In an initial study, Canary analyzed speech recordings with AI and identified early stage Alzheimer’s with 96 percent accuracy.
Dave Arnold retired in his 60s and began spending time volunteering in local schools. But then he started misplacing items, forgetting appointments and losing his sense of direction. Eventually he was diagnosed with early stage Alzheimer’s.
“Hearing the diagnosis made me very emotional and tearful,” he said. “I immediately thought of all my mom had experienced.” His mother suffered with the condition for years before passing away. Over the last year, Arnold has worked for the Alzheimer’s Association as one of its early stage advisors, sharing his insights to help others in the initial stages of the disease.
Arnold was diagnosed sooner than many others. It's important to find out early, when interventions can make the most difference. One promising avenue is looking at how people talk. Research has shown that Alzheimer’s affects a part of the brain that controls speech, resulting in small changes before people show other signs of the disease.
Now, Canary Speech, a company based in Utah, is using AI to examine elements like the pitch of a person’s voice and their pauses. In an initial study, Canary analyzed speech recordings with AI and identified early stage Alzheimer’s with 96 percent accuracy.
Developing the AI model
Canary Speech’s CEO, Henry O’Connell, met cofounder Jeff Adams about 40 years before they started the company. Back when they first crossed paths, they were both living in Bethesda, Maryland; O’Connell was a research fellow at the National Institutes of Health studying rare neurological diseases, while Adams was working to decode spy messages. Later on, Adams would specialize in building mathematical models to analyze speech and sound as a team leader in developing Amazon's Alexa.
It wasn't until 2015 that they decided to make use of the fit between their backgrounds. ““We established Canary Speech in 2017 to build a product that could be used in multiple languages in clinical environments,” O'Connell says.
The need is growing. About 55 million people worldwide currently live with Alzheimer’s, a number that is expected to double by 2050. Some scientists think the disease results from a buildup of plaque in the brain. It causes mild memory loss at first and, over time, this issue get worse while other symptoms, such as disorientation and hallucinations, can develop. Treatment to manage the disease is more effective in the earlier stages, but detection is difficult since mild symptoms are often attributed to the normal aging process.
O’Connell and Adams specialize in the complex ways that Alzheimer’s effects how people speak. Using AI, their mathematical model analyzes 15 million data points every minute, focusing on certain features of speech such as pitch, pauses and elongation of words. It also pays attention to how the vibrations of vocal cords change in different stages of the disease.
To create their model, the team used a type of machine learning called deep neural nets, which looks at multiple layers of data - in this case, the multiple features of a person’s speech patterns.
“Deep neural nets allow us to look at much, much larger data sets built out of millions of elements,” O’Connell explained. “Through machine learning and AI, we’ve identified features that are very sensitive to an Alzheimer’s patient versus [people without the disease] and also very sensitive to mild cognitive impairment, early stage and moderate Alzheimer's.” Based on their learnings, Canary is able to classify the disease stage very quickly, O’Connell said.
“When we’re listening to sublanguage elements, we’re really analyzing the direct result of changes in the brain in the physical body,” O’Connell said. “The brain controls your vocal cords: how fast they vibrate, the expansion of them, the contraction.” These factors, along with where people put their tongues when talking, function subconsciously and result in subtle changes in the sounds of speech.
Further testing is needed
In an initial trial, Canary analyzed speech recordings from phone calls to a large U.S. health insurer. They looked at the audio recordings of 651 policyholders who had early stage Alzheimer’s and 1018 who did not have the condition, aiming for a representative sample of age, gender and race. They used this data to create their first diagnostic model and found that it was 96 percent accurate in identifying Alzheimer’s.
Christian Herff, an assistant professor of neuroscience at Maastricht University in the Netherlands, praised this approach while adding that further testing is needed to assess its effectiveness.
“I think the general idea of identifying increased risk for cognitive impairment based on speech characteristics is very feasible, particularly when change in a user’s voice is monitored, for example, by recording speech every year,” Herff said. He noted that this can only be a first indication, not a full diagnosis. The accuracy still needs to be validated in studies that follows individuals over a period of time, he said.
Toby Walsh, a professor of artificial intelligence at the University of New South Wales, also thinks Canary’s tool has potential but highlights that Canary could diagnose some people who don’t really have the disease. “This is an interesting and promising application of AI,” he said, “but these tools need to be used carefully. Imagine the anxiety of being misdiagnosed with Alzheimer’s.”
