A Predictive Model Offor Attention Deficit Hyperactivity Disorder Based on Clinical Assessment Tools

Area of Science:

• Pediatric neurology
• Child psychology
• Behavioral science

Background:

• Current attention deficit hyperactivity disorder (ADHD) diagnosis relies on subjective parent reports, clinical observations, and varied assessment tools.
• Lack of standardized analysis methods for ADHD diagnostic tools necessitates reliance on clinician experience.
• Existing diagnostic methods for ADHD lack unified standards for interpreting results from rating scales and neuropsychological tests.

Purpose of the Study:

• To develop a standardized method for analyzing ADHD assessment results.
• To enhance the accuracy of attention deficit hyperactivity disorder (ADHD) diagnosis in children.
• To guide clinicians in interpreting data from various ADHD diagnostic tools.

Main Methods:

• Children with and without ADHD were assessed using parent rating scales (PSQ, CBCL) and a neuropsychological test (IVACPT).
• Statistical analysis (t-test with FDR correction) identified significant parameters differentiating ADHD from controls.
• A classification model (LibSVM) was built to predict ADHD using weighted parameters.

Main Results:

• 19 significant parameters were identified (16 from rating scales, 3 from neuropsychological tests).
• Combined rating scales and neuropsychological tests yielded higher classification accuracy (70.44%) than individual tools.
• Key predictive parameters included learning problems, hyperactivity/impulsivity, and activity capacity.

Conclusions:

• Multidimensional assessment is crucial for accurate attention deficit hyperactivity disorder (ADHD) diagnosis.
• Developing new assessment parameters based on physiological and psychological dimensions can further improve ADHD diagnosis.
• The developed predictive model offers potential for better understanding and optimizing ADHD treatment.