Beyond the double banana: improved recognition of temporal lobe seizures in long-term EEG
Ivana Rosenzweig1, András Fogarasi, Birger Johnsen
1*Department of Clinical Neurophysiology, Danish Epilepsy Centre, Dianalund, Denmark; †Department of Neuroimaging, Institute of Psychiatry, King's College London, London, United Kingdom; ‡Department of Neurology, Bethesda Children's Hospital, Budapest, Hungary; §Department of Clinical Neurophysiology, Aarhus University Hospital, Aarhus, Denmark; ‖Department of Clinical Neurophysiology, University of Copenhagen Glostrup Hospital, Glostrup, Denmark; ¶Research Department, BESA GmbH, Gräfelfing, Germany; #The EEG and Epilepsy Unit, Department of Neurology, Tel Aviv Medical Center, Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel; **Great Ormond Street Hospital for Children NHS Trust, Department of Clinical Neurophysiology, London, United Kingdom; ††Department of Clinical Neurophysiology, Rigshospitalet University Hospital, Copenhagen, Denmark; and ‡‡Dutch Epilepsy Clinics Foundation (S.E.I.N.), Department of Clinical Neurophysiology, Heemstede, The Netherlands.
Purpose:
To investigate whether extending the 10-20 array with 6 electrodes in the inferior temporal chain and constructing computed montages increases the diagnostic value of ictal EEG activity originating in the temporal lobe. In addition, the accuracy of computer-assisted spectral source analysis was investigated.
Methods:
Forty EEG samples were reviewed by 7 EEG experts in various montages (longitudinal and transversal bipolar, common average, source derivation, source montage, current source density, and reference-free montages) using 2 electrode arrays (10-20 and the extended one). Spectral source analysis used source montage to calculate density spectral array, defining the earliest oscillatory onset. From this, phase maps were calculated for localization. The reference standard was the decision of the multidisciplinary epilepsy surgery team on the seizure onset zone. Clinical performance was compared with the double banana (longitudinal bipolar montage, 10-20 array).
Results:
Adding the inferior temporal electrode chain, computed montages (reference free, common average, and source derivation), and voltage maps significantly increased the sensitivity. Phase maps had the highest sensitivity and identified ictal activity at earlier time-point than visual inspection. There was no significant difference concerning specificity.
Conclusions:
The findings advocate for the use of these digital EEG technology-derived analysis methods in clinical practice.
