A Semi-autonomous System for Creating a Human-Machine Interaction Corpus in Virtual Reality: Application to the ACORFORMed System for Training Doctors to Break Bad News

A Semi-autonomous System for Creating a Human-Machine Interaction Corpus in Virtual Reality: Application to the ACORFORMed System for Training Doctors to Break Bad News M. Ochs1, P. Blache2, G. Montcheuil2,3,4, J.-M. Pergandi3, J. Saubesty2, D. Francon5, and D. Mestre3 Aix Marseille Universit´e CNRS ENSAM, Universit´e de Toulon, 1LSIS UMR 7296, 2LPL UMR 7309, 3ISM UMR7287 ; 4Bor´eal Innovation, 5Institut Paoli-Calmettes (IPC), Marseille, France Abstract In this paper, we introduce a two-step corpora-based methodology, starting from a corpus of human-human interactions to construct a semi-autonomous system in order to collect a new corpus of human-machine interaction, a step before the development of a fully autonomous system constructed based on the analysis of the collected corpora. The presented methodology is illustrated in the context of a virtual reality training platform for doctors breaking bad news. Keywords: Virtual patient, Virtual reality, Corpus, Health domain, Training 1. Introduction This paper presents a corpus-based methodology elab- orated to design a virtual reality environment aiming at training doctors to break bad news to a virtual patient. Many works have shown that doctors should be trained not only to perform medical or surgical acts but also to develop skills in communication with patients (Baile et al., 2000; Monden et al., 2016; Rosenbaum et al., 2004). Among all possible bad new, doctors can be faced to the complex situation of announcing a damage associated to a care, and that can engage their responsibility: unforeseeable medical situation, dysfunction, medical error, etc. The way doctors deliver bad news related to damage associated with care has a significant impact on the therapeutic process: disease evolution, adherence with treatment recommendations, litigation possibilities (Andrade et al., 2010). However, both experienced clinicians and medical students consider this task as difficult, daunting, and stressful. Training health care professional to break bad news is now recommended by several national agencies (e.g. the French National Authority for Health, HAS)1. Such trainings are organized as workshops during which doctors disclose bad news to actors playing the role of patients. This solution is complex to implement: it requires several persons, it is costly, and time consuming (each 30 mn. session requires one hour of preparation). Our project2 aims at developing a virtual reality training system with an embod- ied conversational agent playing the role of a virtual patient. The approaches classically used to develop an interactive system including an embodied conversational agent consist in analyzing automatically or manually one annotated cor- pus of human-human interactions (Cassell, 2000). Then, a fully autonomous prototype is developed and evaluated 1The French National Authority for Health is an independent public scientific authority with an overall mission of contribut- ing to the regulation of the healthcare system by improving health quality and efficiency. 2ACORFORMed, (Ochs et al., 2017), http://www.lpl- aix.fr/˜acorformed/. through perceptive studies. In the approach presented here, the methodology is based on several steps, making it possi- ble to acquire specific data and information about the re- quired modules and functionalities. More precisely, the procedure starts classically with the collection of natural data …

Content truncated to 500 words