Real-Time Prediction of Panic Attacks Through Sweat Analysis in Wearable Devices

Panic attack has become a major issue not only in adults but children's too. School going students are suffering from eternal stress conditions which cause them to suffer from panic attacks. A panic attack at the wrong time can cost someone their careers and life. We are proposing hardware or an extension to smart watches equipped with machine learning models to predict if a person is going to have a panic attack. This can help early intervention of panic attack and take measures for easing patients or avoiding the panic attack can be taken. A ring made of semi absorbent strap which can be used to trap palm sweat and get input of it which will be crucial for detection of panic attack. The mathematical model we are proposing mostly revolves around two major indicators, cortisol and folic acid. Concentration of these substances in sweat can be a clear indicator of a possibility of a panic attack and we'll build our model upon this. We will be monitoring sweat rate using real time sweat monitors or humidity monitors which will act as a switch for our algorithm. Once sweat rate reaches Smax then further algorithm is launched based on cortisol and folic acid levels.