As with many other AI tools, privacy and bias are additional issues to monitor closely, Walsh said.
Other languages
A related issue is that not everyone is fluent in English. Mahnaz Arvaneh, a senior lecturer in automatic control and systems engineering at the University of Sheffield, said this could be a blind spot.
“The system may not be very accurate for those who have English as their second language as their speaking patterns would be different, and any issue might be because of language deficiency rather than cognitive issues,” Arvaneh said.
The team is expanding to multiple languages starting with Japanese and Spanish. The elements of the model that make up the algorithm are very similar, but they need to be validated and retrained in a different language, which will require access to more data.
Recently, Canary analyzed the phone calls of 233 Japanese patients who had mild cognitive impairment and 704 healthy people. Using an English model they were able to identify the Japanese patients who had mild cognitive impairment with 78 percent accuracy. They also developed a model in Japanese that was 45 percent accurate, and they’re continuing to train it with more data.
The future
Canary is using their model to look at other diseases like Huntington’s and Parkinson’s. They’re also collaborating with pharmaceuticals to validate potential therapies for Alzheimer’s. By looking at speech patterns over time, Canary can get an indication of how well these drugs are working.
Dave Arnold and his wife dance at his nephew’s wedding in Rochester, New York, ten years ago, before his Alzheimer's diagnosis.
Dave Arnold
Ultimately, they want to integrate their tool into everyday life. “We want it to be used in a smartphone, or a teleconference call so that individuals could be examined in their home,” O’Connell said. “We could follow them over time and work with clinical teams and hospitals to improve the evaluation of patients and contribute towards an accurate diagnosis.”
Arnold, the patient with early stage Alzheimer’s, sees great promise. “The process of getting a diagnosis is already filled with so much anxiety,” he said. “Anything that can be done to make it easier and less stressful would be a good thing, as long as it’s proven accurate.”
Indigenous people in Australia dig pits next to a honeypot colony. Scientists think the honey can be used to make new antimicrobial drugs.
These hunts have become rarer, as many of the Tjupan people have moved away and, up until now, the exact antimicrobial properties of the ant honey remained unknown. But recently, scientists Andrew Dong and Kenya Fernandes from the University of Sydney, joined Ulrich, who runs the Honeypot Ants tours in Kalgoorlie, a city in Western Australia, on a honey-gathering expedition. Afterwards, they ran a series of experiments analyzing the honey’s antimicrobial activity—and confirmed that the Indigenous wisdom was true. The honey was effective against Staphylococcus aureus, a common pathogen responsible for sore throats, skin infections like boils and sores, and also sepsis, which can result in death. Moreover, the honey also worked against two species of fungi, Cryptococcus and Aspergillus, which can be pathogenic to humans, especially those with suppressed immune systems.
In the era of growing antibiotic resistance and the rising threat of pathogenic fungi, these findings may help scientists identify and make new antimicrobial compounds. “Natural products have been honed over thousands and millions of years by nature and evolution,” says Fernandes. “And some of them have complex and intricate properties that make them really important as potential new antibiotics. “
In an era of growing resistance to antibiotics and new threats of fungi infections, the latest findings about honeypot ants are helping scientists identify new antimicrobial drugs.
Danny Ulrich
Bee honey is also known for its antimicrobial properties, but bees produce it very differently than the ants. Bees collect nectar from flowers, which they regurgitate at the hive and pack into the hexagonal honeycombs they build for storage. As they do so, they also add into the mix an enzyme called glucose oxidase produced by their glands. The enzyme converts atmospheric oxygen into hydrogen peroxide, a reactive molecule that destroys bacteria and acts as a natural preservative. After the bees pack the honey into the honeycombs, they fan it with their wings to evaporate the water. Once a honeycomb is full, the bees put a beeswax cover on it, where it stays well-preserved thanks to the enzymatic action, until the bees need it.
Less is known about the chemistry of ants’ honey-making. Similarly to bees, they collect nectar. They also collect the sweet sap of the mulga tree. Additionally, they also “milk” the aphids—small sap-sucking insects that live on the tree. When ants tickle the aphids with their antennae, the latter release a sweet substance, which the former also transfer to their colonies. That’s where the honey management difference becomes really pronounced. The ants don’t build any kind of structures to store their honey. Instead, they store it in themselves.
