Open Design and Experimental Comparison of Two MRI-Compatible Flow Olfactometer Architectures

Abstract

Olfactory dysfunction is associated with neurodegenerative diseases such as Alzheimer's and Parkinson's, motivating fMRI studies that require precise odor delivery. Two MRI-compatible flow olfactometer architectures are widely used: the classic air-dilution Lorig layout and a continuous-flow vacuum-switch design. However, these systems have not been systematically compared under identical MRI constraints. In this study, both architectures, built from the same modular base hardware, were benchmarked head-to-head for temporal stimulus precision, stimulus shape, achievable odor vapor concentration range, cross-contamination, airflow stability at the participant interface, and MRI compatibility. Air carrier flow rate was varied between 1 and 10 litres per minute, delivery tube length between 1 and 9 meters, and air dilution ratios between 2.5 and 90%. Concentration measurements used undiluted isoamyl acetate and a photo-ionization detector. The vacuum-switch design reduced onset latency by 35 to 60% and kept it below 500 milliseconds with 9 meters of tubing, while producing the flattest square pulses and reducing residual odor to below 1.6% of the primary response. These gains came at the cost of brief flow dips of about 70% below target during valve transitions and a 20 to 35% lower maximum odor vapor concentration. The Lorig system produced higher peaks and steadier flow, but showed longer, tube-length-dependent latency and stronger cross-contamination of about 4%. Neither configuration reduced temporal signal-to-noise ratio in fMRI measurements. These results define quantitative trade-offs that can guide olfactometer selection for future olfactory fMRI studies.