• Corpus ID: 218466433

Artificial Intelligence in Behavioral and Mental Health Care

@inproceedings{Hudlicka2015ArtificialII,
  title={Artificial Intelligence in Behavioral and Mental Health Care},
  author={Eva Hudlicka},
  year={2015}
}

• E. Hudlicka
• Published 2015
• Computer Science

The past decade has witnessed tremendous progress in the development of affective intelligent virtual agents and serious games (games used for training, coaching, therapy, and other non-entertainment purposes). These developments, in conjunction with an expanding interest in telemental health, have led to a growing recognition that these technologies can significantly enhance clinical practice and improve patient outcomes. This chapter provides a brief history of intelligent virtual affective agents and serious gaming technologies. The state-of-the-art in these areas and recent applications in health care, specifically behavioral health, are presented. Ethical and privacy considerations, as well as future prospects, are also discussed. Intelligent virtual agents (IVAs) are synthetic virtual, computer-controlled characters that can interact with humans (Prendinger & Ishizuka, 2004). Virtual affective agents are IVAs capable of affective interaction, often emphasizing nonverbal interaction; that is, recognition of human emotions, and expression of “emotions” by the agent (Hudlicka et al., 2008; Hudlicka et al., 2009). Affective agents may also include affective user modeling and may explicitly model synthetic emotions within their architectures. These capabilities enable the agents to display degrees of emotional and social intelligence (e.g., awareness of social cues and user goals), to have some ability to form and manage relationships, and to adapt in real-time to the changing states and needs of the human user. Both the visual appearance and the multi-modal interaction capabilities of these agents vary greatly, and can be customized to match the human user’s preferences and needs. The embodiments range from cartoonish, animated characters or faces, to fully articulated full bodies, again, customizable to the user’s preferences and needs. The interaction capabilities range from text-based, multiple-choice user input and