New algorithm ‘boosts accuracy’ of physical activity apps

There are many cell phone apps that track your physical activity. These are useful, not only for your own sense of achievement, but also for doctors who look to these apps to track a patient’s movement and develop tailored treatments. Now, researchers have created a way to make these apps even more accurate.

 A team from Northwestern University in Chicago and Evanston say that previous research surrounding physical activity apps has shown that the majority of people who use them carry their cell phones in a bag, purse or pocket, or they attach it to a belt.

But the researchers note that the location of the phone can have a significant impact on how well the app can pick up a person’s movement.

Therefore, they have created a computer algorithm that can be used in conjunction with a physical activity app that is able to “predict” the location of a mobile phone throughout the day with “near perfect accuracy.”

“Most women carry their phones in a purse. Some people carry theirs on their belt or in their hand. We may change where we carry our phone throughout the day as well,” says first author Stephen Antos, of the McCormick School of Engineering and Applied Science at Northwestern University.

“We wanted to solve this problem and find a way to make these trackers as accurate as possible, no matter where you carry your phone.”

Algorithm ‘accurately detects second-by-second activity’

To create the algorithm, researchers recruited 12 healthy participants who were required to carry out a series of physical activities, including walking, sitting and standing.

During these activities, the participants carried smart phones in different locations – in their purse, backpack, belt, hand or pocket. This same method was also used on two people who had Parkinson’s disease.

The data from this experiment was then used to “train” a computer algorithm to predict where a person is carrying their cell phone, and from this, it was able to successfully detect “second-by-second” physical activity.

Commenting on the findings, published in the Journal of Neuroscience Methods, principal investigator Konrad Kording, of the Northwestern University Feinburg School of Medicine, says:

“While it remains true that smart phone activity trackers are the most accurate when the phone is placed in the pocket or on a belt, with this algorithm we can provide an estimate of error associated with other locations where the phone is carried.”

The researchers note that this algorithm can be applied to patient populations without hesitation, and Kording believes that in the future, smart phone apps will play a major part in helping us to manage our health.

“I believe we will have apps running on smart phones that will know exactly what we’re doing activity-wise and will warn us of diseases before we even know that we have those diseases,” he says.

“In the future, phones will have a major role in motivating people toward behavior that is good for their health.”

Source: Medical news today

Leave a Reply

Your email address will not be published. Required fields are marked *