Dataset Information

A machine learning model for multi-night actigraphic detection of chronic insomnia: development and validation of a pre-screening tool.

ABSTRACT: We propose a novel machine learning-based method for analysing multi-night actigraphy signals to objectively classify and differentiate nocturnal awakenings in individuals with chronic insomnia (CI) and their cohabiting healthy partners. We analysed nocturnal actigraphy signals from 40 cohabiting couples with one partner seeking treatment for insomnia. We extracted 12 time-domain dynamic and nonlinear features from the actigraphy signals to classify nocturnal awakenings in healthy individuals and those with CI. These features were then used to train two machine learning classifiers, random forest (RF) and support vector machine (SVM). An optimization algorithm that incorporated the predicted quality of each night for each individual was used to classify individuals into CI or healthy sleepers. Using the proposed actigraphic signal analysis technique, coupled with a rigorous leave-one-out validation approach, we achieved a classification accuracy of 80% (sensitivity: 76%, specificity: 82%) in classifying CI individuals and their healthy bed partners. The RF classifier (accuracy: 80%) showed a better performance than SVM (accuracy: 75%). Our approach to analysing the multi-night nocturnal actigraphy recordings provides a new method for screening individuals with CI, using wrist-actigraphy devices, facilitating home monitoring.

SUBMITTER: Kusmakar S

PROVIDER: S-EPMC8206690 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:Study objectivesWe evaluated the performance of the Insomnia Severity Index-3 (ISI-3) as a short screening tool to identify clinically significant insomnia derived from the 7-item ISI in an older primary care population.MethodsWe used results from two surveys including the 7-item ISI: Sample 1 (n = 3197) and Sample 2 (n = 247) individuals aged ≥60 years with a diagnosis of osteoarthritis from electronic health records. The 7 items were: difficulty falling asleep, difficulty staying asleep, waking too early, sleep satisfaction, sleep interference with daytime functioning, noticeability of sleep problems by others, and worry about sleep. The ISI-3 included items with highest item-total correlations to the 7-item ISI from Sample 1. A 7-item ISI score ≥15 was defined as clinically significant insomnia and served as the primary criterion for the ISI-3. We derived operating characteristics to determine the diagnostic accuracy and cut-points to maximize sensitivity and specificity for both samples.ResultsThe items with the highest item-total correlations were: sleep dissatisfaction, sleep interference with daily functioning, and worry about sleep problems (r = 0.78-0.81); while difficulty falling asleep, difficulty staying asleep, waking too early and noticeability of sleep problems by others showed lower correlations (r = 0.60-0.74). The ISI-3 achieved high discriminant validity in identifying insomnia (AUC = 0.97-0.98). An ISI-3 score of ≥7 maximized sensitivity (0.94-0.97) and specificity (0.88-0.91) with kappa = 0.68-0.71, 89.1-91.5% agreement.ConclusionsThe ISI-3 can effectively screen for insomnia to trigger a more thorough diagnostic evaluation including the 7-item ISI for research or clinical purposes. Future validation studies are needed in other community and clinical populations.Clinical trialThis manuscript describes secondary analyses of data two National Institutes on Aging-funded clinical trials (ClinicalTrials.gov identifier: NCT01142349, NCT02946957).

Dataset Information

A machine learning model for multi-night actigraphic detection of chronic insomnia: development and validation of a pre-screening tool.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets