Validation of non-REM sleep stage decoding from resting state fMRI using linear support vector machines
A Altmann1, M S Schröter2, V I Spoormaker3
1Max Planck Institute of Psychiatry, Department of Translational Research in Psychiatry, Neuroimaging, Munich, Germany; Stanford Center for Memory Disorders, Department of Neurology and Neurological Sciences, Stanford University, Stanford, CA, USA.
Neuroimage
|November 25, 2015
Summary
Brain connectivity patterns detected by resting-state fMRI (rs-fMRI) can distinguish wakefulness from non-rapid eye movement (NREM) sleep stages. This method offers a way to interpret rs-fMRI data without simultaneous electroencephalography (EEG).
Area of Science:
- Neuroscience
- Cognitive Science
- Sleep Medicine
Background:
- Changes in consciousness correlate with brain connectivity patterns observed in resting-state fMRI (rs-fMRI).
- Simultaneous electroencephalography (EEG) is not always available, necessitating methods to decode sleep patterns directly from rs-fMRI data.
- Accurate interpretation of rs-fMRI data can be improved by accounting for confounding sleep states.
Purpose of the Study:
- To develop and validate a method for classifying sleep stages using rs-fMRI data alone.
- To assess the feasibility of decoding non-rapid eye movement (NREM) sleep stages from rs-fMRI connectivity patterns.
- To identify brain connectivity features that differentiate wakefulness from NREM sleep stages.
Main Methods:
- Linear support vector machine classifiers were trained on combined rs-fMRI and EEG data from 25 subjects.
- Classifiers were trained to distinguish wakefulness (S0) from NREM sleep stages (S1, S2, slow wave sleep [SW], and all combined [SX]).
- Performance was evaluated using leave-one-subject-out cross-validation and an independent dataset of 19 subjects.
Main Results:
- Excellent classification performance (AUCs near 1.0) was achieved for discriminating sleep from wakefulness (S0|SX, S0|S1, S0|S2, S0|SW).
- Good to excellent performance was observed for classifying between sleep stages (S1|S2:~0.9; S1|SW:~1.0; S2|SW:~0.8).
- Reliable classification required fMRI data windows of at least 70 seconds, with subcortical-cortical and within-occipital lobe connectivity patterns being key discriminators.
Conclusions:
- Functional connectivity analysis of rs-fMRI data enables valid classification of NREM sleep stages.
- This approach provides a valuable tool for interpreting rs-fMRI data, especially when EEG is unavailable.
- Brain connectivity alterations during sleep offer significant insights into the neurophysiological changes associated with altered states of consciousness.
Related Concept Videos
01:22Stages of Sleep
1.7K
Sleep progresses through distinct stages, each characterized by specific brain wave patterns and physiological responses ranging from wakefulness to stages of non-rapid eye movement, known as non-REM, to rapid eye movement, referred to as REM. Understanding these stages helps in recognizing how sleep supports various bodily and cognitive functions.
Before sleep begins, in wakefulness, the brain exhibits primarily beta waves, which are high in frequency and low in amplitude, indicating alertness...
Before sleep begins, in wakefulness, the brain exhibits primarily beta waves, which are high in frequency and low in amplitude, indicating alertness...
1.7K
01:15REM Sleep Behavior Disorder
2.5K
REM Sleep Behavior Disorder (RBD) is a sleep disorder characterized by the absence of muscle paralysis that normally occurs during the REM phase of sleep. This absence allows individuals to physically act out their dreams, which are often vivid and disturbing. Common behaviors exhibited during episodes include kicking, punching, and yelling. These actions can be dangerous, potentially leading to injuries for the person with RBD or their bed partner.
RBD is significantly associated with...
RBD is significantly associated with...
2.5K