The Internet of Things and The Future of Sleep

In the 1980s, computer science students at Carnegie Mellon University (Pittsburgh, Pennsylvania), frustrated at going to the vending machine and finding it empty of Coke soft drinks (Coca-Cola Company, Atlanta, Georgia), had the idea of installing microswitches in the machine so they could track how many bottles were in it. The switches were hooked up to the main departmental computer¹ and the students wrote a server program that tracked the number of Coke bottles in the machine. Students could access the information from any computer through a web address on the internet. This was the first use of internet of things (IoT).

IoT now refers to the network of physical objects [i.e., “things”], which are embedded with sensors, processing abilities, software and other technologies, that connect and exchange data with other devices and systems over the internet or other types of communication network.² With these advancements, the use of IoT in the sleep field shows promise in improving the diagnosis and management of sleep disorders, primarily obstructive sleep apnea (OSA).

To discuss IoT requires first defining a few terms:

Big data: Extremely large and complex datasets that can be used to detect patterns, trends and associations, but are too large for traditional data processing software, which may be analyzed by computers to detect patterns, trends and associations. Examples of big data are e-mails, patients’ records and transaction processing systems.^3,4

Bluetooth technology: A wireless technology used to exchange data between fixed devices or mobile devices over short distances via radio waves. Examples of Bluetooth technology are wireless headsets used to talk on mobile phones and wireless computer keyboards.⁵

Cloud computing: The delivery over the internet (i.e., “the cloud”) of computing services (e.g., server [a machine that computes, stores and manages data, devices and systems over a network], storage, databases, networking, software, analytics and intelligence). A provider (e.g., internet service provider) manages and provides cloud computing services to a client (e.g., consumers). Examples of cloud computing are social media, media streaming, personal data storage, productivity software (e.g., Office 365 [Microsoft, Redmon, Washington] and Google Docs [Google, Mountain View, California], and online storage (e.g., Google Drive [Google]).^6,7

Fog computing: A way to prevent network congestion, data is processed from devices without reducing said data's quantity. This is done by introducing a new processing unit between the cloud and the user to enhance reliability, energy efficiency and privacy maintenance and reduce latency.⁸

Internet: A global system of interconnected computer networks that uses a set of rules, called protocols (e.g., hypertext transfer protocol secure [HTTPS]) that allows a group of interconnected computing devices to exchange information with each other. Protocols may be processed by hardware (i.e., the physical part of a computer such as a central processing unit [CPU]), software or a combination of these.^9,10

IoT and Sleep

Polysomnography (PSG) is the gold standard for assessing people with suspected OSA. However, a PSG study involves applying several sensors on a person to record brainwaves, airflow, thoracic and abdominal movements, leg movements and blood oxygen saturation. Patients often find the sensors uncomfortable and may struggle with going to sleep in a sleep center, which may impact a PSG recording.

The ability to wirelessly record brainwaves, airflow, thoracic and abdominal movements, leg movements and blood oxygen saturation throughout a study and relay the information to the cloud so that sleep physicians can access it would be ideal for counteracting the drawbacks of PSG that patients experience. However, a fully wireless PSG study that records all of this information is not available at this time.

Wireless technology used in home PSG studies is more suited to screening for OSA than screening for other sleep disorders such as restless legs syndrome (RLS). A home PSG study involves attaching various sensors (e.g., thoracic belt, abdominal belt, sleep position sensor, oximeter) into a device, which is attached to a patient. Signals from the sensors are relayed from the device to a smart phone or computer via radio waves (i.e., Bluetooth technology) for processing, and from there they are relayed to the internet. A sleep physician can then access the data to determine whether a person has OSA.

Benefits of Increased IoT in the Sleep Field

 IoT could make long-term ambulatory monitoring possible. Surrel et al.¹¹ recently developed a wearable system that can be used to monitor a patient for OSA over several days. The system contained a wearable electrocardiogram device (INYU sensor) that detects and predicts cardiovascular pathologies via monitoring electrocardiogram signals in real time.¹² Surrel designed an online algorithm (i.e., a specialized mathematical formula) that can be used to process changes in heart rate frequency to predict OSA. The system had a classification accuracy of 83.2% (i.e., the number of correct predictions divided by the total number of input samples), when compared with manual scoring, and the algorithm had low utilization of the device’s battery resources, which equated to the battery being able to provide 46 days of continuous OSA monitoring. The wearable device uploads its analysis to an online web service for continuous monitoring, thereby allowing tracking of the evolution of OSA over time.

IoT could make possible home-based monitoring of sleep disorders and treatment outcomes, especially for patients for whom visiting a sleep center would be difficult (e.g. elderly patients). To this end, Yacchirema-Vargas¹³ developed a system that combines fog and cloud computing, IoT and big data analysis and storage. This combination allows data to be seen in real time, thereby overcoming the delay from data collection until a diagnosis. At the fog level, a smart IoT gateway device acts as a router by transmitting (i.e., routing) data between IoT devices and the cloud and preprocesses IoT data to detect events in real-time. Once the data is transmitted to the cloud, it is managed, stored and injected into a big data analyzer for further processing and analyzing. Yacchirema-Vargas believes that, with quicker access to data, health professionals could improve their decision-making when monitoring and guiding sleep apnea treatment.

