A dual Distortion Product Otoacoustic Emission (DPOAE) and Auditory Brainstem Response (ABR) automated newborn hearing screener is proposed. This application focuses on new noise cancellation techniques for DPOAE acquisition in order to increase the reliability and testing bandwidth below 1500 Hz. Currently available systems provide limited noise rejection capabilities which substantially increase testing time by increasing the number of sweeps necessary. The proposed noise reduction techniques utilize a primary microphone and two additional microphones to acquire internal and external noise reference signals. The proposed techniques include: Adaptive Noise Cancellation (ANC), Spectral Enhancement Averaging (SEA) and Adaptive Wiener Filtering (AWF). Each of these methods will be tested with a full complement of noise sources and each will be compared to standard averaging (5A) with artifact rejection. One set of single sweep data will be collected from 12 newborns in order to compare directly the various processing schemes. Phase I testing will focus on 5 frequencies, 0.5, 1, 1.5, 4 and 8 kHz. The performance of the processing strategies will be evaluated using ANOVA techniques using as dependent variables the signal-to-noise ratio (SNR) and number of sweeps required to reach an SNR level.
