JMIR Publications

ABSTRACT

Background:

Neurodevelopmental disorders (NDs) are characterized by heterogeneity, complex and interactions among multiple domains with long-lasting effects in adulthood. Identifying and assessing children at risk for NDs is crucial. As it is well justified in the current literature, many children remain misdiagnosed, missing out on opportunities for effective interventions. Digital tools can contribute to assisting a clinician's assessment of identifying NDs. The concept of using serious games to enhance healthcare has received attention among a growing group of scientists, entrepreneurs, and clinicians.

Objective:

This study aims to define the main principles for detecting NDs.

Methods:

In this study 229 children participated in a serious game. The participants were Greek children of typical development aged 4–12 years. Children suffering from NDs or other neurological disorders or those on medications were excluded from this study. For the ethical issues of this study, parents of children provided a consent form prior to the study. For the study an innovative serious game named ‘Apsou’ was used to measure 18 primary domains featuring speech, language, psychomotor, cognitive, psychoemotional, and hearing abilities. These measurements were based on the child's performance on specific tasks, including gameplay, verbal responding, eye tracking, and impulsive heart rate reacting. All data collected was analyzed using descriptive statistics and Principal Component Analysis (PCA).

Results:

The five principal components, which accounted for about 80% of the data variability, from PCA and varimax rotation explained the 61.44% of the total variance. The results of this study showed the main theoretical principals which are essential for automated detection of Neurodevelopmental Disorders, include communication skills, speech and language development, vocal processing, cognitive skills & sensory functions, and visual spatial skills. The components found in this study are in accordance with the theoretical principles regarding typically developmental domains as they are described in other studies.

Conclusions:

The findings of this study suggest a comprehensive model which can also be a combination of models and techniques for understanding NDs. In extent, the outcomes provide us with information for the creation of machine learning applications for clinical decision-making, enabling highly accurate predictions and classifications for automated screening, diagnosis, prognosis, or intervention planning in NDs. Clinical Trial: 18435-15.05.2020 approved by the Research Ethics Committee of the University of Ioannina, Greece, https://ethics.ac.uoi.gr/