Synchronized oxygen delivery and its optimization method: A bench study

Abstract

Long-term oxygen therapy (LTOT) has been widely used to treat patients with chronic obstructive pulmonary disease. The traditional oxygen delivery methods used for LTOT are continuous flow oxygen (CFO) and demand oxygen delivery (DOD). CFO wastes a considerable amount of oxygen, whereas DOD often makes patients feel uncomfortable because it abruptly supplies a large amount of oxygen at the onset of inhalation. Hence, we developed an algorithm for predicting the onset of inhalation, which allowed oxygen to be supplied smoothly before inhalation. Moreover, we minimized the discomfort index (DI) to offer more comfortable oxygen delivery. By integrating the prediction algorithm and the minimization of DI, the previous synchronized demand oxygen delivery (SDOD) method was modified. We constructed a bench model to validate the modified SDOD. The results showed that the proposed algorithm accurately predicted the onset of inhalation. The difference in the real-time measured and predicted values for the beginning of inhalation was less than 0.10 s. Using the proposed minimization technique, the DI was decreased by 50% under 20 breaths per minute when compared with the DI calculated from a previous study. In conclusion, the modified SDOD could supply oxygen more comfortably while synchronizing with patient breathing patterns.