Intelligent assessment based on Beta Regression for realistic training on medical simulators

Simulators can provide safe training systems by the presentation of several possibilities of situations related to a specific topic. The technology of virtual reality have been used to the conception of realistic and interactive simulators that can be used to this end. Intelligent tools can be coupled to these systems to collect users’ interactions data and use it to assess their performance and skills. Since simulation systems can reduce costs and improve traditional training programs, the presence of good assessment tools coupled to them can help to identify users’ mistakes and improve the acquisition of skills. An assessment system is a knowledge-based system which analyzes that data and provides a decision making support on the users’ proficiency in performing real procedures. In this paper we present a methodology to include intelligent assessment based on Beta Regression to monitor users’ actions in a virtual reality simulator for gynaecological examination. The performance of the assessment methodology was analyzed using Monte Carlo simulation. The results show that this approach is satisfactory to classify user’s skills level.