Automatic Recognition of Auditory Brainstem Response Characteristic Waveform Based on Bidirectional Long Short-Term Memory

Background: Auditory brainstem response (ABR) testing is an invasive electrophysiological auditory function test. Its waveforms and threshold can reflect functional changes in the centers are widely used clinic to diagnose dysfunction hearing. However, identifying its mainly dependent on manual recognition by experimental persons, which could be primarily influenced individual experiences. This also a heavy job clinical practice. Methods: In this work, human ABR was recorded. First, binarization created mark 1,024 sampling points accordingly. The selected characteristic area of data 0–8 ms. marking enlarged expand feature information reduce error. Second, bidirectional long short-term memory (BiLSTM) network structure established improve relevance points, point classifier obtained training. Finally, through thresholding. Results: specific structure, related parameters, effect, noise resistance were explored 614 sets data. results show that average detection time for each 0.05 s, accuracy reached 92.91%. Discussion: study proposed automatic using BiLSTM-based machine learning technique. demonstrated methods recording help doctors making diagnosis, suggesting method has potential future.