Smartphone App Can Detect Early Signs of Stroke
​To check the condition of the heart with the app, the person has to lie down, set the phone on their chest and start the measurement. The app analyses the data and gives an immediate answer whether or not the person has atrial fibrillation. The final diagnosis is confirmed with traditional electrocardiogram registration.

​Atrial fibrillation is a dangerous disease present in 2% of the global population. It causes up to seven million strokes per year.

– In the European Union alone this heart rhythm disorder costs approximately €17 billion every year, said lead author Tero Koivisto, the Vice-Director of the Technology Research Centre (TRC) at the University of Turku.

– Around 70% of strokes due to atrial fibrillation could be avoided with pre-emptive medication. However, atrial fibrillation often occurs randomly and has very little symptoms, if any, which is why it is so hard to detect. Therefore, the pre-emptive medication for a stroke is sometimes never started, says Professor of Cardiology Juhani Airaksinen, who is also the Head of Operational Division at the Heart Centre of Turku University Hospital.

Currently, there are electrocardiogram (ECG) devices that patients can take home for long-term monitoring, but they have many constraints, such as large size and high cost. Moreover, current methods for the detection of atrial fibrillation are infeasible for wide-scale screening of populations or higher risk age groups, such as people over 65 years of age.

Results Are Promising

The researchers tested how well atrial fibrillation can be detected just by using a smartphone application. The study included 16 patients with atrial fibrillation from the Turku Heart Centre and a control group of 20 healthy people. The validity of the app was evaluated with the data collected from both groups.

To detect atrial fibrillation, a smartphone was placed on the chest of the patient, and accelerometer and gyroscope recordings were taken. Patients were advised to lie still in a prone or supine position during the measurements.

– We use the smartphone's inbuilt accelerometer and gyroscope to acquire a heart signal from the patient. Each smartphone has an accelerometer and most models also have a gyroscope. The measurement data is pre-processed by signal processing methods and different analyses, such as autocorrelation and spectral entropy, are then extracted from the data.  Finally, a machine learning algorithm is used to determine if the patient suffers from atrial fibrillation, Koivisto describes the new method.

With this technology, the researchers detected atrial fibrillation in more than 95% of the cases.

Testing Can Be Done at Home

– No extra equipment is needed for the measurement, people just have to download the app with the algorithm we have developed. The app should be available to the public in 2017.

To check the cardiac status, the patient can simply lie down, place the phone on their chest and start the measurement.  The app analyses the results and gives a simple yes/no answer as to whether the patient has atrial fibrillation or not.

– This is a highly cost-effective and easy way to detect atrial fibrillation. The final diagnosis is confirmed with traditional ECG registration. The method can also be used for screening large populations as there are over 2 billion smartphone users in the world. In the future, a secure cloud service could be created to store and analyse larger masses of data, Koivisto envisions.

The study was funded by Tekes and the Academy of Finland.

The research results were published at the European Society of Cardiology Congress in Rome in August: Tero Koivisto, Olli Lahdenoja, Tero Hurnanen, Timo Knuutila, Tuija Vasankari, Tuomas Kiviniemi, Antti Saraste, Juhani Airaksinen, Mikko Pänkäälä, “Detecting atrial fibrillation via existing smartphones without any add-ons”, ESC Congress (European Heart Journal supplement), 2016.

>> Technology Research Center

Photo: Sami Nieminen

Published date 9/2/2016 2:15 PM ,  Modified date 9/2/2016 2:43 PM

20014 Turun yliopisto, Finland
Tel. +358 29 450 5000

People search

Follow us: 
Facebook   Twitter   Instagram   Youtube   LinkedIn
© University of Turku