The University of Southampton

Patient-centric model for remote management, treatment and rehabilitation of autistic children

Pervasive Healthcare and Telemedicine

Quantitative EEG is used in medical laboratories to determine the brain connectivity in autistic children but due to the artificial nature of this lab-based approach its validity under a real-life scenario is doubtful. Also therapeutic interventions are executed typically in a clinical setting and have a limited extension in time while it has been proved the beneficial effect of an intensive intervention. Moreover, although pointed out by several researcher the need of a personalized treatment for each child, there is a little knowledge about how to identify the most suitable treatment or integration of treatments for a specific child. The MICHELANGELO project intends to bring the assessment and the therapy of the autism out of the clinical environment and develop a patient-centric home-based intervention solution requiring a minimal human involvement and therefore extremely cost effective. The main outcomes of the project are intended to address both the assessment and the therapeutic intervention in autistic children :

1. Autism Assessment • A camera-based system that – triggered by a wearable EEG solution – will take snapshots of the scenes according to the eye movement and will allow to identify the stimuli that in a natural environment (the child’s home) cause significant responses in the autistic child; • The same pervasive wearable EEG system that - used in conjunction with an eye-tracking device - measures the brain activity while the patient is presented with the reproduction of the identified stimuli and allows to better characterize them. The eye tracking system will record the eyes’ fixations and will enable to see how the brain connectivity changes with naturalistic stimuli and allows determining the particular properties of the objects (e.g. shape, colour etc.) the child is looking at and most responsible for the abnormal brain wave behaviour. The system - being pervasive in nature - makes the patient “less aware” about the artificial nature of the experiment and therefore affects the cognitive activity far less than the currently employed systems. • A set of advanced and sophisticated signal processing algorithms enabling accurate characterisation of stimulation-specific brainwave anomalies and connectivity between different brain regions and hence giving vivid insight into the process of information integration ability of the brain in a stimulus-specific way. 2. Autism Therapy • The design of a personalized intervention protocol based on a heterogeneous strategy where well consolidated developmental/behavioural thera peutic approaches are combined with neurophysiologically established new connectivity-guided neuro feedback techniques. Intervention will use also ICT-based solutions such as virtual reality, serious game and robotics. • A set of unobtrusive tools for the continuous monitoring of the autistic children during the therapeutic program allowing the adaptation and personalization of the intervention. They include a wearable multi-parametric platform measuring relevant vital signs together with a camera-based system for the recognition of repetitive behaviors. • Sophisticated algorithms for advanced imaging techniques (Diffusion Tensor Imaging and fMRI) which are used by the doctor to check the anatomical and functional connectivity of the brain at different steps during the therapy and assess its effectiveness. • A novel system enabling intensive intervention in “patient-centric” way in the home settings with minimal human intervention leading to significant cost reduction while substantially enhancing effective clinical outcome. Finally the MICHELANGELO project will set the bases to validate the results of its research work in an exploratory study executed in Italy and in France.

Primary investigator

Secondary investigators


  • 7 EU partners

Associated research groups

  • Electronic and Software Systems
  • Pervasive Systems Centre
Share this project FacebookTwitterWeibo