Due to a globally aging population, there is a growing demand for smart home technology which can serve to monitor the health and safety of adults. Therefore, sleep monitoring has emerged as a crucial tool to improve the health and autonomy of adults. While polysomnography (PSG) is an effective and accurate tool for sleep monitoring, it is obtrusive as the user must wear the instruments during the experiment. Therefore, there has been a growing interest in deploying unobtrusive sleep monitoring devices, specifically for long-term patient monitoring. This thesis proposes multiple algorithms applicable to unobtrusive pressure sensitive sensor arrays in order to assess sleep quality. These algorithms can be listed as adaptive movement detection, sensor data fusion and bed occupancy detection. This thesis also investigates long-term sleep pattern changes from previously recorded data. The methods developed in the thesis can be of interest for future clinical remote patient monitoring systems.

Sleep related breathing irregularities and sleep disturbances affect a surprisingly large number of people in the society. Due to the high risks of chronic and acute health situations associated with sleep disturbances, a robust sleep monitoring system is needed. While the current golden standard for sleep monitoring is the Polysomnograph (PSG), other approaches also require attachments to patients' body. Furthermore, these monitoring techniques are performed in sleep clinics. Therefore, they interfere with natural sleep patterns. Finally these techniques are usually expensive. The Intelligent Bed Sensor is proposed as a non-restraining home-based sleep monitoring system. It uses 144 pressure sensors embedded in a bed sheet for measuring the pressure that the patient's body exerts on the bed. The main theoretical contribution of this work is a new methodology for analyzing periodicity in pressure data via Computer Vision techniques. We prove that the Intelligent Bed Sensor is capable of detecting individual respiration cycles, apnea events and movements accurately.

Sleep related breathing irregularities and sleep disturbances affect a surprisingly large number of people in the society. Due to the high risks of chronic and acute health situations associated with sleep disturbances, a robust sleep monitoring system is needed. While the current golden standard for sleep monitoring is the Polysomnograph (PSG), other approaches also require attachments to patients' body. Furthermore, these monitoring techniques are performed in sleep clinics. Therefore, they interfere with natural sleep patterns. Finally these techniques are usually expensive. The Intelligent Bed Sensor is proposed as a non-restraining home-based sleep monitoring system. It uses 144 pressure sensors embedded in a bed sheet for measuring the pressure that the patient's body exerts on the bed. The main theoretical contribution of this work is a new methodology for analyzing periodicity in pressure data via Computer Vision techniques. We prove that the Intelligent Bed Sensor is capable of detecting individual respiration cycles, apnea events and movements accurately.

