Researchers at the University of Buffalo have developed a necklace that could help passively track nutrition by listening to the sound of the wearer chewing.
University at Buffalo computer scientist Wenyao Xu is developing the wearable, called AutoDietary, with a small microphone on the back that records the sound of people eating food. The sounds are sent to a smartphone via Bluetooth and compared against a database of chewing and swallowing sounds for different foods, including cookies, peanuts, and apples.
“There is no shortage of wearable devices that tell us how many calories we burn, but creating a device that reliably measures caloric intake isn’t so easy,” Xu said in a statement.
Xu has already used AutoDietary in a study of 12 people, published last month in IEEE Sensors Journal. Each participant was given water and different foods, including apples, carrots, potato chips, cookies, peanuts and walnuts, to eat. The system was able to identify the correct food 85 percent of the time.
AutoDietary, Xu, explained, could be used to help people with a number of conditions including diabetes, obesity, bowel disorders. One of the downsides to the necklace that Xu identified is its inability to differentiate between very similar foods, for example corn flakes and frosted flakes, but he added that he is working on integrating a biosensor into the system that would be able to better track the nutritional data of food.
Over the years, a couple of research groups and companies have attempted to develop a passive nutrition-tracking wearable in form factors.
In 2014, researchers at the University of Alabama developed a diet-tracking sensor that collects information based on the wearer's chewing. The sensor, called Automatic Ingestion Monitor, or AIM, is fitted around the user's ear and monitors vibrations from jaw movement. AIM is programmed to ignore jaw motions from talking. This data, paired with pictures the user would take of their meals, would give them information about how much and how often they eat.
That same year, GE Research developed a prototype of a microwave that could someday measure calories in the foods that a user warms up. This first prototype is only capable of measuring the caloric content of some liquids. After combing through U.S. Department of Agriculture nutritional information on 6,500 foods, GE Researchers found that they could get a mostly accurate calorie count from the food's fat content, water content, and weight.
There have also been a few attempts at passive nutrition tracking wearables by smaller startups. Two of the startups, Ontario-based Airo and HealBe claimed they had developed wristworn passive nutrition tracking devices, and another, Telspec, is a miniature spectrometer that claimed to be able to deliver information about allergens, chemicals, nutrients, calories, and ingredients in food that it scanned to a user's smartphone.