Visual and semi-automatic non-invasive detection of interictal fast ripples: A potential biomarker of epilepsy in children with tuberous sclerosis complex

Abstract

Objectives: We aim to establish that interictal fast ripples (FR; 250-500 Hz) are detectable on scalp EEG, and to investigate their association to epilepsy.

Methods: Scalp EEG recordings of a subset of children with tuberous sclerosis complex (TSC)-associated epilepsy from two large multicenter observational TSC studies were analyzed and compared to control children without epilepsy or any other brain-based diagnoses. FR were identified both by human visual review and compared with semi-automated review utilizing a deep learning-based FR detector.

Results: Seven out of 7 children with TSC-associated epilepsy had scalp FR compared to 0 out of 4 children in the control group (p = 0.003). The automatic detector has a sensitivity of 98% and false positive rate with average of 11.2 false positives per minute.

Conclusions: Non-invasive detection of interictal scalp FR was feasible, by both visual and semi-automatic detection. Interictal scalp FR occurred exclusively in children with TSC-associated epilepsy and were absent in controls without epilepsy. The proposed detector achieves high sensitivity of FR detection; however, expert review of the results to reduce false positives is advised.

Significance: Interictal FR are detectable on scalp EEG and may potentially serve as a biomarker of epilepsy in children with TSC.

Keywords: Epilepsy; Fast ripples; High frequency oscillations; Tuberous sclerosis complex.