Researchers at the University of Tsukuba develop a machine-learning algorithm for automatically classifying the sleep stages of lab mice. Combining two techniques, they achieve 96.6% accuracy, which may help accelerate sleep research
Tsukuba, Japan – Researchers at the University of Tsukuba have created a new artificial intelligence program for automatically classifying the sleep stages of mice that combines two popular machine learning methods. Dubbed “MC-SleepNet,” the algorithm achieved accuracy rates exceeding 96% and high robustness against noise in the biological signals. The use of this system for automatically annotating data can significantly assist sleep researchers when analyzing the results of their experiments.
Scientists who study sleep often use mice as animal models to better understand the ways the activity in the brain changes during the various phases. These phases can be classified as awake, REM (rapid eye movement) sleep, and non-REM sleep. Previously, researchers who monitored the brainwaves of sleeping mice ended up with mountains of data that needed to be laboriously labeled by hand, often by teams of students. This represented a major bottleneck in the research.
Now, researchers at the University of Tsukuba have introduced a program for automatically classifying the stage of sleep that a mouse experienced based on its electroencephalogram (EEG) and electromyogram (EMG) signals, which record electrical activity in the brain and body, respectively. They combined two machine learning techniques, convolutional neural networks (CNN) and long short-term memory (LSTM) recurrent neural networks, to achieve accuracies that surpass those of the best existing automatic methods.
“Machine learning is an exciting new field of research with important applications that combine medicine with computer science. It allows us to automatically classify new data based on labeled examples,” corresponding author Kazumasa Horie explains. This is especially valuable when the patterns to look for are not well known, as with sleep stages. In this way, the algorithm can ‘learn” how to make complex decisions without being explicitly programed. In this project, the accuracy was very high because of the large dataset used. With over 4,200 biological signals, it was the biggest dataset of any sleep research so far. Also, by implementing a CNN, the algorithm showed high robustness against individual differences and noise.
The main advance in this work was to divide the task between the two machine learning methods. First a CNN was used to extract features of interest from the recordings of the electrical activity in the brain and body. These data were then passed to an LSTM to determine which features were most indicative of the sleep phase the mouse was experiencing. “We are optimistic that we can translate this work into classifying sleep stages in humans”, senior author Hiroyuki Kitagawa says. In the meantime, this program can already speed up the work of researchers in the field of sleep, which may lead to a much clearer understanding of how sleep operates.
Learn more: Machine Learning That Works Like a Dream
Go deeper with Bing News on:
- Parental enforcement of bedtime yields solid results for sleep-deprived teenson January 9, 2020 at 4:38 pm
Among the key findings: Parent-enforced bedtimes--along with later school start times--are the greatest predictors of sleep duration, daytime energy level, and depressive symptoms. More than 50 ...
- Sleep Number’s Climate360 smart bed warms and cools sleepers for better comforton January 9, 2020 at 11:45 am
With more than 700 million sleep sessions of data and research, our digital health platform is linking quality sleep to individualized wellness, and our 360 smart beds are delivering proven higher ...
- Parents aren't powerless when it comes to sleep-deprived teenagerson January 9, 2020 at 9:32 am
Among the key findings: Parent-enforced bedtimes—along with later school start times—are the greatest predictors of sleep duration, daytime energy level, and depressive symptoms. More than 50 percent ...
- Just one bad night's sleep 'raises levels of an Alzheimer's biomarker by 17%'on January 9, 2020 at 7:12 am
Just one bad night’s sleep can raise levels of an Alzheimer’s biomarker in the blood, research suggests. Scientists from Uppsala University in Sweden deprived 15 healthy men of a decent night’s kip.
- Sleep Number partners with Mayo at CES; Climate360 bed takes home trophyon January 8, 2020 at 7:05 pm
in Las Vegas to announce new products. Of course, Sleep Number Corp. did that (keep reading), but it also made a different kind of announcement Wednesday from its booth in the Sands Expo and ...
Go deeper with Google Headlines on:
Go deeper with Bing News on:
Sleep research and machine learning
- Going Beyond Machine Learning To Machine Reasoningon January 9, 2020 at 8:05 pm
Machine learning is necessary for systems to gain value from data, but it just pushes the problem of intelligence to the next level: reasoning.
- Machine learning shapes microwaves for a computer's eyeson January 9, 2020 at 1:02 pm
"The transmitter and receiver act together and are designed together by the machine learning algorithm," said Mohammadreza Imani, research assistant in Smith's lab. "They are jointly designed and ...
- New method uses MRI, machine learning to forecast genetic mutations in glioma tumorson January 9, 2020 at 9:56 am
Researchers at Osaka University have developed a computer method that uses magnetic resonance imaging (MRI) and machine learning to rapidly forecast genetic mutations in glioma tumors, which occur in ...
- Machine Learning As A Service (MLaaS) Market Forecast 2017-2024 by Technology, Leading Companies and Demandon January 7, 2020 at 9:50 pm
Machine learning refers to the artificial intelligence software which predicts the outputs without any explicit program. Instead of supervising the unsupervised program, machine learning learns from ...
- Machine Learning as a Service Market 2020 Analysis Forecast Research and Development Report 2025 – Market Research Engineon January 7, 2020 at 8:26 pm
New York, January 08, 2020: The Machine Learning as a Service Market is segmented on the lines of its component, organization size, end-user, application, and regional. Based on component segmentation ...