For pre-surgical evaluation of patients with focal seizures, stereoelectroencephalographic (SEEG) signals are routinely recorded to identify the epileptogenic zone network (EZN). This network consists of distant brain regions involved in seizure initiation. However, the mechanisms underlying typical SEEG patterns that occur during the transition from interictal to ictal activity in distant brain nodes of the EZN remain elusive.We analyzed SEEG signals recorded from the EZN (involving three distant neocortical brain regions) in 10 patients during the transition from interictal to ictal activity. This transition consisted of a sequence of periods during which SEEG signals from remote sites exhibited stereotypical patterns of activity: sustained preictal spiking activity preceding a fast activity occurring at seizure onset, followed by the ictal activity. Spectral content and nonlinear correlation of SEEG signals were analyzed. In addition, we developed a novel neuro-inspired computational model consisting of two bidirectionally coupled neuronal populations, each represented by a layered neural mass model adapted to neocortical architectonics.The proposed model captured the essential characteristics of the patient signals, including the quasi-synchronous onset of rapid discharges in distant interconnected epileptogenic zones. Statistical analysis confirmed the dynamic correlation/de-decorrelation pattern in both patient and simulated signals. Results suggest that bidirectional connections of remote neuronal populations (from pyramidal cells to VIP+ interneurons) play a key role in this transition, while PV+ interneurons intervene in the emergence of rapid discharges at seizure onset. This study provides significant insight into the abnormal changes in glutamatergic and GABAergic dynamics that occur during the transition to seizures.