The workers feed their harvest to their fellow ants called repletes, stuffing them up to the point that their swollen bellies outgrow the ants themselves, looking like amber-colored honeypots—hence the name. Because of their size, repletes don’t move, but hang down from the chamber’s ceiling, acting as living feedstocks. When food becomes scarce, they regurgitate their reserves to their colony’s brethren. It’s not clear whether the repletes die afterwards or can be restuffed again. “That's a good question,” Dong says. “After they've been stretched, they can't really return to exactly the same shape.”
These replete ants are the “treat” the Tjupan women dug for. Once they saw the round-belly ants inside the chambers, they would reach in carefully and get a few scoops of them. “You see a lot of honeypot ants just hanging on the roof of the little openings,” says Ulrich’s mother, Edie Ulrich. The women would share the ants with family members who would eat them one by one. “They're very delicate,” shares Edie Ulrich—you have to take them out carefully, so they don’t accidentally pop and become a wasted resource. “Because you’d lose all this precious honey.”
Dong stumbled upon the honeypot ants phenomenon because he was interested in Indigenous foods and went on Ulrich’s tour. He quickly became fascinated with the insects and their role in the Indigenous culture. “The honeypot ants are culturally revered by the Indigenous people,” he says. Eventually he decided to test out the honey’s medicinal qualities.
The researchers were surprised to see that even the smallest, eight percent concentration of honey was able to arrest the growth of S. aureus.
To do this, the two scientists first diluted the ant honey with water. “We used something called doubling dilutions, which means that we made 32 percent dilutions, and then we halve that to 16 percent and then we half that to eight percent,” explains Fernandes. The goal was to obtain as much results as possible with the meager honey they had. “We had very, very little of the honeypot ant honey so we wanted to maximize the spectrum of results we can get without wasting too much of the sample.”
After that, the researchers grew different microbes inside a nutrient rich broth. They added the broth to the different honey dilutions and incubated the mixes for a day or two at the temperature favorable to the germs’ growth. If the resulting solution turned turbid, it was a sign that the bugs proliferated. If it stayed clear, it meant that the honey destroyed them. The researchers were surprised to see that even the smallest, eight percent concentration of honey was able to arrest the growth of S. aureus. “It was really quite amazing,” Fernandes says. “Eight milliliters of honey in 92 milliliters of water is a really tiny amount of honey compared to the amount of water.”
Similar to bee honey, the ants’ honey exhibited some peroxide antimicrobial activity, researchers found, but given how little peroxide was in the solution, they think the honey also kills germs by a different mechanism. “When we measured, we found that [the solution] did have some hydrogen peroxide, but it didn't have as much of it as we would expect based on how active it was,” Fernandes says. “Whether this hydrogen peroxide also comes from glucose oxidase or whether it's produced by another source, we don't really know,” she adds. The research team does have some hypotheses about the identity of this other germ-killing agent. “We think it is most likely some kind of antimicrobial peptide that is actually coming from the ant itself.”
The honey also has a very strong activity against the two types of fungi, Cryptococcus and Aspergillus. Both fungi are associated with trees and decaying leaves, as well as in the soils where ants live, so the insects likely have evolved some natural defense compounds, which end up inside the honey.
It wouldn’t be the first time when modern medicines take their origin from the natural world or from the indigenous people’s knowledge. The bark of the cinchona tree native to South America contains quinine, a substance that treats malaria. The Indigenous people of the Andes used the bark to quell fever and chills for generations, and when Europeans began to fall ill with malaria in the Amazon rainforest, they learned to use that medicine from the Andean people.
The wonder drug aspirin similarly takes its origin from a bark of a tree—in this case a willow.
Even some anticancer compounds originated from nature. A chemotherapy drug called Paclitaxel, was originally extracted from the Pacific yew trees, Taxus brevifolia. The samples of the Pacific yew bark were first collected in 1962 by researchers from the United States Department of Agriculture who were looking for natural compounds that might have anti-tumor activity. In December 1992, the FDA approved Paclitaxel (brand name Taxol) for the treatment of ovarian cancer and two years later for breast cancer.
In the era when the world is struggling to find new medicines fast enough to subvert a fungal or bacterial pandemic, these discoveries can pave the way to new therapeutics. “I think it's really important to listen to indigenous cultures and to take their knowledge because they have been using these sources for a really, really long time,” Fernandes says. Now we know it works, so science can elucidate the molecular mechanisms behind it, she adds. “And maybe it can even provide a lead for us to develop some kind of new treatments in the future.”
