It has been a practice by both scientists and law enforcers to train sniffing dogs to help them with their jobs. These animals have a seen sense of smell, one that’s better than most humans. Many even choose these canines to sniff out cancer.
But there’s also another option. There’s a human with a keen sense of smell. So accurate is she that the scientists have developed their own version of the e-nose. This has been designed to spot Parkinson’s and the inspiration came from a woman named Joy Milne.
Years back, Joy Milne made the news. Headlines of newspapers were dedicated to her when scientists discovered how she could actually sniff out Parkinson’s disease on people who suffer from the neurodegenerative and debilitating disorder. They were impressed at her ability that they wanted to create a technology that could spot what she could.
Since they met her, researchers from all over have been working hard to build devices that could diagnose the patients who may suffer from the disease through odor compounds that seep out of the human skin. Their hard work and diligence may have paid off because now, they have reported in ACS Omega that they have finally developed a portable, artificially intelligent olfactory system that they call “e-nose.” This tool could one day diagnose the disease so long as it’s available in the doctor’s office.
Parkinson’s disease (or what is oftentimes referred to as PD) causes motor symptoms, such as tremors, rigidity, and trouble walking. Non-motor symptoms such as depression and dementia become apparent as well to those who suffer from it.
Although Parkinson’s comes with no cure at the moment, the best way to cope with it is through early diagnosis and treatment. By catching the disease early, doctors are able to improve the quality of life, relieve symptoms, and prolong survival for their patients.
Sadly, they are oftentimes unable to identify the disease until their patients start developing motor symptoms, and by then, irreversible neuron loss has already taken place. Hence, research has gone into this and what scientists recently discovered was that those suffering from PD secrete increased sebum. This is an oily, waxy substance that is produced by the skin’s sebaceous glands. Along with it is an increased production of yeast, enzymes, and hormones as well. When all these are combined, the patient emits a specific odor.
Human “super smellers” don’t happen very often. Milne is one of those rare and talented folks that have been blessed with the gift. She first caught scent of the disease’s “musky, oily odor” when she smelled it on her now-late husband Les. Around 12 years after she first caught smell on him, as GNN reported, he was diagnosed with Parkinson’s by the time he was 45 years old.
What Started It All
Milne shared her discovery and scientists were shocked at the ability she had. They then started their research to look more into this rare talent from the Scottish native.
“Finding changes in the oils of the skin in Parkinson’s is an exciting discovery that was sparked by a simple conversation between a member of the public and a researcher,” said Professor David Dexter. He was the Deputy Director of Research at Parkinson’s UK when this took place. He added, “More research is needed to find out at what stage a skin test could detect Parkinson’s, or whether it is also occurs in other Parkinson’s related disorders, but the results so far hold real potential. Both to change the way we diagnose the condition and it may even help in the development of new and better treatments for the 145,00 people living with Parkinson’s in the UK.”
Milne was immediately asked to join the research. She had to smell several dozen tee-shirts that were used by those with and without the disease. She was able to point out which of the shirts belonged to Parkinson’s patients. She was then asked to smell a shirt worn by a “healthy” participant. In as short as eight months, the said participants had received a positive diagnosis.
The tests showed how Milne was 98 percent accurate in diagnosing the disease that were spread about in 50 different shirt samples. They immediately conducted further research and the experts also discovered that what her nose was smelling was a compound contained in the sebum, which is the waxy, lipid-based biofluid whose main function was to moisturize and protect the skin. This sebum can be found mainly on the forehead and upper back. When there is excessive production of the substance, the smell becomes more apparent and this may also be a symptom of Parkinson’s. Credit should be given to Milne’s nose because scientists were able to identify specific biomarkers found in the sebum that are considered red flags.
The New Technology Developed
In order to come up with the e-nose, the researchers made use of gas chromatography (GC)-mass spectrometry to analyze the odor compounds present in the sebum of those who suffer from PD. However, these instruments are bulky, slow, and expensive. Jun Liu, Xing Chen, and the other colleagues wanted to develop a better and easier model that’s portable and inexpensive. This new GC system will be able to diagnose PD through scent, which makes it ideal for point-of-care testing.
The researchers are now developing an e-nose that combines GC with a surface acoustic wave sensor that’s designed to measure gaseous compounds through the interaction with a sound wave, and the same time via machine learning algorithms. In order to do so, the experts collected sebum samples from 31 PD patients and 32 healthy controls. They did this by swabbing their upper back with the use of a gauze.
After which, they analyzed the vaporous organic compounds that come out from the gauze with the e-nose. They had come upon three odor compounds: octanal, hexyl acetate, and perillic aldehyde. They observed how these were significantly varying from both groups. They used the data collected to build a model for PD diagnosis.
The next step was to analyze the sebum from an additional 12 PD patients and 12 healthy controls. They saw that the model they built had a 70.8 percent accuracy in predicting PD. The model was 91.7 percent sensitive in identifying true PD patients. As of now, the machine’s specificity was only at 50 percent, which meant that it also came with a sufficient rate of false positives. When they used machine learning algorithms to analyze the entire odor profile of the person, the accuracy of diagnosis increased and went up to 79.2 percent.
There are so many steps yet needed to be passed and before the e-nose is made available for clinic use, the team needs to test it out some more on people. Their goal is to ameliorate the accuracy of the models they already have. They are also looking into factors such as race to consider what needs to be improved, the researchers say.
The article has already been published in ACS.
Recommended For You