Introduktion: Sömn är en erkänt viktig del för idrottslig prestation. Intresset och möjligheterna att övervaka den har ökat betänkligt till följd av antalet tillgängliga sömnmonitorerande bärbara verktyg (SBV) idag. Denna utrustning bidrar med en möjlighet att kontinuerligt samla in data, men till följd av den snabba utvecklingen av dessa verktyg finns det ett ständigt behov av mer valideringsarbete.   Syfte: Syftet med litteraturöversikten var att undersöka och redogöra hur väl olika sömnmonitorerande bärbara verktyg kan monitorera sömn i relation till polysomnografi (PSG), nivå I bland friska individer.   Metod: En strukturerad litteratursökning utfördes i databaserna SPORTdiscus och PubMed under tidsperioden 2023-03-15 till och med 2023-04-06 med syftet att identifiera artiklar som jämförde kommersiellt tillgängliga SBV mot PSG, nivå I bland friska individer. Vi utvärderade kvaliteten och risken för bias för de inkluderade studierna genom en modifierad checklista av Downs & Black (1998).  Resultat: Efter en granskning av 213 artiklar inkluderades 11 av dem, innehållandes totalt 368 deltagare. Resultatet visade att SBV har svårigheter att korrekt estimera tid spenderad i respektive sömnfas i jämförelse mot PSG. Förmågan att uppskatta Total Sleep Time (TST) och Wake After Sleep Onset (WASO) varierade stort mellan de olika modellerna. Gällande Sleep Onset Latency (SOL), hjärtfrekvensen (HR) och hjärtfrekvensvariabiliteten (HRV) var det få statistiskt signifikanta resultat samt antal studier som undersökte dessa tre variabler.  Konklusion: I dagsläget har SBV problem att uppmäta tiden spenderad i respektive sömnfas, och uppvisar varierande resultat i fråga om estimeringen av TST och WASO. Gällande SOL, HR och HRV fanns det inte tillräckligt med underlag i de inkluderande studierna för att dra slutsatser om förmågan hos SBV att uppskatta dessa variabler. Individer bör vara försiktiga vid appliceringen av data från verktygen, samt medvetna om de eventuella risker som kan tillkomma när denna data introduceras inom den atletiska populationen. / Introduction: Sleep is widely recognized as an important factor for athletic performance. The interest and possibility of measuring it has increased dramatically following the number of available wearable devices today. These wearables provide an opportunity to continuously collect data in a home environment, but because of the rapid increase of commercial availability, there remains a constant need for more validation of the most recent models.   Purpose: The purpose of this review was to examine the capability of different wearables to monitor sleep in comparison to polysomnography (PSG), level I in healthy individuals.  Method: A structured literature review was performed in the databases SPORTDiscus and PubMed under the period 2023-03-15 until 2023-04-06 with the aim of identifying relevant articles that compared commercially available wearables to PSG level I in healthy individuals. We assessed the quality and risk of bias of the included studies with a modified questionnaire from Downs & Black (1998).  Results: After screening 213 articles, 11 of them were included, which in total amounted to 368 participants. The result showed that wearables struggle to correctly estimate the time spent in each sleep-stage in comparison with PSG. The capability to assess Total Sleep Time (TST) and Wake After Sleep Onset (WASO) varied between the different models. There were few statistically significant results of the ability to measure Sleep Onset Latency (SOL), Heart Rate (HR) and Heart Rate Variability (HRV).  Conclusion: Wearables have, at present time, a difficult time correctly estimating time spent in each sleep-stage and show varied results in monitoring TST and WASO. Too few studies analyzed SOL, HR and HRV to draw conclusions regarding these variables. Individuals should be cautious when implementing the data from these devices, and aware of the potential risks when it is used with athletes.

The new Health 4.0 concept brings the idea of combining modern technologies from field of science and technology with research in healthcare and medicine. This work realizes a system of secured actigraph data transfer and preprocessing based on the concept of Health 4.0. The system is successfully designed, implemented, tested and secured. With the help of a non-invasive method of monitoring the movement and temperature of the subject using the GENEActiv actigraph allows the system to securely transfer, process and evaluate the subject's sleep data using the machine learning algorithm XGBoost. The proposed system is in accordance with the valid law of the Czech Republic and meets legal requirements.

我們提出的智慧型睡眠監測系統，是基於影像分析技術進行睡眠品質觀測，並利用所得到的數據來推斷最佳的喚醒時間。此系統命名為iWakeUp，利用非接觸式的方法來收集影像資料並進行後續處理，此裝置將被安裝在一般的臥室來幫助睡眠者，以期成為增進智慧家庭生活品質的一環。在此論文中，我們將會描述iWakeUp的各個模組包括測定動作量、推斷睡眠階段乃至於如何建立喚醒機制。更特別的是，我們考慮了喚醒時間與喚醒機制的關係，於較早的時間喚醒必須具有更高的信心度，否則將付出較大的代價，反之亦然。另外為了處理晨間臥室中的光影變化，不同的背景模型也已被整合測試，以期讓系統可以提升長時間觀測的準確度。最後，我們也進行了使用iWakeUp的臨床實驗，結果指出使用iWakeUp喚醒的睡眠者具有較低的嗜睡感與更好的活力。 / We present a video-based monitoring system to determine the sleep status and optimal wakeup time in this thesis. We envision a smart living space in which a data collection and processing module named iWakeUp is installed in the bedroom to record and monitor sleep in a non-invasive manner. We describe the overall structure of the iWakeUp system, including the procedure to measure amount of motion, the method for inferring wake/sleep status from the acquired video and the logics for deciding the optimal wakeup time. In particular, a time-dependent decision rule has been incorporated to account for unequal penalties when classification error occurs. Furthermore, various background modeling techniques have been examined to address lighting changes at dawn in the bedroom for long-term monitoring. Validation experiments are carried out to compare the alertness level upon awakening with/without reported a lower level of sleepiness and higher level of vigorousness in comparison to the control group.