Related Experiment Video
Updated: Jun 22, 2025
06:01
Semi-Automated Analysis of Peak Amplitude and Latency for Auditory Brainstem Response Waveforms Using R
Published on: December 9, 2022
2.5K
Performance Comparison of Convolutional Neural Network-Based Hearing Loss Classification Model Using Auditory
Jun Ma1, Seong Jun Choi2, Sungyeup Kim3
1Department of Software Convergence, Soonchunhyang University, Asan 31538, Republic of Korea.
Diagnostics (Basel, Switzerland)
|June 27, 2024
Summary
This study shows AlexNet, a deep learning model, can accurately classify hearing loss from auditory brainstem response (ABR) images with 95.93% accuracy, aiding in automated diagnosis.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Audiology
Background:
- Hearing loss diagnosis relies on auditory brainstem response (ABR) data.
- Automating ABR analysis can improve diagnostic efficiency.
Purpose of the Study:
- To evaluate Convolutional Neural Network (CNN) models for hearing loss classification using ABR images.
- To compare the performance of six CNN architectures: VGG16, VGG19, DenseNet121, DenseNet-201, AlexNet, and InceptionV3.
Main Methods:
- Utilized a dataset of 7990 preprocessed ABR images.
- Systematically tested six CNN models for classification accuracy.
- Performed comparative analysis based on accuracy and computational efficiency.
Main Results:
- AlexNet achieved the highest accuracy at 95.93% in classifying hearing loss.
- Demonstrated the effectiveness of deep learning models in ABR image analysis.
Conclusions:
- Deep learning, specifically AlexNet, shows significant potential for automated hearing loss diagnosis from ABR graphs.
- Further refinement of models can enhance clinical application and diagnostic accuracy.
