Diagnosing Epilepsy with Normal Interictal EEG Using Dynamic Network Models.

Area of Science:

• Neuroscience
• Medical Diagnostics
• Computational Biology

Background:

• Scalp electroencephalogram (EEG) is crucial for epilepsy diagnosis, but routine EEGs have limited diagnostic value.
• A significant percentage of epilepsy cases (20-30%) are misdiagnosed due to the low detection rate of interictal epileptiform discharges (IEDs) in routine EEGs.

Purpose of the Study:

• To develop and validate a novel method using EEG network properties to enhance the speed and accuracy of epilepsy diagnosis.
• To differentiate epilepsy from non-epileptic conditions, even when routine EEGs lack IEDs.

Main Methods:

• A multicenter study analyzed routine scalp EEGs from 218 patients with suspected epilepsy and normal initial EEGs.
• A logistic regression model, named EpiScalp, was trained using spectral and network-derived EEG features to classify epilepsy.
• The model was trained on 90% of the data and tested on the remaining 10%, with 10-fold cross-validation performed on the training set.

Main Results:

• EpiScalp demonstrated high diagnostic performance with an area under the curve (AUC) of 0.940.
• The tool achieved an accuracy of 0.904, a sensitivity of 0.835, and a specificity of 0.963 in classifying patients.

Conclusions:

• EpiScalp offers an accurate diagnostic aid from a single initial EEG, proving effective even in challenging epilepsy cases with normal initial EEGs.
• This approach may revolutionize epilepsy diagnosis by providing an objective measure of epilepsy likelihood from initially uninformative EEGs.