The University of Tokyo developed an original machine learning algorithm and used it to evaluate waking and sleep states through arm acceleration. They then turned the arm acceleration data from 100,000 U.K Biobank volunteers into sleep data and analysed it. The team came up with 16 sleep patterns from their trial.
The volunteers whose data researchers used for the study were between 30 and 60. The participants were measured for seven days with a wristband accelerometer. Researchers used the data they had received and converted it into 21 indicators.
The team looked into sleep patterns associated with social jet lag
Afterwards, they used clustering and reduction methods and grouped the sleep pattern into eight clusters. Some of the clusters they considered were linked to social jet lag. Others were linked to mid-onset awakening.
The team also looked into insomnia. This disorder enabled them to extract clusters with a correlation to sleep disorders and lifestyle. Two of the most significant sleep indicators researchers evaluated were immediate waking time and sleep duration. These indicators are affected in people with sleep disorders.
The researchers split their clusters into eight groups. These groups were associated with evening and morning types. Some of the clusters also had a link to insomnia. Moreover, they could link seven sleep patterns to the sleeping disorder.
This study helped the team analyse sleep patterns on a large scale. They were also able to evaluate sleep patterns which people link to lifestyles, including morning/evening persons and social jet lag. These parameters can be difficult to evaluate using short-term PSG measurements.
The team didn’t use conventional methods to come up with these clusters. Moreover, they hope that scientists could use this technique to classify and diagnose sleep disorders like insomnia.
This data could help people with sleep disorders
The study comes at a good time since the number of people who have complained about having insufficient sleep and feeling anxious about sleep has gone high. Furthermore, more people are reporting sleep disorders. Fortunately, this study can serve as a tool to help people.
The study also helps scientists understand the sleep-wake cycle, its impact on biological function and how abnormal situations arise.
