Human respiratory behaviour is extremely sensitive to instrumentation and observation. The use of a mouthpiece to measure ventilatory flow with a pneumotachograph (PNT) introduces a major perturbation to breathing. Likewise, the emotional and cognitive impacts of being observed are bound to induce breathing modifications. Measuring human breathing under true ecological conditons is currently impossible. This study evaluates the feasibility and reliability of sonar-derived measurements of chest wall displacements (no physical contact with the subject, no visible signal emitted). Such measurements were compared with PNT measurements of ventilatory flow in 21 healthy subjects during tidal breathing. Inspiratory time (TI), expiratory time (TE), and breathing frequency ( f ) were simultaneously measured with a PNT and derived from the mathematical treatment of the thoraco-abdomnimal motion signals obtained from an ultrasound emitting-receiving device projecting non-audiblle acoustic waves through the air on the chest wall. Data were tested for interchangeability using Passing-Bablock regression analysis. All the measurements proved interchangeable, as follows: sonar vs PNT (median and [IQR] in sec): TI 1.93 [1.6-2.29] vs 1.68 [1.43-2.07], TE 2.78 [2.2-3.36] vs 2.48 [2.49-3.67], f 13.09 [10.69-16.03] vs 13.18 [10.98-16.03]. We conclude that a sonar device can be reliably used to assess breathing pattern. In physiological studies, this has the potential to eliminate both the “instrumental” and the “observer” effects and therefore to provide descriptions of breathing under true ecological conditions. Beyond this, our results pave the way to multiple medical and non-medical applications of non-contact respiratory measurements.
