Decoding Brain Eeg Signals To Find State Of Mind Using Ai

Abstract

Emotions are primordial for human beings and they play a key role in human intelligence. Emotion is basically connected with sight, human correspondence and logical decision making. Now a days the need for attested and dependable remedies for the recognition of human emotional states is obligatory due to the rise in interest of upcoming researchers towards establishing some significant emotional interactions between humans and computers.By analyzing the features of electroencephalography (EEG) signals, which are generated from EEG sensors that noninvasively measure the electrical activity of neurons inside the human brain, we select the optimal combination of the features for recognition. Then the signals are pre-processed using Hjorth parameters that measure signal activity of time-series data. The classification of signals obtained is based on supervised pixel classification. By using convolutional neural networks(CNN) the signal feature obtained is compared with the parameters set and thus it detects the state of mind whether the patient is happy, depressed or anger and the output generated will be in a text format. This is very much beneficial in many sectors especially in health sector  where dealing with patients diagnosed from Locked In Syndrome, coma and various neuropsychiatric disorders .By detecting the emotional state of the patient which means to detect whether they are depressed or anger or happy and it will help the doctors in treating them in a better way and the patient can recover soon. In case of depression it is mandatory to treat those kind of patients or else they may take some psychotic decisions like suicide or they may become mentally weak. Similarly if the emotional state detected is anger, then we can make them happy by doing what they like which will help  to make them normal In patients diagnosed with locked in syndrome their full body will be paralyzed except their eye muscles They can think,feelemotions,sense smell but cannot move. If we detect their state of mind it will be more beneficial in treating them..The accuracy of the overall project will be around 83%.Our approach shows better performance compared to existing algorithms