Objective: Our aim was to evaluate the feasibility of our developed intelligent cardiopulmonary training system (ICTS) and of the percentage of time spent within the target HR range (%time) as an indicator of adherence to training intensity. Methods: In this noncontrolled trial, nine participants with sedentary lifestyles were recruited from the outpatient rehabilitation department of a teaching hospital. All participants received twelve 30 min sessions of cycling ergometer exercises (5 min warm up, 20 min training phase, and 5 min cool down) with the ICTS three times per week. Training intensity was determined at 60−80% heart rate reserve using cardiopulmonary exercise (CPET) pretests. During training, pedaling resistance was automatically adjusted by the ICTS to keep the user’s heart rate at the predetermined intensity range. Workload-peak and peak oxygen uptake (VO2-peak) were measured during the pretests and post-tests. We recorded the percentage of time spent within the target heart rate range (%time) during the 20 min training phase for each training session as an indicator of adherence. The correlation between %time and gains in VO2-peak was assessed. Results: After 4 weeks of training on the ICTS, workload-peak and VO2-peak significantly improved by 13.6 ± 7.2 w (mean ± SD, p = 0.008) and 1.5 ± 1.1 mL/kg/min (p = 0.011), respectively. The 12-session average %time ranged from 10.6% to 93.1% among the participants, and five participants achieved an average %time >80%. A positive correlation between average %time and training efficacy was found (rs = 0.85, p = 0.004). Conclusions: Cardiopulmonary training with an ICTS is feasible, and the percentage of time spent within the target heart rate range seems to be a reasonable indicator for monitoring training-intensity adherence.
Keywords: aerobic exercise; artificial intelligence; cardiorespiratory fitness; physical fitness; rehabilitation.