An Expert System to Design Intervention Strategies for Children with Disabilities in Multi-Sensory Stimulation Environments

The Multi-Sensory Environments (MSEs) have the objective of improving the development of thought, intelligence, and social skills of children with and without disabilities. However, in developing countries, due to the lack of personnel and resources, the process of designing intervention plans for children with disabilities in MSEs is carried out from an empirical perspective. For these reasons, in this paper, we present a system that relies on a multi-sensory environment consisting of 7 rehabilitation modules and an expert system to automatically determine which rehabilitation activities are appropriate for children with disabilities. To this aim, the expert system uses a model based on decision trees regressors with multi-outputs and suggests a therapy plan selecting pairs activity-module (from a total of 7 modules). We tested the system with a group of experts using three criteria: completeness of the plan, coherence (activities versus child’s profile), and applicability (relation between activities and each module). Currently, the system can suggest plans for children with different disabilities such as Infantile Cerebral Palsy (ICP), autism, hypogenesis of the corpus callosum, intellectual disability, epilepsy, and hydranencephaly.