Classification of passive auditory event-related potentials using discriminant analysis and self-organizing feature...

Area of Science:

• Neuroscience
• Auditory Neuroscience
• Computational Neuroscience

Background:

• Central Auditory Processing Disorder (CAPD) affects auditory perception.
• Event-related potentials (ERPs) reflect neural activity in response to auditory stimuli.
• Accurate classification of CAPD is crucial for effective intervention.

Purpose of the Study:

• To evaluate the efficacy of Discriminant Analysis (DA) and Self-Organizing Feature Maps (SOFM) for classifying auditory ERPs.
• To determine if SOFM can reliably identify children with varying degrees of auditory perception deficits, indicative of CAPD.
• To investigate the nature of auditory processing deficits in CAPD.

Main Methods:

• Collected auditory ERPs (P1, N1, P2, N2) from children with severe/marked auditory perception deficits and controls.
• Selected 18 ERP amplitude parameters for statistical analysis.
• Trained and compared DA methods with SOFM configurations for classification.
• Validated the trained SOFM on an independent cohort of 37 subjects.

Main Results:

• SOFM demonstrated superior classification performance compared to DA methods.
• The validated SOFM achieved 96% accuracy in classifying subjects with behavioral auditory perception deficits.
• Results support the existence of non-peripheral auditory processing deficits within CAPD.

Conclusions:

• SOFM is a reliable and effective tool for classifying auditory ERPs in the context of CAPD.
• The findings confirm that CAPD involves non-peripheral auditory processing impairments.
• This classification approach aids in understanding and diagnosing central auditory processing disorders.