Neurology & Neuroscience

Stress has become part of students’ academic life. Adolescents are particularly vulnerable to the problems associated with academic stress as transitions occur at an individual and social level. Albeit, it becomes imperative to understand the academic stress in order to derive efficient intervention strategies. Hence, the study was aimed for

“Evaluation of stress among medical students during examination using Artificial Intelligence based Graphology and its correlation with salivary cortisol”. This study employed a quantitative research design where 43 medical students (19 males and 24 females) enrolled as subjects, within the age group of 18-26 years. Subjects were monitored to follow the protocol prior to the collection of salivary samples. Salivary samples were collected during pre examination and post examination for cortisol estimation by Competitive ELISA method. Cortisol the “stress hormone” spike during

times of high stress in the body. Salivary cortisol has been used as a biomarker of psychosocial stress and can be indirectly used to assess psychobiological mechanisms that trigger the hypothalamus-pituitary-adrenal axis. Written examination manuscript images were captured for Artificial Intelligence based Graphology analysis. Graphology can be a useful tool in spotting health problems before they become too severe, and is excellent at identifying stress in the individual. The negative traits, which reflects the stress was determined by handwriting were explored in this study. Artificial intelligence (AI) aims to mimic human cognitive functions. In medicine, applications of artificial intelligence have been innovative. The diverse areas of AI include: neural networks, programming languages, genetic algorithms, speech/handwriting recognition etc. The results of the study indicated statistically significant elevated salivary cortisol levels during examination. Students with raised salivary cortisol during post examination levels showed raised negative graphology traits score & low academic performance (P< 0.05). Handwriting was processed through trained Convolutional Neural Network Model of AI. Compared the Negative traits extracted by CNN with

Negative traits determined by Graphology. This study is the first of its kind in India to prove that empirically AI based graphology can be used to assess the stress levels as a screening assistive tool for Psychologist, counselors & academic mentors. It would facilitate the development of effective counselling modules and intervention strategies

in order to help the students to alleviate stress.