Implementation of expert systems to support the functional evaluation of stand-to-sit activity

BACKGROUND Functional evaluation of sit-to-stand and stand-to-sit activities is often used by physiotherapists in patients with neurological and musculoskeletal disorders. The observation of the way these activities are executed is essential in identifying kinesiological problems. There are different methodologies used to describe the stand-to-sit activity and its evaluation is not yet standardized, which makes the practical application of resources on clinical observation difficult. The objective of this study is to automate the decision making process of an evaluation protocol, developed in previous study, and facilitate its utilization by professionals in the area. METHODS A decision-making system has been implemented through a computational tool, more specifically an Expert System that due its inherent characteristics emulates the decision-making process of a human expert in the domain area. A Shell called Expert Sinta was used to develop two knowledge bases, i.e. two expert systems, one for the anterior view and another for the lateral view of stand-to-sit activity. Variables, values, associated rules and confidence factors, objectives, and additional information questions were defined by the expert of domain and once implemented each expert system generates a number of questions to its user. These questions serve as a guide to physiotherapists and support the standardization of the activity evaluation. The developed systems were evaluated by physiotherapists through the application of a questionnaire that evaluates the knowledge base and the usability of the system. The physiotherapists' answers were then evaluated through statistical estimation and percentage analysis. RESULTS When asked about the systems' "utility for clinical practice of the physiotherapist", 67% of evaluators answered positively. An interesting finding was that most physiotherapists (i.e. 92%) considered that the systems are suitable for educational purposes, which was not the main objective of this study. CONCLUSIONS The developed expert systems can support the physiotherapist in evaluating stand-to-sit activity through a conclusion suggestion about the "level of inadequacy" for the "degree of inadequacy" searched during its execution. Results of experts' evaluation analyzed through statistical methods indicate that the automation of protocols contributed to the standardization of the evaluation of stand-to-sit activity and that it has application for teaching purposes.