The use of time‐of‐flight camera to assess respiratory rates and thoracoabdominal depths in patients with chronic respiratory disease. (29th January 2023)

• Record Type: Journal Article
• Authors: Van Hove, Olivier; Andrianopoulos, Vasileios; Dabach, Ali; Debeir, Olivier; Van Muylem, Alain; Leduc, Dimitri; Legrand, Alexandre; Ercek, Rudy; Feipel, Véronique; Bonnechère, Bruno
• Is Part Of: Clinical respiratory journal. Volume 17:Number 3(2023)
• Page Start: 176
• Page End: 186
• Publication Date: 2023-01-29
• Subjects: assessment -- breathing -- Kinect sensor -- respiratory diseases -- validation; Respiratory organs -- Diseases -- Periodicals; Respiratory organs -- Periodicals; 616.24
• DOI: 10.1111/crj.13581
• Languages: English
• ISSNs: 1752-6981

Abstract

Abstract: Introduction: Over the last 5 years, the analysis of respiratory patterns presents a growing usage in clinical and research purposes, but there is still currently a lack of easy‐to‐use and affordable devices to perform such kind of evaluation. Objectives: The aim of this study is to validate a new specifically developed method, based on Kinect sensor, to assess respiratory patterns against spirometry under various conditions. Methods: One hundred and one participants took parts in one of the three validations studies. Twenty‐five chronic respiratory disease patients (14 with chronic obstructive pulmonary disease (COPD) [65 ± 10 years old, FEV1 = 37 (15% predicted value), VC = 62 (20% predicted value)], and 11 with lung fibrosis (LF) [64 ± 14 years old, FEV1 = 55 (19% predicted value), VC = 62 (20% predicted value)]) and 76 healthy controls (HC) were recruited. The correlations between the signal of the Kinect (depth and respiratory rate) and the spirometer (tidal volume and respiratory rate) were computed in part 1. We then included 66 HC to test the ability of the system to detect modifications of respiratory patterns induced by various conditions known to modify respiratory pattern (cognitive load, inspiratory load and combination) in parts 2 and 3. Results: There is a strong correlation between the depth recorded by the Kinect and the tidal volume recorded by the spirometer: r = 0.973 for COPD patients, r = 0.989 for LF patients and r = 0.984 for HC. The Kinect is able to detect changes in breathing patterns induced by different respiratory disturbance conditions, gender and oral task. Conclusions: Measurements performed with the Kinect sensors are highly correlated with the spirometer in HC and patients with COPD and LF. Kinect is also able to assess respiratory patterns under various loads and disturbances. This method is affordable, easy to use, fully automated and could be used in the current clinical context. Respiratory patterns are important to assess in daily clinics. However, there is currently no affordable and easy‐to‐use tool to evaluate these parameters in clinics. We validated a new system to assess respiratory patterns using the Kinect sensor in patients with chronic respiratory diseases. Abstract : The non‐contact respiratory pattern measurement system presented in this paper is easy to use, non‐invasive, and doesn't need to be calibrated, and can therefore be easily used in the daily clinic. This type of technique allows the measurement of changes in the thoracoabdominal depth induced by different types of breathing and/of pathologies. This system should be used in the future to assess the patients quickly and on a regular basis. It allows the detection of variations in respiratory patterns and its asynchrony over time and in relation to the constraints on the respiratory system.