, neurological, and mental conditions are diagnosed and managed on the basis of qualitative information, since there is no easy way of carrying out human research to try and identify quantitative biomarkers, and animal models are no equivalent enough. This is for example the case of depression (350 million people are affected in the world). Having a truly wearable EEG system will allow for long term brain monitoring and consequently the acquisition of signals that can be used to try to find new biomarkers for those diseases. The latter could potentially be used not just to better understand the condition and its progression but also to improve patients’ management and to carry out research on new treatments. This demonstrates the potential of the system to be used not only as a stand-alone sleep monitoring system for diagnosis but also as an enabling technology for large clinical trials in sleep medicine which would not be possible with the use of traditional PSG systems.
The technology that has been created, as a result, is unique because of its following aspects.
1. Size: Including batteries, the prototype system weights only 6g with a diameter of 23mm. This represents a massive reduction in both weight and size with respect to any other commercial technology. This also results in a device that is extremely comfortable to sleep with or to use for extended periods of time.
2. Power Consumption: The system can provide continuous monitoring of EEG signals for up to 36 hours. An average night of sleep is generally 8 hours. Hence, on a single charge, the system can monitor 4 nights of sleep without the need for the battery to be recharged.
3. Comfort: The system has been designed such that the user only has to peel an adhesive and click on the sensor on to their forehead. It takes less than 30 seconds to put the sensor on and start brain monitoring.
4. Ease of Use: Brain monitoring can be started and stopped with just a single tap using the accompanying user-friendly smartphone app. The app itself provides visual feedback about the status of the sensor. Additionally, data being recorded is streamed wirelessly to the smartphone and can be displayed in real-time on this app or recorded for offline access and analysis later.
5. Reliability: The system, once attached, stays in place for the whole night resulting in data being acquired with high integrity. This is made possible by designing the system enclosure in which the weight is uniformly distributed. This, together with the adhesive, results in a highly reliable electrode attachment.
6. Intelligent Signal Analysis: A novel algorithm for sleep staging has been implemented on the smartphone to provide sleep scores in real-time while the EEG signals are being acquired. The result can be visualised using an interactive bar plot showing different sleep stages. Additionally, other sleep parameters such as the total sleep time, time in bed, sleep latency, REM latency, wake up time, etc. are also computed that can assist with the diagnosis of several sleep disorders.
7. High Quality Data: Its ability to provide high quality raw one-channel EEG signals makes it a very useful tool for data collection in various neurological research applications.
It is estimated that 30% of the population in Europe, 20% in Japan and 50% in the USA have one or more sleep related problem. The social and financial impacts of these are huge. Only in Europe the overall annual costs of sleep disorders are over €35 billion. However, the direct costs of sleep disorders account for only 2% of their total costs since the major financial impact results from the indirect part they play in work related injuries, loss of productivity and motor vehicle accidents. As a representative example, twenty percent of all fatal road accidents are caused by someone falling sleep, with an alarming 37% of drivers having reported to have fallen asleep whilst driving at least once. Furthermore, poor quality sleep is associated with physical as well as psychological impairments, with sleep disorders representing a twofold risk factor for later occurrence of depression.
Diagnosis of sleep disorders is an expensive procedure that requires performing a sleep study to monitor multiple parameters including neural activity (EEG), eye movements (EOG), and muscle activity (EMG), amongst others. Together, these signals are used to perform sleep scoring to detect the presence of various sleep disorders. The main problem associated with diagnosing sleep disorders is the lack of diagnostic sleep clinics, where patients can be monitored using polysomnography (PSG). PSG is very time-consuming as it requires at least one night of sleep recording which is subsequently scored by a sleep specialist. Due to increasing healthcare costs and limited resources, access to PSG is severely limited. In cases where it is available, PSG is very expensive and uncomfortable for the patient, requires long time from medical specialists in order to both attach the sensors and interpret the signals, and has got very long waiting times. In the UK alone it is estimated that the demand exceeds the supply by a factor of 50 and this number is increasing. Although portable PSG systems are being used to screen patients at home and try to increase the diagnostic yield, these systems still require long time from qualified medical specialists to later on interpret the signals, and also, do not normally include the EEG channels which makes them only of some benefit for those disorders that involve respiratory disturbances. Furthermore, even in these cases, the lack of neurophysiological channels makes impossible sleep staging (hypnogram) and the use of actigraphy results in under or over estimation of many of the diagnostic indices.
The sleep monitoring system that has been created as part of this proof-of-concept grant is able to achieve unprecedentedly low levels of power consumption, which will effectively have a massive impact on the size and overall usability of a wearable sleep diagnostic system. As a result of its low power consumption, small size, and ease of use, it could be used to alleviate some of the aforementioned problems with the existing sleep monitoring systems. For example, it could be used on its own to collect and analyse sleep EEG signals providing the full patient’s hypnogram and potentially additional sleep diagnostic features (such as spindles and micro arousals). Together with other already existing cardio-respiratory sensors it can also provide additional diagnostic information for respiratory related sleep disorders.
In addition, a number of sleep, neurological, and mental conditions are diagnosed and managed on the basis of qualitative information, since there is no easy way of carrying out human research to try and identify quantitative biomarkers, and animal models are no equivalent enough. This is for example the case of depression (350 million people are affected in the world). Having a truly wearable EEG system will allow for long term brain monitoring and consequently the acquisition of signals that can be used to try to find new biomarkers for those diseases. The latter could potentially be used not just to better understand the condition and its progression but also to improve patients’ management and to carry out research on new treatments. This demonstrates the potential of the system to be used not only as a stand-alone sleep monitoring system for diagnosis but also as an enabling technology for large clinical trials in sleep medicine which would not be possible with the use of traditional PSG systems.