Advancements and Challenges in Automatic Sleep Scoring for Rodent Research
Abdelrahman Rayan
Postdoctoral Fellow at Genzel Lab, Donders Institute for Brain, Cognition and Behavior, Radboud University, Nijmegen, Netherlands. Follow on LinkedIn and X (Twitter).
As a postdoctoral researcher in the Genzel Lab, I have interdisciplinary background that spans medicine, neuroscience, and machine learning. My doctoral research in Germany was dedicated to unraveling the neural correlates of experience and learning within the hippocampus. Furthermore, I examined how experiences influence the dynamics of neural substrates of the cognitive map and their role in different navigational strategies. In my current role at Genzel’s Lab, I am engaged in probing the neural mechanisms responsible for memory formation and their subsequent transformation into semantic knowledge, using the object space task as a behavioral paradigm, alongside advanced data-driven analytical methods. Additionally, I am developing innovative automatic sleep scoring tools. These tools are instrumental in dissecting the specific contributions of various sleep stages to memory formation and processing, enhancing our understanding of the cognitive and neural architectures that support learning and memory.
Automated sleep scoring for rodents
Sleep scoring is crucial in both sleep research and clinical practice. Traditionally, it depends on manual scoring by experts, which is time-consuming and suffers from inconsistencies. Human sleep is classified into non-rapid eye movement (NREM) with three stages (N1-N3) and rapid eye movement (REM). In rodent sleep research, it’s typically simplified to REM and NREM, with classification based on muscle tone and theta-delta ratio in EEG recordings, but this overlooks the complexity of sleep architecture in rodents.
While human sleep stages are well-defined, there’s no consensus on a standardized manual for sleep scoring in simpler setups for various species, including rodents and non-human primates. Rodent sleep is usually semiautomatically scored with manual corrections, starting with a threshold for muscle activity to separate wake and sleep periods, and then classifying sleep into NREM and REM based on theta-delta ratios. This simplification ignores the complexity of rodent sleep architecture, including transitional sleep states.
Automatic sleep scoring tools are crucial for more accurate classification of sleep states. They save time compared to manual scoring, produce consistent results with large, well-trained datasets, and can identify new sleep substates. However, they struggle with atypical data and require standard data acquisition protocols. The main challenge is developing a unified system or ensuring different methods are validated against common standards for comparability across studies.
"Automated sleep scoring tools significantly enhance efficiency and consistency in sleep research by overcoming the limitations of manual scoring, yet they require standardization across studies, particularly in rodents, to reliably capture the complex architecture of their sleep states."
Article and infographic based on:
Rayan, Szabo & Genzel (2024). The pros and cons of using automated sleep scoring in sleep research: Comparative analysis of automated sleep scoring in human and rodents: advantages and limitations. Sleep.
Infographic designed by:
Dr. Maria Hrozanova
Postdoctoral fellow at the Department of Public Health and Nursing, Norwegian University of Science and Technology (NTNU), Norway, and a member of the ESRS Digital and Communication Committee.
Recent publications from ESRS members
- Baur et al. (2023). Concentration-effect relationships of plasma caffeine on EEG delta power and cardiac autonomic activity during human sleep. J Sleep Res.
- Pevernagie et al. (2023). Looking for clues in the hypnogram – the human eye and the machine. Sleep.
- Kristoffersen et al. (2023). The long-term effect of work schedule, shift work disorder, insomnia and restless legs syndrome on headache among nurses: A prospective longitudinal cohort study. Cephalalgia.
- Schiza et al. (2023). Co-existence of OSA and respiratory diseases and the influence of gender. Expert Rev Respir Med.
- Minhas et al. (2023). A Novel Approach to Quantify Microsleep in Drivers with Obstructive Sleep Apnea by Concurrent Analysis of EEG Patterns and Driving Attributes. IEEE J Biomed Health Inform.
- Barateau et al. (2023). Narcolepsy Severity Scale-2 and Idiopathic Hypersomnia Severity Scale to better quantify symptoms severity and consequences in Narcolepsy type 2. Sleep.
- Winter et al. (2023). Vagus nerve stimulation for the treatment of narcolepsy. Brain Stimul.