Principles underlying real-time nystagmus analysis of horizontal and vertical eye movements recorded with electro-, infra-red-, or video-oculographic techniques.

New methods for separating fast and slow phases of human nystagmus in real time are presented that are adaptive to the time-dependent noise properties of the input eye movement signal and therefore applicable to different recording techniques and directions. The methods employ a statistical filter technique to track slow-phase eye movements, uninfluenced by fast phases and blink artifacts, and fuzzy-logic techniques to identify fast-phase eye movements. Because these two techniques are decoupled from one another, highly accurate phase separation and slow-phase velocity profiles are achieved. In addition, the tracking of the variance of slow-phase and fast-phase eye movement recording permits a quality control of the analysis for different recording techniques and a variety of ocular nystagmus responses. Because blinks impose different eye velocity profiles on the recordings, depending on the type of recording technique and direction (horizontal, vertical), blink detection and its effect on fast-phase amplitude must be individually adjusted to each recording technique. Results are illustrated in the context of simultaneously recorded video-oculographic, infra-red, and electro-oculographic recordings of vestibulo-ocular reflex and optokinetic reflex responses causing horizontal or vertical eye movements.