gms | German Medical Science

Fourth International Symposium and Workshops: Objective Measures in Cochlear Implants

Medical University of Hannover

01.06. bis 04.06.2005, Hannover

Objective classification of responses obtained with Neural Response Imaging

Meeting Abstract

Search Medline for

  • corresponding author L. Litvak - Advanced Bionics: Corporation, Sylmar, Ca, USA
  • G. Emadi - Advanced Bionics: Corporation, Sylmar, Ca, USA
  • P. Boyle - Advanced Bionics Ltd, Cambridge, UK

Medical University of Hannover, Department of Otolaryngology. Fourth International Symposium and Workshops: Objective Measures in Cochlear Implants. Hannover, 01.-04.06.2005. Düsseldorf, Köln: German Medical Science; 2005. Doc05omci025

The electronic version of this article is the complete one and can be found online at: http://www.egms.de/en/meetings/omci2005/05omci025.shtml

Published: May 31, 2005

© 2005 Litvak et al.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by-nc-nd/3.0/deed.en). You are free: to Share – to copy, distribute and transmit the work, provided the original author and source are credited.


Outline

Text

Introduction

Most cochlear implants are now implanted in young children. This population may provide minimal or no feedback and hence present a challenge to program accurately. Modern cochlear implant systems support objective measurements which can complement minimal behavioural measures and greatly ease programming. However, the inherent overlap of a large stimulus artefact with the much smaller neural response may also be difficult to interpret leading to a different set of challenges.

Materials and Methods

This paper describes a rigorous, largely automatic, statistically based method for determining whether or not a neural component is present. A principle component analysis approach was taken to reduce the noise in the data. Neural Response Imaging (NRI) datasets were analyzed using the automatic system and also by a set of experienced observers as in typical clinical practice. In particular, trained observers classified each trace as a response or as a non-response.

Results

1076 traces of Neural Response Imaging (NRI) data were analyzed. Three experienced observers produced relatively poor agreement (agreeing on only approximately 50% of the measurements). The objective method correctly identified 97% of records labeled as responses by all observers. A similar number of non-responses were correctly identified.

Conclusions

Because further averaging can be used to increase certainty of the measurement in borderline cases, the objective classification method may be used clinically to automatically acquire neural growth curves. This should enhance the interpretation of neural measurements and therefore improve programming of difficult cases.