Don’t prick your fingers! Scientists detect diabetes through smartphone cameras: accuracy rate is more than 80%

According to the University of California, a new method of diagnosing diabetes by using a new type of mobile phone camera was reported on February 8 by the University of California, San Francisco. There are no symptoms in the early stage of diabetes, so it is easy to be ignored. In addition, scientists have so far decided to carry out a noninvasive diabetes detection tool. < / P > < p > the system is based on the use of photoplethysmography. It uses photoelectric sensors to detect the difference of reflected light intensity after absorption by human blood and tissues, and trace the change of vascular volume in the cardiac cycle, and calculate the heart rate from the pulse waveform obtained. < / P > < p > when smart phone cameras appeared, scientists realized that they could capture PPG measurements. The researchers speculate that using PPG data captured by smartphone cameras could detect vascular damage caused by diabetes. < / P > < p > the researchers first detected 2.6 million PPG records of 53870 diabetic patients through deep neural network, and developed a deep learning algorithm based on this, and verified whether it can effectively identify diabetic patients and healthy subjects. < / P > < p > after developing the algorithm, the researchers tested the ability to detect diabetes using PPG data from smartphones, which were collected through the phone’s flash and camera. < / P > < p > the results showed that the system accurately detected about 80% of the subjects with diabetes. When the algorithm is combined with other basic patient data, its prediction potential is further improved. < / P > < p > researchers say that in this era of smartphone popularity, using it to detect diabetes will effectively reduce the risk of diabetes. At present, we are transforming the algorithm into an application. Next, we will test how this new digital tool can be better integrated into the practice of diabetes screening. Video Number assistant internal test online! Four functions let you send 1g video on the computer