Abstract
PURPOSE: To evaluate the proof-of-concept feasibility of using low-cost, commercially available contact microphones (CMs) for bedside voice assessment under simulated hospital noise conditions. METHODS: Two low-cost CMs were tested against a reference accelerometer and headset air microphone using two vocally trained adults. Participants performed sustained vowels, pitch glides, and connected speech under four noise conditions: quiet-lab, quiet-hospital, multi-talker babble, and simulated hospital noise. The selected acoustic parameters, commonly used in clinical assessment, include smoothed cepstral peak prominence (CPPS), fundamental frequency (fo), shimmer, jitter, harmonics-to-noise ratio (HNR), noise-to-harmonics ratio (NHR), and low-to-high spectral ratio (L/H ratio). Data were analyzed using generalized estimating equations. RESULTS: CPPS and fo demonstrated no significant device effects and remained stable across noise conditions (p > 0.05). Breathy voice significantly reduced CPPS (speech: β = − 0.48, p ≤ 0.01; vowel: β = − 0.62, p ≤ 0.01) and increased jitter and shimmer (β = 0.74 and 0.75, respectively; p ≤ 0.01). Device-related variability was most evident in shimmer and NHR, with accelerometer values differing from CMs. Noise conditions minimally influenced primary measures in CMs compared to the headset microphone. CONCLUSION: This feasibility study suggests that low-cost CMs may preserve clinically relevant acoustic measures with stability across noisy conditions. Preliminary findings indicate potential advantages over conventional microphones for bedside voice assessment, though validation with clinical populations in real, rather than staged, conditions is needed. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s00405-025-09970-0.