Finally, IoT could be used to improve a person’s sleep quality and sleep behavior. Takeuchi et al.¹⁴ developed a cloud-based health care IoT (HIT) system that continuously acquires health-related information via activity monitors on participant's non-dominant wrists such as momentary symptoms, biological signals and surrounding environmental information. The HIT system consists of a cloud server and a smartphone app that is set up to collect data on a person’s daily thoughts and behavior in their normal environment at or close to the time of the behavior. Additionally, the HIT app can connect with various IoT devices via Bluetooth technology to allow data to be transferred from the IoT devices to the HIT server, which can then store, integrate and manage data uploaded from the app and send personalized messages (i.e., push-type feedback messages) to app users.

In the feedback messages, a person is given information to improve their performance. In the Takeuchi study, participants were assigned to a feedback group or non-feedback group (i.e., control). Personalized sleep feedback messages were sent to participants in the feedback group, based on the sleep data. A sample message was, “You accumulated X minutes of sleep debt yesterday. Your current overall debt is X minutes. Sleep debt has adverse effects on physical and mental health. Adjust your daytime behavior to cancel your debt."  Questionnaires were then used to ask the participants, who were office workers, to indicate their current mood (e.g., depressed, anxious, stressed) and physical symptoms (e.g., sleepiness, fatigue and neck and shoulder stiffness) five times a day via a smartphone app.

Takeuchi found that compared to the control group, the feedback group had more stable sleep timing and less sleepiness, fatigue and neck and shoulder stiffness. Thus, the push feedback messages seem to promote sleep self-management and help individuals to address sleep behavior problems.

Limitations of IoT

Despite such encouraging findings, IoT has some limitations. The first is that although big data can be encrypted, large volumes of data are highly vulnerable to security threats, data breaches and cyberattacks. Thus, patients’ health information and sensitive data can be misappropriated by cybercriminals.

A second limitation of the increased use of IoT is total reliance on the technology. If any part of a IoT system (e.g., a fog device or the Internet) were to crash, no backup alternatives exist to prevent a disruption in the collection, analysis and transmission of sleep data. A third limitation is that implementing IoT and keeping the software needed for IoT updates is expensive.

Nevertheless, scientists continue to work to determine how best to incorporate IoT into the sleep field to improve the diagnosis and treatment of sleep disorders to a greater extent than is possible with current technology.   

References

1. Carnegie Mellon University Computer Science Department. The "only" Coke machine on the Internet. Carnegie Mellon University: Pittsburg, PA. https://www.cs.cmu.edu/~coke/history_long.txt
2. Oracle.com. What is IoT? https://www.oracle.com/internet-of-things/what-is-iot/
3. Oracle.com. What Is Big Data? https://www.oracle.com/big-data/what-is-big-data/
4. Kumari P. What is Big Data?/Uses of Big Data/Types Of Big Data/Big Data Analytics Interview Questions. LinkedIn.com. 30 May 2023. https://www.linkedin.com/pulse/whatbig-data-uses-types-analytics-interviewquestions-jha/
5. CISA.gov. Understanding Bluetooth Technology. 1 Feb 2021. https://www.cisa.gov/news-events/news/understandingbluetooth-technology
6. IBM.com. What is cloud computing. https://www.ibm.com/topics/cloud-computing
7. Frankenfield J. What is Cloud Computing? Pros and Cons of Different Types of Services. Investopedia. https://www.investopedia.com/terms/c/cloud-computing.asp
8. Elhadad A, Alanazi F, Taloba AI, Abozeid A. Fog Computing Service in the Healthcare Monitoring System for Managing the Real-Time Notification. J HealthcEng. 2022 Mar 15;2022:5337733. doi: 10.1155/2022/5337733
9. Rouse M. Internet. Techopedia.com. https://www.techopedia.com/definition/2419/internet
10. Cloudflare.com. How does the internet work? https://www.cloudflare.com/learning/network-layer/how-does-the-internet-work/
11. Surrel G, Rincón F, Murali S, Atienza D. Low-power wearable system for real-time screening of obstructive sleep apnea. 15^th IEEE Computer Society Annual Symposium on VLSI (ISVLSI). Held at: Pittsburgh, PA on 2016. IEEE. 2016; p. 230-235
12. Surrel G, Aminifar A, Rincon F, et al. Online obstructive sleep apnea detection on medical wearable sensors. IEEE Transactions on Biomedical Circuits and Systems. 2018;12:762-773. doi: https://doi.org/10.1109/TBCAS.2018.2824659
13. Yacchirema D, Sarabia-Jácome D, Palau CE, Esteve M. System for monitoring and supporting the treatment of sleep apnea using IoT and big data. Pervasive and Mobile Computing. 2018;50:25-40
14. Takeuchi H, Suwa K, Kishi A, et al. The effects of objective push-type sleep feedback on habitual sleep behavior and momentary symptoms in daily life: mHealth intervention trial using a health care internet of things system. JMIR mHealth and uHealth. 2022;10:e39150. doi: https://doi.org/10.2196/39150