CLAS-hdEEG: high-density EEG software platform for real-time delta wave detection and closed-loop auditory stimulation

Objective.Closed-loop auditory stimulation (CLAS) can enhance slow oscillations during non-rapid eye movement sleep; however, its broader use has been limited by technical and temporal constraints. We introduce and validateCLAS-hdEEG, a high-density electroencephalographic (hd-EEG) software platform integrated within the Magstim EGI environment, capable of real-time delta wave detection and phase-targeted auditory stimulation with low detection-to-stimulation latency, achieved through event-based detection of delta wave extrema.Approach.The CLAS-hdEEG software continuously streams 128-channel EEG data, applies an online moving average filter to Fz for real-time delta wave detection, and triggers brief pink noise bursts at the detected peak or trough of ongoing oscillations. Detection thresholds were based on established amplitude crossings, a minimum peak-to-peak difference of75μV, and duration limits of 160-1700 ms. Synchronization between neural detection and stimulus onset was verified using digital input hardware (DIN) markers recorded by the EGI acquisition system. The system was validated in 14 healthy participants during N3 sleep under peak, trough, and sham stimulation conditions. Phase-targeting success was defined as auditory bursts occurring within predefined 90∘phase windows of the oscillation (0∘-90∘for in-phase blocks and 180∘-270∘for anti-phase blocks).Main results.Across all sessions, auditory stimuli were delivered with a mean detection-to-stimulation latency of20.03±0.5msand an overall phase-targeting success rate of 0.913, indicating that 91.3% of stimuli occurred within the predefined 90∘target phase windows. Online and offline detections showed strong agreement (F1≈0.80), confirming that real-time processing preserved the fidelity of offline algorithms.Significance.This study establishes the technical and temporal validity of CLAS-hdEEG, the first hd-EEG software platform for CLAS within the EGI environment. By combining low detection-to-stimulation latency with high-density spatial sampling, this framework provides a robust and extensible platform for investigating how auditory stimulation interacts with large-scale neural dynamics across diverse states of consciousness.