A Prospective Study on Electronystagmography (ENG) to Detect Vestibular Disorders Using Simplified GUI

Abstract

Electronystagmography (ENG) refers to the clinical test that monitors the involuntary movements of the eye due to an effect called Nystagmus. The procedure is also used to detect disorders of inner ear that are responsible for vertigo and imbalance. In clinical practice, various tests such as caloric test, calibration test, gaze test, pendulum tracking test, opto-kinetics test, positional test and Rhomberg's test were generally performed to diagnose the vestibular dysfunction. Such procedure were found to be very time consuming and cumbersome for the patients. This prospective study investigates the screening of the vestibular dysfunction by integrating the readymade hardware unit with a compact software tool to reduce the clinical test duration. Electrooculography (EOG) based physiological signal is used to record the eye movements during ENG. The proposed research study aims in developing a simplified procedure towards the analysis of ENG signals. The proposed methodology not only helps in reducing the time for performing such procedures but also helps in reducing the discomforts to the subject. The procedure showed the effectiveness in terms of recognizing the required biomarkers for various activities. For the pilot study, 25 healthy male and 25 female test subjects in the age group of 19-21 years were considered. Experimental set up was arranged to record EOG signals with all activities with time duration of 10 minutes. Two features namely spike rhythmicity and auto regressive feature using Burg's method were extracted from the recorded raw signal. A typical Elman feedback neural network was introduced to perform the binary classification to identify the pattern, (Nystagmus left ear-Nystagmus right ear, Vertical-Horizontal and Gaze-Pendulum) respectively. An overall classification accuracy (CA) of 93.8% was achieved that confirms the effectiveness of the test required to detect nystagmus. The system needs to be validated with large experimental data to confirm the accuracy for clinical screening and validation.