Algorithms, modelling and kinetics

This article summarises the pros and cons of different algorithms developed for estimating breath-by-breath (B-by-B) alveolar O 2 transfer ( ) in humans. is the difference between O 2 uptake at the mouth and changes in alveolar O 2 stores (∆ V O 2s ), which for any given breath, are equal to the alveolar volume change at constant plus the O 2 alveolar fraction change at constant volume , where V A i −1 is the alveolar volume at the beginning of a breath. Therefore, can be determined B-by-B provided that V A i −1 is: (a) set equal to the subject’s functional residual capacity (algorithm of Auchincloss, A) or to zero; (b) measured (optoelectronic plethysmography, OEP); (c) selected according to a procedure that minimises B-by-B variability (algorithm of Busso and Robbins, BR). Alternatively, the respiratory cycle can be redefined as the time between equal F O 2 in two subsequent breaths (algorithm of Grønlund, G), making any assumption of V A i −1 unnecessary. All the above methods allow an unbiased estimate of at steady state, albeit with different precision. Yet the algorithms “per se” affect the parameters describing the B-by-B kinetics during exercise transitions. Among these approaches, BR and G, by increasing the signal-to-noise ratio of the measurements, reduce the number of exercise repetitions necessary to study kinetics, compared to A approach. OEP and G (though technically challenging and conceptually still debated), thanks to their ability to track ∆ V O 2s changes during the early phase of exercise transitions, appear rather promising for investigating B-by-B gas exchange.