Ethical implications of emotion mining in medicine.

• Emotion mining is a novel technique used in Artificial Intelligence to extract and analyse the emotions of individuals and populations. • In recent years these techniques have been applied to the field of Medical Artificial Intelligence (AI). • The use of emotion mining has widespread implications for both individual patient health and for population health. • Existing medical curriculums and ethical frameworks are not fit for purpose when it comes to these new e-health technologies. • A new approach is suggested for appraising medical AI, for both clinicians and policy makers. Background: The tools of emotion mining were developed in the field of computer science to detect and evaluate human emotions. The development of social media and digital phenotyping has empowered researchers to apply these techniques to the field of medicine. In recent years we have seen an evolution of novel health technologies that utilise data-driven emotion mining for mental health diagnostics and assessment. Aim: In this review of psychiatric AI we provide a high level overview of these technologies and their integration into clinical practice. The aim of this paper is to describe novel ethical scenarios posed by these models through a series of case studies, and establish how existing ethics frameworks can be adapted to meet these new challenges. With the techniques of emotion mining being used increasingly for predictions of self-harm, suicide and homicide this review comes at a crucial time. Methods: The article is presented as a narrative review and therefore does not detail the methodology of a systematic literature review. Instead, we offer a high level overview of AI in mental health and explore a number of case studies which highlight the novel ethical challenges these tools present. Results and Conclusion: Despite the growth of these technologies, there has not been a parallel growth in the ethical debate. For digital health to be implemented in an equitable manner, clinicians must be ethically equipped to appraise these systems. In the final section of our review we provide practical recommendations for healthcare professionals who are faced with implementing these new systems. [ABSTRACT FROM AUTHOR]