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

Abstract

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

Understanding the brain with attention: A survey of transformers in brain sciences

“…The attention mechanism forms the core of Transformers. Compared to the limited windows in the RNN, [50] Gated Recurrent Unit (GRU), [56,57] and Long Short-Term Memory (LSTM), [58,59] the attention mechanism has a theoretically infinite window and computing space. In the attention module, the word vector passes through three fully connected layers to create the query (q), key (k), and value (v) vectors.…”

Multi-Transfer Learning Techniques for Detecting Auditory Brainstem Response

“…Deep learning models are particularly good at identifying difficult-to-identify inputs and detecting ABRs because they may learn by focusing on their commonalities [11][12][13]. In the field of deep learning, [14] used 614 subjects with an average age between 18 and 90 years, 348 men and 266 females, using a long shortterm memory (LSTM) model rich to 92.91% accuracy. With a different strategy, [15] offered a technique based on spectral feature extraction that would speed up detection without compromising accuracy.…”

Feed forward Neural Networks for Accurate Thyroid Detection in Healthcare

“…As a result, it's getting more and more important to be able to quickly and accurately spot pathology in ultrasound thyroid knobs. As of late, the use of counterfeit thinking advancement in medication has consistently grown, especially in the disciplines of imaging [3]- [5] and signal [6]. Developing a PC-aided mechanized thyroid indication framework by utilizing data from ultrasound images in the most effective manner is a significant area of flow study [7,8].…”