Abstract: Objective: The objective is to find the characteristics of arterial blood sample waveform in different respiration models. Methods: Six post-operative patients with normal heart function and negative Allen test, were 4 male and 2 female, (59.00±16.64)year, (71.67±0.37)kg, left ventricular ejection fraction(LVEF) (61.33±2.16)%, had been placed the arterial catheterization and central venous catheterization for continuous collecting arterial in 3 different kinds of respiration models: normal breathing, no breathing and deep breathing. We selected two breaths cycles of waveform from each patient for data calculations of magnitudes and time interval. Compare the adjacent highest and lowest values of patients to verify whether there are periodic wave-like signal changes in arterial and venous blood gas in the three breathing states. In addition, statistical t-test analysis was performed on the change amplitude of the periodic wave-like signal of the patient's arterial and venous blood gas to compare whether there is a difference. Results: The heart beat numbers for drawing blood into pipe were 15-16, and all covered more than 2 breathing cycles. There were significant changes of arterial PaO₂ (i.e. the highest high values compare to the next lowest values, P<0.05) in three different breathing models(normal, no breathing and high breathing), the magnitudes of which were (9.96±5.18)mmHg, (5.33±1.55)mmHg and (13.13±7.55)mmHg, with (8.09±2.43)%, (5.29±2.19)% and (10.40±2.68)% from their mean respectively. PO₂ in venous blood gas did not show wavy changes under normal breathing, 20 s breath holding and high tidal volume ventilation. The amplitudes were (1.63 ± 0.41) mmHg, (1.13 ± 0.41) mmHg and (1.31 ± 0.67) mmHg, which were (3.91 ± 1.22)%, (2.92 ± 1.12)%, (3.33 ± 1.81)%, respectively, which were significantly lower than that of arterial blood gas under the same state, but there was no significant difference between groups. Conclusion: With continuous beat-by-beat arterial blood sampling and ABG analyzing method in three different breathing models, We obtain a clear evidence of the biggest periodic parameters ABG waveform in high breathing models, which followed by normal breathing models, no breathing was the smallest, and the wave variation amplitude of venous oxygen partial pressure was not obvious in the three respiratory states, which implies the oscillatory information of the arterial blood with comes from the gas exchanging in the lung.

Key words: irregular breathing pattern, beat-by-beat arterial blood, waveform breathing, respiration regulation