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.”
With conventional fuel cells as their model, researchers learned to use similar chemical reactions to make a fuel from microbes in pee.
Plus, urine contains water, phosphorus, potassium and nitrogen, the key ingredients plants need to grow and survive. Human urine could replace about 25 percent of current nitrogen and phosphorous fertilizers worldwide and could save water for gardens and crops. The average U.S. resident flushes a toilet bowl containing only pee and paper about six to seven times a day, which adds up to about 3,500 gallons of water down per year. Plus cows in the U.S. produce 231 gallons of the stuff each year.
Pee power
A conventional fuel cell uses chemical reactions to produce energy, as electrons move from one electrode to another to power a lightbulb or phone. Ioannis Ieropoulos, a professor and chair of Environmental Engineering at the University of Southampton in England, realized the same type of reaction could be used to make a fuel from microbes in pee.
Bacterial species like Shewanella oneidensis and Pseudomonas aeruginosa can consume carbon and other nutrients in urine and pop out electrons as a result of their digestion. In a microbial fuel cell, one electrode is covered in microbes, immersed in urine and kept away from oxygen. Another electrode is in contact with oxygen. When the microbes feed on nutrients, they produce the electrons that flow through the circuit from one electrod to another to combine with oxygen on the other side. As long as the microbes have fresh pee to chomp on, electrons keep flowing. And after the microbes are done with the pee, it can be used as fertilizer.
These microbes are easily found in wastewater treatment plants, ponds, lakes, rivers or soil. Keeping them alive is the easy part, says Ieropoulos. Once the cells start producing stable power, his group sequences the microbes and keeps using them.
Like many promising technologies, scaling these devices for mass consumption won’t be easy, says Kevin Orner, a civil engineering professor at West Virginia University. But it’s moving in the right direction. Ieropoulos’s device has shrunk from the size of about three packs of cards to a large glue stick. It looks and works much like a AAA battery and produce about the same power. By itself, the device can barely power a light bulb, but when stacked together, they can do much more—just like photovoltaic cells in solar panels. His lab has produced 1760 fuel cells stacked together, and with manufacturing support, there’s no theoretical ceiling, he says.
Although pure urine produces the most power, Ieropoulos’s devices also work with the mixed liquids of the wastewater treatment plants, so they can be retrofit into urban wastewater utilities.
This image shows how the pee-powered system works. Pee feeds bacteria in the stack of fuel cells (1), which give off electrons (2) stored in parallel cylindrical cells (3). These cells are connected to a voltage regulator (4), which smooths out the electrical signal to ensure consistent power to the LED strips lighting the toilet.
Courtesy Ioannis Ieropoulos
Key to the long-term success of any urine reclamation effort, says Orner, is avoiding what he calls “parachute engineering”—when well-meaning scientists solve a problem with novel tech and then abandon it. “The way around that is to have either the need come from the community or to have an organization in a community that is committed to seeing a project operate and maintained,” he says.
Success with urine reclamation also depends on the economy. “If energy prices are low, it may not make sense to recover energy,” says Orner. “But right now, fertilizer prices worldwide are generally pretty high, so it may make sense to recover fertilizer and nutrients.” There are obstacles, too, such as few incentives for builders to incorporate urine recycling into new construction. And any hiccups like leaks or waste seepage will cost builders money and reputation. Right now, Orner says, the risks are just too high.
Despite the challenges, Ieropoulos envisions a future in which urine is passed through microbial fuel cells at wastewater treatment plants, retrofitted septic tanks, and building basements, and is then delivered to businesses to use as agricultural fertilizers. Although pure urine produces the most power, Ieropoulos’s devices also work with the mixed liquids of the wastewater treatment plants, so they can be retrofitted into urban wastewater utilities where they can make electricity from the effluent. And unlike solar cells, which are a common target of theft in some areas, nobody wants to steal a bunch of pee.
When Ieropoulos’s team returned to wrap up their pilot project 18 months later, the school’s director begged them to leave the fuel cells in place—because they made a major difference in students’ lives. “We replaced it with a substantial photovoltaic panel,” says Ieropoulos, They couldn’t leave the units forever, he explained, because of intellectual property reasons—their funders worried about theft of both the technology and the idea. But the photovoltaic replacement could be stolen, too, leaving the girls in the dark.
The story repeated itself at another school, in Nairobi, Kenya, as well as in an informal settlement in Durban, South Africa. Each time, Ieropoulos vowed to return. Though the pandemic has delayed his promise, he is resolute about continuing his work—it is a moral and legal obligation. “We've made a commitment to ourselves and to the pupils,” he says. “That's why we need to go back.”
Urine as fertilizer
Modern day industrial systems perpetuate the broken cycle of nutrients. When plants grow, they use up nutrients the soil. We eat the plans and excrete some of the nutrients we pass them into rivers and oceans. As a result, farmers must keep fertilizing the fields while our waste keeps fertilizing the waterways, where the algae, overfertilized with nitrogen, phosphorous and other nutrients grows out of control, sucking up oxygen that other marine species need to live. Few global communities remain untouched by the related challenges this broken chain create: insufficient clean water, food, and energy, and too much human and animal waste.
The Rich Earth Institute in Vermont runs a community-wide urine nutrient recovery program, which collects urine from homes and businesses, transports it for processing, and then supplies it as fertilizer to local farms.
One solution to this broken cycle is reclaiming urine and returning it back to the land. The Rich Earth Institute in Vermont is one of several organizations around the world working to divert and save urine for agricultural use. “The urine produced by an adult in one day contains enough fertilizer to grow all the wheat in one loaf of bread,” states their website.
Notably, while urine is not entirely sterile, it tends to harbor fewer pathogens than feces. That’s largely because urine has less organic matter and therefore less food for pathogens to feed on, but also because the urinary tract and the bladder have built-in antimicrobial defenses that kill many germs. In fact, the Rich Earth Institute says it’s safe to put your own urine onto crops grown for home consumption. Nonetheless, you’ll want to dilute it first because pee usually has too much nitrogen and can cause “fertilizer burn” if applied straight without dilution. Other projects to turn urine into fertilizer are in progress in Niger, South Africa, Kenya, Ethiopia, Sweden, Switzerland, The Netherlands, Australia, and France.
Eleven years ago, the Institute started a program that collects urine from homes and businesses, transports it for processing, and then supplies it as fertilizer to local farms. By 2021, the program included 180 donors producing over 12,000 gallons of urine each year. This urine is helping to fertilize hay fields at four partnering farms. Orner, the West Virginia professor, sees it as a success story. “They've shown how you can do this right--implementing it at a community level scale."
