Accurate Identification of EEG Recordings with Interictal Epileptiform Discharges Using a Hybrid Approach: Artificial Intelligence Supervised by Human Experts

Overview

Journal Epilepsia

Specialty Neurology

Date 2022 Feb 20

PMID 35184276

Authors

Mustafa Aykut Kural

Jin Jing

Franz Furbass

Hannes Perko

Erisela Qerama

Birger Johnsen

Steffen Fuchs

M Brandon Westover

Sandor Beniczky

Affiliations

Soon will be listed here.

Abstract

Objective: To evaluate the diagnostic performance of artificial intelligence (AI)-based algorithms for identifying the presence of interictal epileptiform discharges (IEDs) in routine (20-min) electroencephalography (EEG) recordings.

Methods: We evaluated two approaches: a fully automated one and a hybrid approach, where three human raters applied an operational IED definition to assess the automated detections grouped into clusters by the algorithms. We used three previously developed AI algorithms: Encevis, SpikeNet, and Persyst. The diagnostic gold standard (epilepsy or not) was derived from video-EEG recordings of patients' habitual clinical episodes. We compared the algorithms with the gold standard at the recording level (epileptic or not). The independent validation data set (not used for training) consisted of 20-min EEG recordings containing sharp transients (epileptiform or not) from 60 patients: 30 with epilepsy (with a total of 340 IEDs) and 30 with nonepileptic paroxysmal events. We compared sensitivity, specificity, overall accuracy, and the review time-burden of the fully automated and hybrid approaches, with the conventional visual assessment of the whole recordings, based solely on unrestricted expert opinion.

Results: For all three AI algorithms, the specificity of the fully automated approach was too low for clinical implementation (16.67%; 63.33%; 3.33%), despite the high sensitivity (96.67%; 66.67%; 100.00%). Using the hybrid approach significantly increased the specificity (93.33%; 96.67%; 96.67%) with good sensitivity (93.33%; 56.67%; 76.67%). The overall accuracy of the hybrid methods (93.33%; 76.67%; 86.67%) was similar to the conventional visual assessment of the whole recordings (83.33%; 95% confidence interval [CI]: 71.48-91.70%; p > .5), yet the time-burden of review was significantly lower (p < .001).

Significance: The hybrid approach, where human raters apply the operational IED criteria to automated detections of AI-based algorithms, has high specificity, good sensitivity, and overall accuracy similar to conventional EEG reading, with a significantly lower time-burden. The hybrid approach is accurate and suitable for clinical implementation.

Citing Articles

Can ChatGPT 4.0 Diagnose Epilepsy? A Study on Artificial Intelligence's Diagnostic Capabilities.

Brigo F, Broggi S, Leuci E, Turcato G, Zaboli A J Clin Med. 2025; 14(2).

PMID: 39860325 PMC: 11765833. DOI: 10.3390/jcm14020322.

Artificial Intelligence in Pediatric Epilepsy Detection: Balancing Effectiveness With Ethical Considerations for Welfare.

Mourid M, Irfan H, Oduoye M Health Sci Rep. 2025; 8(1):e70372.

PMID: 39846037 PMC: 11751886. DOI: 10.1002/hsr2.70372.

Expert level of detection of interictal discharges with a deep neural network.

Tjepkema-Cloostermans M, Tannemaat M, Wieske L, van Rootselaar A, Stunnenberg B, Keijzer H Epilepsia. 2024; 66(1):184-194.

PMID: 39530797 PMC: 11742546. DOI: 10.1111/epi.18164.

AI and EEG: Should EEGers RIP (Rest in Peace)?.

King-Stephens D Epilepsy Curr. 2024; 24(2):111-113.

PMID: 39280053 PMC: 11394416. DOI: 10.1177/15357597241227085.

Machine-learning for the prediction of one-year seizure recurrence based on routine electroencephalography.

Lemoine E, Toffa D, Pelletier-Mc Duff G, Xu A, Jemel M, Tessier J Sci Rep. 2023; 13(1):12650.

PMID: 37542101 PMC: 10403587. DOI: 10.1038/s41598-023-39799-8.

