Clinical practice of EEG revisited: improved spike identification, localization, and characterization

2021.11.02 | Graduate School of Health

On Thursday 11 November at 13:00, Mustafa Aykut Kural defends his dissertation entitled "Clinical practice of EEG revisited: improved spike identification, localization, and characterization".

In our studies, we hypothesized that 1) The operational definition of IEDs, recently issued by the International Federation of Clinical Neurophysiology (IFCN) has high diagnostic accuracy; 2) The specific combination of IFCN criteria having high inter-rater agreement (optimized IFCN criteria) gives high diagnostic accuracy of identifying IEDs; 3) The repetition of similar morphological criteria of identifying IEDs enables better pattern recognition defining IEDs; 4) Young trainees can learn and apply the previously validated criteria for identifying IEDs, and improve their diagnostic performance; 5) Using source montages used for identification of IEDS can contribute to map the cortical generators of widely distributed EEG abnormalities, such as the triphasic waves (TWs).

In the first study, seven raters applied the IFCN criteria and analysis in source space, in 100 EEG epochs. In the second study, three raters applied the combinations of IFCN criteria with high interrater agreement in 70 EEG epoch. In the third study, three raters scored the repetition rate of these visual features (criteria) in a balanced dataset of 20 minutes routine EEG recording for each patient. In the fourth study, seven junior raters scored 70 EEG samples: first without being familiar with the IFCN criteria, and second, after a training session, learned and applied the operational definition of IEDs. In the fifth study, we determined the cortical generators of triphasic waves using different source localization methods.

The validation studies show that high diagnostic accuracy (specificity over 95%; sensitivity: 81-89%) can be achieved using a threshold of five IFCN criteria, using the specific combination of the specific combination of IFCN criteria. We need more repetition for less typical IEDs. Our trainees significantly improved their diagnostic accuracy after implementing the IFCN criteria. Source montages helped identifying the bilateral, extended cortical network generating the triphasic potentials.

The summary is written by the PhD student.     

The defence is public and takes place at the Clinical Neurophysiology Department, Aarhus University Hospital, Palle Juul-Jensens Boulevard 165, 2nd floor, J209, 8200 Aarhus N. Please read full press release for more information. 

Contact:

PhD student Mustafa Aykut Kural

Mail: muskur@rm.dk

Phone: (+45) 24811885

Read full press release