Volta

NeuroSpectNet: A Novel Architecture for Early Stage Detection of Executive Dysfunction From Resting-State EEG

We present NeuroSpectNet, a novel architecture for the early-stage detection of executive dysfunction from resting-state electroencephalography. The proposed approach is based on power spectral density representations and introduces frequency-domain regularization to support robust, subject-independent learning under minimal preprocessing. NeuroSpectNet is evaluated against established neuropsychological reference measures of executive functioning using Trail Making Test Parts A and B derived from the LEMON dataset. In clinical screening contexts, missed cases (impaired individuals incorrectly classified as unimpaired) represent the most critical error. Across the separately trained TMT-A and TMT-B models, NeuroSpectNet achieves a mean accuracy of 78.04%. In a separate, comparable difference-score analysis, NeuroSpectNet achieves a statistically well-supported reduction in those false-negative classifications of 34.9% compared to prior work, which corresponds to a …

Reference & gravity metrics

Secondary-market trade history

References (0)

Cited by (1)