An algorithm for actigraphy-based sleep/wake scoring: Comparison with polysomnography.

Area of Science:

• Sleep science
• Biomedical engineering
• Neuroscience

Background:

• Sleep disturbances are common in neuropsychiatric diseases like depression and dementia.
• Accurate sleep/wake cycle assessment is crucial for managing these conditions.
• Actigraphy offers an unobtrusive alternative to polysomnography (PSG) for sleep monitoring.

Purpose of the Study:

• To evaluate a novel algorithm for sleep/wake discrimination using wrist actigraphy.
• To compare the accuracy of this new algorithm against existing state-of-the-art methods.
• To validate actigraphy's accuracy against polysomnography (PSG) as the gold standard.

Main Methods:

• Polysomnography (PSG) and wrist actigraphy were simultaneously recorded in 20 participants (11 men, 9 women, age 71.1±5.0) with suspected neuropsychiatric sleep disturbances.
• 37 features were extracted from actigraphy data for each 1-min epoch.
• A newly developed algorithm modeling temporal sleep structure was compared against four state-of-the-art algorithms using leave-one-subject-out cross-validation.

Main Results:

• The new algorithm achieved 84.9% sensitivity (sleep) and 74.2% specificity (wake), with an overall accuracy of 79.0%.
• It demonstrated significantly higher specificity compared to existing algorithms, with only a minor decrease in sensitivity.
• Four out of five sleep parameters were estimated more accurately by the new algorithm.

Conclusions:

• The novel algorithm shows improved accuracy, particularly specificity, in distinguishing sleep and wake states from actigraphy data.
• Its ability to model sleep architecture contributes to its enhanced performance in patients with sleep disorders.
• This algorithm holds promise for unobtrusive sleep assessment in neuropsychiatric populations.