Graphology or handwriting analysis work on principle that while writing, our hand is controlled by our subconscious mind. The graphic action reflects the state of the subconscious itself. It will be unique for each individual. Instead of graphologist who interprets individual character which is prolonged and susceptible to errors, the work automatically determine the personality trait using Deep learning and Task fMRI to accelerate the process and reduce error .The dataset consists of 129 different person's handwriting on a phone note or tablet pc. The emotional state of participants is assessed by the DASS depression-anxiety-stress scale. DASS-Depression centers around things identified with low inspiration and confidence, DASS-uneasiness to fear and panic and DASS-worry to strain and fractiousness. The characteristics comprise of sorrow, nervousness, stress and the mind state is named Normal, Mild, Moderate, Severe, Extremely Severe. Emotional state is related to handwriting. X position,Y position, Time, pen on, azhimuth, altitude angle, applied pressure. The features also include stroke order, direction and speed. The independent components obtained in analysis of task fMRI and the features obtained by the recognition are fused to classify the emotional state using Deep learning multilayer perceptron as positive nature, negative nature , idealism, broadminded, meanminded, reserved, selfcentered, roll model, average ,attentive. The obtained emotional state proves to be a basic model for Brain computer interaction.