Content area
Full text
* Corresponding author. E-mails: [email protected] ; [email protected]
1. Introduction
Medical education requires trainees and practising doctors to develop expertise in diagnosis or clinical reasoning. These skills are traditionally acquired through clinical practise, and they may be enhanced with the help of simulations with mannequins, role-playing games or virtual patients (VPs) (Rombauts 2014). More broadly, the literature uses the term virtual patient to refer to simulations such as case presentations, interactive patient scenarios, high-fidelity mannequins, VP games, high-fidelity software simulations, human standardised patients – who are actors playing the role of interviewed patients paid for educational purposes – or virtual standardised patients (Talbot et al.2012a). VPs allow health professionals to practise their skills by interacting with a software ‘that simulates real-life scenarios’ (Cook et al.2010). In our work, VP refers to virtual standardised patients. For the last few decades, VPs have allowed doctors to train clinical and history-taking skills through simulated scenarios in digital environments (Ellaway et al.2006; Danforth et al.2009).
Interactivity with a VP might be enhanced through a dialogue system, but such a component needs to address several phenomena to achieve a natural, user-friendly dialogue (Figure 1). As shown, medical doctors tend to begin by eliciting initial clues from the patient by using broad questions. Then, they use follow-up questions to focus on specific details. The system needs to deal with this behaviour by processing context information (ellipsis and anaphora) and updating its information state, so that it avoids providing redundant answers. In addition, term variants referring to the same concept need to be mapped accurately (e.g. hypertension ↔ high blood pressure) by means of linguistic and terminological knowledge.
Figure 1.
Sample dialogue (D: doctor; P: patient, simulated by the system) and relation to the patient record. Phenomena to be addressed are shown in bold: discourse phenomena, linguistic variation (blue circles) and termino-ontological variation (green boxes). The patient record structure is simplified due to space constraints. We show real replies of the current version of the French system, but we translated them to English.
[Figure omitted. See PDF]
This work describes our endeavour to create a dialogue system featuring unconstrained natural language interaction in a simulated consultation with a VP. We built this system to simulate history-taking in an...