Machine Learning-based Panic Syncsight: An IoT-Integrated Panic Attack Detection System for Enhanced Therapeutic Insights

People who experience panic attacks usually find it very difficult to convey what they experience. Apart from this, another major challenge is to actually know when the panic attack happens and how severe it was physically. Real-time healthcare trackers and anxiety level trackers lack specificity and accuracy in detecting panic attacks. The proposed system is specifically designed to detect panic attacks and ultimately detect the severity of the condition the patient is in. Usage of advanced sensors and random forest machine learning algorithms pushes it to 80% to 95% accuracy with proper calibration. The system aims to detect and record the time and duration of the panic attack by reading the physical response of the user. Since the human body shows the same physical symptoms such as increased pulse, heart rate even during a normal physical workout, advanced sensors are also being included such as Galvanic Skin Conductance sensor. The data is being trained in such a way that the system would detect only panic attacks and not other non-panic attack events. The data thus confirmed as panic attacks are being recorded and the details are being sent to the concerned persons via IOT System that have integrated with system. This data is accessible only to the physician handling the patient and the guardian. This in short helps both the physician treating the individual and the individual who is facing the panic attacks himself/herself.