Related Experiment Videos
Computer recognition of brain stem auditory evoked potential wave V by a neural network
1Section of Medical Informatics, University of Pittsburgh, PA 15261.
Proceedings. Symposium on Computer Applications in Medical Care
|January 1, 1991
Summary
A neural network model accurately identifies wave V peaks in brain stem auditory evoked potential (BAEP) tests. This automated approach achieves 85% accuracy, aiding in objective analysis of auditory pathway function.
Area of Science:
- Computational Neuroscience
- Auditory Neurophysiology
Background:
- Brain Stem Auditory Evoked Potential (BAEP) tests are crucial for assessing auditory pathway function.
- Identifying specific waveforms, like wave V, is critical for accurate diagnosis.
- Manual interpretation of BAEPs can be subjective and time-consuming.
Purpose of the Study:
- To develop and evaluate a connectionist model using a neural network simulator for automated recognition of the wave V peak in BAEPs.
- To improve the objectivity and efficiency of BAEP analysis.
Main Methods:
- Digitized and normalized ipsilateral and contralateral BAEP waveforms.
- Developed a neural network with two architectures (40 and 16 hidden units) using back-propagation.
- Trained the network on 50 BAEP recordings and tested on an independent set.
Main Results:
- The best-performing network achieved 85% accuracy in identifying wave V peaks on an independent test set.
- The model demonstrated robust performance after 60 epochs (3000 presentations).
Conclusions:
- A neural network model can reliably identify the wave V peak in BAEP tests.
- This automated method offers a promising tool for objective and efficient analysis of auditory pathway function.