A System Based on AI and M1 Enhanced to Investigate Physiological Markers for User Forecasting Decision-Making

Abstract

By creating an explicatory AI report for stress prediction using wearable sensors, we fill this gap in our work. We based the explanatory AI's appearance and feel on a typical blood test report because medical professionals and patients are probably familiar with theseUsing physiological data, we provide in this study a novel, comprehensible AI technique for stress prediction. Users and medical professionals may decide which biological characteristics, in addition to any anomalies connected to health, have the greatest influence on the prediction of stress based on the report. Both quantitative and qualitative indicators were used to assess the usefulness of the study on explainable AI. It was demonstrated that the stress prediction accuracy was on par with modern standards. There was shown to be a correlation between the ratios of the individual physiological indicators used to make the stress prediction and the actual results. A qualitative survey with psychiatrists was also conducted in addition to these quantitative evaluations to further bolster the credibility and efficacy of the AI system's explanatory report on the stress. Additional explanatory elements pertaining to the patient's various emotional states, such as melancholy, relaxation, anxiety, or happiness, will be added in future study.

Ähnliche Arbeiten