PPRS, Colmar, France.
Unité d'exploration des Rythmes Veille Sommeil, Centre Hospitalier de Rouffach, Rouffach, France.
PLoS One. 2023 Oct 20;18(10):e0291593. doi: 10.1371/journal.pone.0291593. eCollection 2023.
Polysomnographic sleep architecture parameters are commonly used to diagnose or evaluate treatment of sleep disorders. Polysomnography (PSG) having practical constraints, the development of wearable devices and algorithms to monitor and stage sleep is rising. Beside pure validation studies, it is necessary for a clinician to ensure that the conclusions drawn with a new generation wearable sleep scoring device are consistent to the ones of gold standard PSG, leading to similar interpretation and diagnosis. This paper reports on the performance of Somno-Art Software for the detection of differences in sleep parameters between patients suffering from obstructive sleep apnea (OSA), insomniac or major depressive disorder (MDD) compared to healthy subjects. On 244 subjects (n = 26 healthy, n = 28 OSA, n = 66 insomniacs, n = 124 MDD), sleep staging was obtained from PSG and Somno-Art analysis on synchronized electrocardiogram and actimetry signals. Mixed model analysis of variance was performed for each sleep parameter. Possible differences in sleep parameters were further assessed with Mann-Whitney U-test between the healthy subjects and each pathology group. All sleep parameters, except N1+N2, showed significant differences between the healthy and the pathology group. No significant differences were observed between Somno-Art Software and PSG, except a 3.6±2.2 min overestimation of REM sleep. No significant interaction 'group'*'technology' was observed, suggesting that the differences in pathologies are independent of the technology used. Overall, comparable differences between healthy subjects and pathology groups were observed when using Somno-Art Software or polysomnography. Somno-Art proposes an interesting valid tool as an aid for diagnosis and treatment follow-up in ambulatory settings.
多导睡眠图睡眠结构参数通常用于诊断或评估睡眠障碍的治疗。由于多导睡眠图（PSG）存在实际限制，因此开发用于监测和分期睡眠的可穿戴设备和算法正在兴起。除了纯粹的验证研究外，临床医生还需要确保使用新一代可穿戴睡眠评分设备得出的结论与金标准 PSG 一致，从而得出相似的解释和诊断。本文报告了 Somno-Art 软件在检测患有阻塞性睡眠呼吸暂停（OSA）、失眠或重度抑郁症（MDD）的患者与健康受试者之间睡眠参数差异方面的性能。在 244 名受试者（n = 26 名健康受试者、n = 28 名 OSA 受试者、n = 66 名失眠受试者、n = 124 名 MDD 受试者）中，从 PSG 和 Somno-Art 分析同步心电图和活动记录仪信号中获得睡眠分期。对每个睡眠参数进行混合模型方差分析。使用 Mann-Whitney U 检验进一步评估健康受试者与每个病理组之间睡眠参数的可能差异。除 N1+N2 外，所有睡眠参数在健康和病理组之间均显示出显著差异。除 REM 睡眠高估 3.6±2.2 分钟外，Somno-Art 软件与 PSG 之间未观察到显著差异。未观察到“组”*“技术”的显著交互作用，表明病理差异独立于所使用的技术。总体而言，当使用 Somno-Art 软件或多导睡眠图时，在健康受试者和病理组之间观察到可比较的差异。Somno-Art 提出了一种有趣的有效工具，可作为在门诊环境中进行诊断和治疗随访的辅助工具。
