Machine learning for olfactory testing in Parkinson’s disease and REM sleep disorder


  • The 3-Sniffin’-stick model shows potential as a brief screening test to stratify individuals with clinical Parkinson’s disease (PD) and its prodromal form, isolated rapid eye movement (REM) sleep behavior disorder (iRBD), according to olfactory dysfunction.

Why this matters

  • Olfactory dysfunction occurs in up to 90% of individuals with early PD, demonstrating concordance with dopaminergic deficit; thus, olfactory tests including the 16-item Sniffin sticks test are commonly used for screening in PD, but may not always be easily applicable to busy clinical settings.

  • The potential demonstrated by the 3-Sniffin’-stick model in PD stratification alludes to the applicability of machine learning models to neurological disease screening, which may potentially extend to other areas of medicine; however, further research is required in both PD and other disease populations.