Impact of different respiratory monitoring techniques on respiration-dependent stroke-volume measurements assessed by real-time magnetic resonance imaging

Abstract Purpose Quantitative blood flow measurements in thoracic vessels are dependent on the patient's respiration due to intrathoracic pressure changes. Several registration techniques for tracing the patient's breathing curve exist exploiting various physiological characteristics. In the presented study two registration techniques were investigated to estimate the time-shift between the associated respiration curves and the impact on respiration-dependent hemodynamic measures. Materials and methods Using flow-sensitive real-time magnetic resonance imaging (3.0 T, temporal resolution = 24–26 ms) data were acquired during a 13 s time-interval under normal physiological (no) and forced breathing (fo) in the ascending aorta (AAo) and inferior vena cava (IVC). Breathing curves were obtained by using (1) an abdominal placed respiratory belt (=standard, RB) and (2) by applying a self-developed edge-detection software (ED). Respiration curves were divided into four intervals (end-expiration, inspiration, end-inspiration and expiration) to generate respiration-dependent stroke volumes (SVs) and cardiac indices (CIs). Data were available from 12 healthy controls (16.2 ± 8.8 yrs) and 18 Fontan-patients (18.6 ± 7.1 yrs). Results Respiration curves acquired with RB differs from and are shifted compared to those obtained by ED (controls: average shift = 207 ± 168 ms, Fontan-patients: average shift = 106 ± 235 ms). This time-shift results in statistical significant differences in cardiac indices CI AAo,fo in controls (ΔCI AAo,fo : expiration: +0.16 L/min/m 2 , p  = 0.018), in CI IVC,no in Fontan-patients (ΔCI IVC,no : end-expiration: +0.94 L/min/m 2 , p  = 0.002 and end-inspiration: −1.04 L/min/m 2 , p  = 0.017) and in CI IVC,fo in Fontan-patients (ΔCI IVC,fo : end-expiration: +1.31 L/min/m 2 , p  = 0.009; inspiration: −2.26 L/min/m 2 , p  = 0.008 and end-inspiration: −1.87 L/min/m 2 , p  = 0.029). Conclusions A time-shift between both applied respiratory tracking techniques was observed resulting in significant differences in respiration-dependent CIs, which could influence the clinical decision-making.