Interictal epileptiform discharge and subclinical burst analysis-based pediatric epilepsy seizure severity assessment using CFQFL

Abstract The identification of the seizure burden is facilitated by the Pediatric Epilepsy Seizure (PES) severity analysis, thus enhancing the long-term neurological outcomes. Nevertheless, the existing works overlooked the Interictal Epileptiform Discharge (IED) and subclinical burst for risk scoring, thereby leading to delayed medical interventions. Hence, this paper proposes a novel Spike Frequency Index (SFI) and Burst Rhythmic Index (BRI)-based risk scoring and Cantor Function-centric Quantum Fuzzy Logic (CFQFL)-based PES severity evaluation. The proposed system collects and pre-processes the ElectroEncephaloGram (EEG) signals for removing the artifacts and improving the signal quality, followed by data augmentation and signal decomposition. Then, the time–frequency and signal energy analysis is carried out. Further, the features are extracted. Moreover, by utilizing Renyi Scaled Sine-Hyperbolic Transfer Learning-centric Long Short-Term Memory (RSSinHTL-LSTM), the PES is predicted. Also, the explainability is provided for the classified outcome. Lastly, PES severity is evaluated using CFQFL. According to the experimental results, the proposed classifier attained an accuracy of 98.97%, precision of 98.86%, recall of 98.79%, and F-Measure of 98.56% in PES detection. In addition, severity assessment was enhanced by the CFQFL model with a fuzzification time of 2561 ms, thus outperforming traditional models. Overall, precise PES detection and effective risk scoring were facilitated by the proposed framework, thus supporting clinical intervention in pediatric epilepsy.