Dynamic accuracy of lung simulation hardware: A comparative evaluation

Simulation of human respiratory mechanics is critical for the evaluation, certification, and the clinical reliability of ventilatory and sleep therapy devices. Despite the widespread use of lung simulation devices, their dynamic accuracy remains underexplored. This study presents the first systematic comparison of lung simulators: ASL5000 (IngMar Medical, Pittsburgh, USA), Alveo (IMT Analytics, Buchs, Switzerland), SmartLung 2000 (IMT Analytics, Buchs, Switzerland) and Training & Test Lung (TTL, Michigan Instruments, Kentwood, USA). Using sinusoidal flow stimuli and frequency-domain analysis, we evaluated each simulator' ability to replicate the theoretical behavior of the linear single-compartment lung model, serving as a reference. Our results in healthy, restrictive, and obstructive lung model configurations indicate that the ASL5000, while accurately simulating in the typical breathing frequency range, exhibits resonance and attenuation along with significant phase error at higher frequencies. Alveo exhibits minimal frequency and flow amplitude dependence, demonstrating accurate simulation capabilities. However, pre-launch testing revealed a resistance bias, particularly when using the obstructive parameter set. Passive simulators such as SmartLung 2000 and TTL show consistent behavior across the frequency spectrum but exhibit flow amplitude dependence due to their parabolic resistance. Based on these findings, we emphasize the importance of considering simulator-specific characteristics and limitations for designing bench tests and interpreting experimental results. To our knowledge, this is the first study to systematically evaluate the frequency response and simulation accuracy of multiple lung simulation devices.