From the ACS Press Room
‘E-nose’ could Someday Diagnose Parkinson’s Disease by ‘Smelling’ Skin (GC)-mass spectrometry to analyze odor compounds in the sebum of people with PD. But the instruments are bulky, slow and exACS Omega pensive. Jun Liu, Xing Chen and colleagues wanted to develop a fast, easy to use, portaA couple of years ago, a woman named Joy ble and inexpensive GC system to diagnose Milne made headlines when scientists discov- PD through smell, making it suitable for ered that she could “smell” Parkinson’s dispoint-of-care testing. ease (PD) on people with the neurodegenerative disorder. Since then, researchers have The researchers developed an e-nose, combeen trying to build devices that could diag- bining GC with a surface acoustic wave sennose PD through odor compounds on the sor — which measures gaseous compounds skin. Now, researchers reporting in ACS through their interaction with a sound wave Omega have developed a portable, artificially — and machine learning algorithms. The intelligent olfactory system, or “e-nose,” that team collected sebum samples from 31 PD could someday diagnose the disease in a doc- patients and 32 healthy controls by swabbing their upper backs with gauze. They analyzed tor’s office. volatile organic compounds emanating from PD causes motor symptoms, such as tremors, the gauze with the e-nose, finding three odor rigidity and trouble walking, as well as non- compounds (octanal, hexyl acetate and perilmotor symptoms, including depression and lic aldehyde) that were significantly different dementia. Although there’s no cure, early di- between the two groups, which they used to agnosis and treatment can improve one’s build a model for PD diagnosis. quality of life, relieve symptoms and prolong survival. However, the disease usually isn’t Next, the researchers analyzed sebum from identified until patients develop motor symp- an additional 12 PD patients and 12 healthy toms, and by that time, they’ve already expe- controls, finding that the model had an accurienced irreversible neuron loss. Recently, racy of 70.8% in predicting PD. The model scientists discovered that people with PD se- was 91.7% sensitive in identifying true PD crete increased sebum (an oily, waxy sub- patients, but its specificity was only 50%, instance produced by the skin’s sebaceous dicating a high rate of false positives. When glands), along with increased production of machine learning algorithms were used to anyeast, enzymes and hormones, which com- alyze the entire odor profile, the accuracy of bine to produce certain odors. Although hu- diagnosis improved to 79.2%. Before the eman “super smellers” like Milne are very ra- nose is ready for the clinic, the team needs to re, researchers have used gas chromatography test it on many more people to improve the “Artificial Intelligent Olfactory System for the Diagnosis of Parkinson’s Disease”
March 2022
The Southwest RETORT
20
