Automatic phase reversal detection in routine EEG.

Area of Science:

• Neuroscience
• Medical Technology
• Signal Processing

Background:

• Electroencephalograph (EEG) is a crucial, safe, and accessible tool for assessing brain function and diagnosing neurological disorders.
• Analyzing long-term EEG data, essential for studying conditions like epilepsy, is time-consuming and challenging for human experts.
• Phase reversal (PR) in EEG is a significant indicator of neurological disorders.

Purpose of the Study:

• To develop and validate an automated method for detecting phase reversal (PR) in long-term electroencephalograph (EEG) recordings.
• To assess the efficiency and accuracy of the automated PR detection method compared to expert neurologist interpretations.
• To accelerate the interpretation process of complex neurological data from long-term EEG.

Main Methods:

• An automated phase reversal (PR) determination technique was developed for analyzing long-term electroencephalograph (EEG) data.
• The method was applied to retrospective pathological EEG recordings from Selcuk University Hospital (SUH) and Boston Children's Hospital (BCH) datasets.
• Performance was evaluated by comparing the automated PR detection results with those identified by specialist neurologists.

Main Results:

• The automated method achieved classification success rates of 83.22% and 85.19% in the SUH dataset.
• For the BCH dataset, the highest classification success was 75% and the lowest was 58.33%.
• Overall classification success was 84.20% for SUH and 66.7% for BCH, demonstrating the method's potential.

Conclusions:

• The developed automated PR detection method can significantly expedite the interpretation of long-term EEG recordings.
• This technique aids in identifying indicators of neurological disorders, supporting clinical evaluation and diagnosis.
• Automating PR detection enhances the efficiency of neuroscientific research and clinical practice.