Auditory brainstem evoked potentials peak identification by finite impulse response digital filters

Area of Science:

• Neuroscience
• Signal Processing
• Biomedical Engineering

Background:

• Auditory brainstem evoked potentials (ABEP) are crucial for assessing auditory pathway function.
• Differentiating ABEP components is essential for accurate diagnosis and analysis.
• Traditional methods for ABEP component identification can be time-consuming and subjective.

Purpose of the Study:

• To develop and validate a digital filtering technique for automatic identification of ABEP components.
• To differentiate slow, medium, and fast frequency bands within ABEP signals.
• To establish a reliable method for labeling ABEP peaks I, III, and V.

Main Methods:

• Application of linear phase finite impulse response (FIR) digital filters to ABEP signals.
• Analysis of ABEP power spectrum to define frequency bands: <240 Hz (slow), 240-483 Hz (medium), >500 Hz (fast).
• Correlation of filtered components with known ABEP peak latencies (I, III, V) and validation on normal and patient data.

Main Results:

• Specific frequency bands were identified within ABEP signals.
• Medium frequency components corresponded to peaks I, III, and V.
• A 'pedestal' in the slow frequency band coincided with peak V, enabling sequential labeling of peaks V, III, and I.
• The automated method showed high reliability and agreement with manual analysis when the 'pedestal' was present.

Conclusions:

• Linear phase FIR filtering provides an effective method for differentiating ABEP components.
• The developed automated peak identification procedure is reliable and accurate, especially in the presence of a 'pedestal'.
• This technique offers a valuable tool for objective and efficient analysis of ABEP data in clinical and research settings.