As technology advances, sedentarism is the new way of life. It is all about comfort and comfort above all costs! But it aggravates the health risk and brings in a load of health problems, thereby weakening the body. Scientists have made commendable progress in finding out why the human body fails and what are the indicators of approaching death.
Bruce Schatz from the University of Illinois Urbana-Champaign and his team have done an extensive study to map out the important indicators of five-year mortality. They used a machine learning model to develop an algorithm that could find predict if a person would die in the coming years.
Smartphone’s Motion Sensor can predict death risk?
Earlier studies used a motion sensor to gather data on participants and predict mortality rates using that. But Bruce Schatz and his team decided not to conduct the same test again using wearables like smartwatches or fitness bands. Instead, they decide to sieve through the data and find correlations between them.
Bruce used the data of about one-tenth of the participants through a machine-learning model. The ML models developed an algorithm that estimated five-year mortality risk. It used acceleration during a 6-minute walk as a primary factor to calculate the mortality rate. But the question arises – is it accurate enough?
The scientists determined the c-index score. It is a metric used in biostatistics to assess accuracy. The result was 0.72, which is considered fairly accurate for measuring the mortality rate. But they use the data from wearables and not smartphones, so how could they predict if someone’s life is at a risk?
It turns out that wearables aren’t that common among non-affluent and non-tech-savvy people. But since the study uses the accelerating data of a 6-minute walk, it is can be done using the smartphone too. Even if the users don’t stay close to the phone all day long, all they need to do is perform some physical activity for 6-minutes, and that data will be enough to predict the mortality rate. “If people carry phones around, you could do a weekly or daily prediction, and that’s something you cannot get by any other method,” said Bruce Schatz.