References

Gotman J, Ives J, Gloor P . Automatic recognition of inter-ictal epileptic activity in prolonged EEG recordings. Electroencephalogr Clin Neurophysiol. 1979; 46(5):510-20. DOI: 10.1016/0013-4694(79)90004-x. View

da Silva Lourenco C, Tjepkema-Cloostermans M, van Putten M . Machine learning for detection of interictal epileptiform discharges. Clin Neurophysiol. 2021; 132(7):1433-1443. DOI: 10.1016/j.clinph.2021.02.403. View

V K, Rajagopalan S, Bhardwaj S, Panda R, Reddam V, Ganne C . Machine learning detects EEG microstate alterations in patients living with temporal lobe epilepsy. Seizure. 2018; 61:8-13. DOI: 10.1016/j.seizure.2018.07.007. View

Basiri R, Shariatzadeh A, Wiebe S, Aghakhani Y . Focal epilepsy without interictal spikes on scalp EEG: A common finding of uncertain significance. Epilepsy Res. 2018; 150:1-6. DOI: 10.1016/j.eplepsyres.2018.12.009. View

Jing J, Sun H, Kim J, Herlopian A, Karakis I, Ng M . Development of Expert-Level Automated Detection of Epileptiform Discharges During Electroencephalogram Interpretation. JAMA Neurol. 2019; 77(1):103-108. PMC: 6806668. DOI: 10.1001/jamaneurol.2019.3485. View

Pillai J, Sperling M . Interictal EEG and the diagnosis of epilepsy. Epilepsia. 2006; 47 Suppl 1:14-22. DOI: 10.1111/j.1528-1167.2006.00654.x. View

Aanestad E, Gilhus N, Brogger J . Interictal epileptiform discharges vary across age groups. Clin Neurophysiol. 2019; 131(1):25-33. DOI: 10.1016/j.clinph.2019.09.017. View

McBride A, Shih T, Hirsch L . Video-EEG monitoring in the elderly: a review of 94 patients. Epilepsia. 2002; 43(2):165-9. DOI: 10.1046/j.1528-1157.2002.24401.x. View

Benbadis S, Tatum W . Overintepretation of EEGs and misdiagnosis of epilepsy. J Clin Neurophysiol. 2003; 20(1):42-4. DOI: 10.1097/00004691-200302000-00005. View

10.

Uldall P, Alving J, Hansen L, Kibaek M, Buchholt J . The misdiagnosis of epilepsy in children admitted to a tertiary epilepsy centre with paroxysmal events. Arch Dis Child. 2006; 91(3):219-21. PMC: 2065931. DOI: 10.1136/adc.2004.064477. View

11.

LaFrance Jr W, Benbadis S . Avoiding the costs of unrecognized psychological nonepileptic seizures. Neurology. 2006; 66(11):1620-1. DOI: 10.1212/01.wnl.0000224953.94807.be. View

12.

Kural M, Tezcan Aydemir S, Levent H, Olmez B, Ozer I, Vlachou M . The operational definition of epileptiform discharges significantly improves diagnostic accuracy and inter-rater agreement of trainees in EEG reading. Epileptic Disord. 2021; 24(2):353-358. DOI: 10.1684/epd.2021.1395. View

13.

Chadwick D . Diagnosis of epilepsy. Lancet. 1990; 336(8710):291-5. DOI: 10.1016/0140-6736(90)91815-r. View

14.

Baldini S, Pittau F, Birot G, Rochas V, Tomescu M, Vulliemoz S . Detection of epileptic activity in presumably normal EEG. Brain Commun. 2020; 2(2):fcaa104. PMC: 7566453. DOI: 10.1093/braincomms/fcaa104. View

15.

Kural M, Duez L, Sejer Hansen V, Larsson P, Rampp S, Schulz R . Criteria for defining interictal epileptiform discharges in EEG: A clinical validation study. Neurology. 2020; 94(20):e2139-e2147. PMC: 7526669. DOI: 10.1212/WNL.0000000000009439. View

16.

Benbadis S, Lin K . Errors in EEG interpretation and misdiagnosis of epilepsy. Which EEG patterns are overread?. Eur Neurol. 2008; 59(5):267-71. DOI: 10.1159/000115641. View

17.

Furbass F, Kural M, Gritsch G, Hartmann M, Kluge T, Beniczky S . An artificial intelligence-based EEG algorithm for detection of epileptiform EEG discharges: Validation against the diagnostic gold standard. Clin Neurophysiol. 2020; 131(6):1174-1179. DOI: 10.1016/j.clinph.2020.02.032. View

18.

Ren S, He K, Girshick R, Sun J . Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks. IEEE Trans Pattern Anal Mach Intell. 2016; 39(6):1137-1149. DOI: 10.1109/TPAMI.2016.2577031. View

19.

Gotman J, Gloor P, Ray W . A quantitative comparison of traditional reading of the EEG and interpretation of computer-extracted features in patients with supratentorial brain lesions. Electroencephalogr Clin Neurophysiol. 1975; 38(6):623-39. DOI: 10.1016/0013-4694(75)90163-7. View

20.

Obeid I, Picone J . The Temple University Hospital EEG Data Corpus. Front Neurosci. 2016; 10:196. PMC: 4865520. DOI: 10.3389/fnins.2016.00196. View