Accurate identification of EEG recordings with interictal epileptiform discharges using a hybrid approach: Artificial intelligence supervised by human experts
Kural MA, Jing J, Fürbass F, Perko H, Qerama E, Johnsen B, Fuchs S, Westover MB, Beniczky S.
Epilepsia. 2022;00:1–10.
https://doi.org/10.1111/epi.17206
Hot Off The Press – This scientific paper by Aykut Kural et al. 2022 compared the #precision of three different artificial intelligence based algorithms (SpikeNet, Encevis and Persyst) for the detection of interictal epileptiform discharges (#IEDs) in conventional electroencephalography (#EEG) recordings. The use of a hybrid system, manual assessment of automatically detected IEDs, and a fully automated method was compared to gold-standard video-EEG.
The hybrid system was associated with adequate #sensitivity and a significantly increased #specificity, as compared to the fully automated technique, for all three algorithms. Furthermore, the #accuracy of the hybrid system was comparable to the standardized evaluation of video-EEG recordings by clinical experts. However, the use of the hybrid method significantly lowered the physicians workload.
