Fast Automated Auditory Brainstem Response (AABR) using new non linear Maximum Length Sequence (MLS) reconstruction and automated signal detection

The automated auditory brainstem response (AABR) is one of the successful screening tools used in Universal Newborn Hearing Screening (UNHS) but often hindered by unacceptably long test times. This present study has developed and used a new non linear Maximum Length Sequence (MLS) reconstruction algorithm which aims to account for the non linearity aspect in the auditory system and to reduce the time required to complete an AABR assessment for UNHS. 492 AABR waveforms from 126 neonates who underwent UNHS were included in the final data analysis. Those AABRs were acquired at 35 dBnHL and at six different stimulus repetition rates (33, 90, 125, 250 and 418 clicks per seconds). The recording was made using standard clicks (at 33 and 90 cps) and Maximum Length Sequence (MLS) with linear and new non linear MLS reconstruction (at 90, 125, 250 and 418 cps). The time to AABR detection for all stimulus repetition rates were determined using variance ratio analysis (Fsp at 99% confidence level). The result showed that all MLS median detection time was significantly different than the standard click at 33 cps (Mann Whitney U test, p< 0.001), linear and non linear MLS at 418 cps median test time was significantly different than the standard click at its maximum rate 90 cps (Mann Whitney U test, p< 0.001) and MLS linear and MLS non linear reconstruction median test time was statistically different (Friedman test, p< 0.003). The best median time to detection was 3.59 s provided by the MLS non linear reconstruction stimulus at 418 cps. This study concludes that the combination MLS non linear reconstruction and Fsp holds significant promise to reduce UNHS test time