Employing Classification Techniques on SmartSpeech Biometric Data towards Identification of Neurodevelopmental Disorders

Early detection and evaluation of children at risk neurodevelopmental disorders and/or communication deficits is critical. While the current literature indicates a high prevalence disorders, many remain undiagnosed, resulting in missed opportunities for effective interventions that could have had greater impact if administered earlier. Clinicians face variety complications during disorders’ procedures must elevate their use digital tools to aid early efficiently. Artificial intelligence enables novelty taking decisions, classification, diagnosis. The research investigates efficacy various machine learning approaches on biometric SmartSpeech datasets. These datasets come from new innovative system includes serious game which gathers children’s responses specifically designed speech language activities manifestations, intending assist clinical disorders. were used by utilizing algorithms Radial Basis Function, Neural Network, Deep Learning Networks, variation Grammatical Evolution (GenClass). most significant results show improved accuracy (%) when using eye tracking dataset; more specifically: (i) class Disorder with GenClass (92.83%), (ii) Autism Spectrum Disorders Networks layer 4 (86.33%), (iii) Attention Deficit Hyperactivity (87.44%), (iv) Intellectual Disability (86.93%), (v) Specific (88.88%), (vi) Communication (88.70%). Overall, indicated be nearly top competitor, opening up additional probes future studies toward automatically classifying assisting assessments