AI can detect Type 2 diabetes from 20 seconds of speech

By Stephen Beech

Diabetes can be detected in just 20 seconds from the way someone talks, reveals new research.

Artificial intelligence (AI) technology could transform screening for the chronic condition, say scientists.

The largest study of its kind concluded that AI-based analysis of short recordings — which can be done over the phone — could be a fast, noninvasive and scalable screening tool for diabetes.

Type 2 diabetes is becoming increasingly common and early detection is key to preventing serious complications, such as heart disease and nerve damage.

However, many cases still go undetected — putting pressure on healthcare systems.

In the U.K., around 30% of diabetes cases are undiagnosed and around 60% of the £10.7 billion ($14.5 billion) NHS spent on diabetes each year goes on managing complications.

Screening currently involves blood tests or GP appointments.

The NHS includes diabetes screening in the health checks it offers to people aged 40-plus every five years.

But research has shown that fewer than half of eligible adults (40.4%) attend the useful, but time-consuming, appointments.

Previous studies have linked Type 2 diabetes to changes in speech — such as increased hoarseness and roughness and inability to control breath and voice as well while speaking.

Researchers at London-based tech firm thymia, working with colleagues from RMIT University in Melbourne, Australia, first developed an AI model to detect the changes.

They trained it on 63,283 voice samples from 21,129 people in the U.K. and U.S. who'd reported whether they had been diagnosed with diabetes.

They then tested the speech model on remote 20-second recordings by people reading one of Aesop's fables.

The first of two evaluations compared the model's performance on recordings made by 7,319 adults in the U.K.

It found that the speech model gave a higher risk score to people who reported having Type 2 diabetes than to those who did not report having the condition 80% of the time, considered to be clinically useful.

The model performed well across different sexes and ages.

However, performance was lower on recordings from Black participants.

That was likely due to the low number of Black people reporting Type 2 diabetes, according to the research team.

More from this section

Woman dies after insect bite becomes weeping wounds

Recognizing and shifting ADHD procrastination

Baby monitor helps mom spot daughter’s rare cancer

Performance was also lower on recordings from people with heart disease, high blood pressure or obesity.

That was believed to be because the conditions often coincide with Type 2 diabetes and may cause similar vocal changes.

The second evaluation involved a subgroup of 801 participants who took HbA1c blood tests at home within three months of the speech recording.

The HbA1c blood test measures average blood sugar levels in the previous two to three months and is the gold-standard test for Type 2 diabetes.

It can also be used to detect prediabetes, where blood sugar levels are higher than normal but not high enough to be classed as diabetes.

The speech model gave a higher risk score to people with Type 2 diabetes than to those who did not have the condition, based on HbA1c tests, 75% of the time.

It correctly picked up 82% of the participants with Type 2 diabetes.

The speech model was also able to distinguish between people at low, medium and high risk of Type 2 diabetes, based on their HbA1c results.

None of the people the model classed as low risk had blood results in the diabetic or prediabetic range.

The researchers concluded that Type 2 diabetes can be detected at "clinically useful" levels from 20 seconds of speech.

Giedrė Čepukaitytė, a research scientist at thymia, said: "Pending further clinical validation of the tool, one option would be for GPs to use short recordings to triage patients.

"Those at higher risk could then be given blood tests to confirm their status."

The researchers say speech-based screening would sit alongside blood testing, not replace it.

Čepukaitytė is due to present the findings at the annual meeting of the European Association for the Study of Diabetes in Milan.

She said: "This is the largest real-world study of speech-based screening for Type 2 diabetes to date which also checks the model's predictions against blood test results as well as against what people reported about their own diagnosis.

"Those flagged up as higher risk by the model had blood results to match.

"This has the potential to change what screening looks like.

"A speech sample can be taken over the phone or through an app, so we can reach far more of the people who need a blood test than current pathways do, particularly those who never get to a health check.

"Our model opens a new route to screening for diabetes.

"It is not a replacement for a blood test, and it should never stop anyone who thinks they need one from getting one."

Čepukaitytė added: "Our next step is to test the model in clinical settings and to understand how well it works for every group of people, because a screening tool has to work for everyone."