Happy Appy Happy Talk: New app identifies diseases through speech patterns

The VAMP (Voice Analytics for Medical Professionals) app from Kraków startup Techmo, identifies changes in people’s speech patterns for early detection of diseases such as diabetes, cancer and heart problems. Techmo

A clever new app can determine a patient’s medical condition simply by registering changes in voice patterns.

The app called VAMP (Voice Analytics for Medical Professionals) from Kraków startup Techmo, identifies changes in people’s timbre, projection, and syntax for early detection of diseases such as diabetes, cancer and heart problems.

Multiple medical conditions are known to cause detectable changes in how we communicate, which are one of the first signs of emerging illnesses.Techmo

Multiple medical conditions are known to cause detectable changes in how we communicate, which are one of the first signs of emerging illnesses. 

With 70 percent of all deaths worldwide caused by diabetes, cancer and heart diseases according to WHO, early diagnosis is the key to saving lives.  The AI used in VAMP can not only detect the symptoms but assess the risk of the disease developing.

The AI used in VAMP can not only detect the symptoms but assess the risk of the disease developing.Techmo

Professor Bartosz Ziółko from the AGH University of Science and Technology in Kraków, Techmo’s CEO, said: "Doctors aren't used to basing their diagnosis on what can be caught in the voice. But when they start to think about it, they are usually able to catch changes related to certain diseases in their patients’ voices.”

Together with the Medical University in Białystok, they devised the voice analysis app to help doctors and patients with the diagnosis process. For example, diabetes can lead to making cognitive errors more often (such as errors in inflection or confusing consonants).

Techmo CEO Professor Bartosz Ziółko from the AGH University of Science and Technology in Kraków, said that the app can be installed on a smartphone so individuals can regularly monitor their health.Techmo/Facebook

Atherosclerosis, on the other hand, can make the voice shiver more. Some psychiatric illnesses, such as depression or bipolar disease, can be associated with mood changes. And they can be recognized by the emotional character of the speech and the words you use.

Techmo’s solution doesn’t require any specialised equipment and can be easily installed on a smartphone. With the users' permission it can gather data while they make phone calls and record the changes over a period of time. The AI then compares the current voice patterns with older samples.


Although the app's prototype is ready for voice detection of atherosclerosis or diabetes it still needs more data and testing before it can be officially registered.

The startup has also applied for an EU grant to develop technology for COVID-19 detection.

With the users' permission it can gather data while they make phone calls and record the changes over a period of time.Techmo

Posting on Facebook the team said: “Techmo is responding to the EU's call to fight against coronavirus. With significant experience in research on voice detection, Techmo decided to respond to the European Union's call and submit a project on voice detection of coronavirus.

As is well known, SARS-CoV-2 attacks the respiratory tract, making changes in the voice a significant symptom that we can detect.”