Physical activity is an integral part of a healthy lifestyle, but more and more people are not active, work as a sedentary worker, and spend their free time in front of a computer screen or television. Insufficient physical activity is closely associated with complicated cardiovascular diseases, such as myocardial infarction. There is a growing debate about effect physical activity profile have on health, i.e. duration of physical activity sessions, their frequency and distribution in time in regards to other sessions. The hypothesis is that individualized physical activity profiles for target groups of people could help reduce the number of deaths caused by cardiovascular diseases. Current methods for assessing physical activity are biased (questionnaires, diaries), and therefore there is a need for more comfortable and precise means to assess physical activity. With the rapid development of technology, it is possible to evaluate the physical activity objectively using smart wristbands capable of counting the steps.
In order to contribute to the solution of the listed problems, we present a method for assessing the physical activity profile using the pedometer data received from the smart wristbands. The presented method is capable of separating if the steps are concentrated over a given time interval (e.g., the weekend), or are distributed evenly throughout the monitoring period. The method to differentiate different physical activity profiles has been studied with a database of 40 patients with cardiovascular diseases and the amount of steps that they walked. The patients wear Fitbit Charge 2 (Fitbit, San Francisco, USA) for a minimum of a week.
Four prevailing physical activity profiles are distinguished that gain the different values of the estimate. Physical activity profiles, in which the steps are distributed evenly, acquire the estimate values of close to zero, while the profiles in which the steps are concentrated in a short interval of time acquires the estimate values of close to one.
Analysing the daily physical activity profiles of patients with cardiovascular diseases, it has been observed that this group of patients is associated with a higher estimate. This shows that the subjects spend more time being physically passively, their physical activity is concentrated in a short interval of time. The difference between the physical activity profiles of women and men is also noted. The women’s group is associated with a lower estimate, which indicates that women’s physical activity is evenly distributed (most likely due to housework).
Our judgement of the physical activity profile is evaluated objectively and it is able to recognize the different profiles of physical activity, so the individual profile for cardiac patients can be applied. The proposed solution is also suitable for smart wristbands or other devices that track physical activity.
Duration:
2018 - 2018
