Discrimination accuracy per tested methodology.

The accuracy achieved by the TFFO methodology is significantly better than other reaction latency related implementations, proving the added value of the analysis beyond latency differences between healthy subjects and schizophrenia patients (P-Value<1%). Moreover, the results obtained did not include any false positives, i.e., no healthy subject was classified as a schizophrenia patient. Therefore, it is desirable in such problems to have high specificity rather than high sensitivity. Generally, frontal area EEG input correlation to cognitive-related features and mental disorders is evident. Recent work in patients with schizophrenia [13;22;24] and Parkinson’s disease [25] has shown high input variance in that segment. Such evident correlations can explain the obtained results.
Discrimination accuracy per tested methodology.