PNAS: Classification of 16 Adult Sleep Patterns Based on Large-Scale Sleep Analysis

In recent years, due to the diversification of life>

A research group at the University of Tokyo, led by Prof.

The research team first focused on arm acceleration data from about 100,000 people in the UK Biobank

Thus, by analyzing sleep at scale, they revealed the landscape of human sleep phenotypes

These results were obtained through the "Ueda Biotiming Project" ERATO program funded by the Japan Science and Technology Agency (JST)

(1) ACCEL: The research team's original sleep judgment algorithm

(2) UK Biobank: A large research database containing genetic and health information on approximately 500,000 UK participants

(3) Sleep data: time-series data at 30-second intervals, marked as sleeping or awake

(4) Dimensionality reduction method: a method to reduce the dimension of data

(5) Clustering method: A method of classifying data into clusters according to the similarity between them

(6) Upper and lower percentages: In any given percentage, when the values ​​are in descending order, it is called the upper percentage

(7) Polysomnography (Polysomnography, PSG): Polysomnography is to connect multiple electrodes and sensors to the subject to measure brain waves, eye movements, breathing status and electrocardiogram status