ehiscence and Evaluation of Surgical Outcome based on Electrocochleography ECochG Waveforms Background Superior Canal Dehiscence SCD abnormal opening in the top of the balance canal Electrocochleography ID: 916907
Download Presentation The PPT/PDF document "Diagnosis of Superior Canal D" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Diagnosis of Superior Canal Dehiscence and Evaluation of Surgical Outcome based on Electrocochleography (ECochG) Waveforms
Background
Superior Canal Dehiscence (SCD): abnormal opening in the top of the balance canal
Electrocochleography (
ECochG
) is a
method for recording the electrical
potential of the cochlea
ECochG
can provide measurement of:
Stimulus-related cochlear Potentials (SP)
Compound Action Potential (AP) of
the auditory nerve
Slide2Diagnosis of Superior Canal Dehiscence and Evaluation of Surgical Outcome based on Electrocochleography (ECochG) Waveforms
Background
The AP/SP ratio is a possible predictor of the presence of an abnormal opening
It could also be a method to determine if a proper closure of the opening has occurred during surgery
Slide3What Students Will DoTraining and evaluating a Neural Network (NN) using an existing set of labeled ECOG data to:Diagnose indicate existence of
SCD condition in patients based on ECOG data in effected ear
Will use contralateral ear as normal control
Evaluate surgical outcome
Developing a binary classifier using NNs:
Input: two relevant parameters (
AP and SP
) specified by a
technician
Output: existence of the condition
Developing a deep NN by using the actual waveforms as the input to extract existing patterns
to eliminate subjective evaluation of AP and SP values
Evaluating and comparing performance of the two NNs
Slide4Deliverables:Minimum: Training and evaluating a NN based on the subjective AP and SP values as the inputExpected: Training and evaluating a deep NN based on the full E-COG waveforms as the input
Maximum: Comparing performance of the two networks in diagnosis of the condition and evaluation of success rate of the surgical procedure
Group Size:
1-2
Skills:
Programming skills such as Python/MATLAB; familiarity with libraries such as
PyTorch
and
Tensorflow
is a plus
Knowledge of signal processing, NNs and deep learning algorithms
Mentors:
Dr. Mahya Shahbazi, Dr. Russell Taylor,
Dr. Francis Creighton, Dr. Chris
Razavi
, Dr. Deepa
Galaiya
Contact:
m
ahya.sh@jhu.edu, rht@jhu.edu,
francis.creighton@jhmi.edu,
crazavi1@jhmi.edu
, gdeepa1@jhmi.edu