Robust algorithm for objective hearing screening of newborns using Automated Auditory Brain-stem Response

At present, 2 to 4 out of every 1000 births are affected with hearing impairments. In enforcing mandatory hearing screening for neonates, robust algorithms are required to make the process fast and efficient. In this paper, we present a novel algorithm to automate objective hearing screening using auditory brainstem response (referred to as ABR). An effective stimulus delivery mechanism, an efficient signal processing algorithm and an automatic peak detection algorithm are consolidated to reduce test time while maintaining accuracy. Simultaneously, compensation for ambient noise levels in clinical environments as opposed to sound proof environments are also considered. The Chirp Stimulus, Empirical Mode Decomposition and High Curvature Detection have been rigorously verified by MATLAB® simulations for data collected by ADInstruments® PowerLab. The algorithms utilized in screening reduce testing time to 8% of the gold standard hearing screening procedure, i.e. Click Stimulus based synchronized averaging. Moreover, the resultant ABR waveforms acquired are de-noised making them comparatively convenient to diagnose. The findings highlighted in the paper provide a superior methodology for robust and accurate newborn hearing screening compared to existing gold standard procedure.