| Cortical auditory evoked potential (CAEP) tests are used to evaluate cochlear implant (CI) patient auditory pathways, but when the CI device processes sound stimuli it produces an electrical artifact which obscures the relevant information in the neural response. Currently there are multiple methods which attempt to extract the neural response from the contaminated CAEP, but there is no gold standard which can quantitatively confirm the effectiveness of these methods. To address this crucial shortcoming, this work employs time-frequency analysis, using the continuous wavelet transform (CWT), to quantify how much artifact energy remains in the neural response recovered by these methods. The proposed CWT evaluation tool calculates the two-dimensional correlation coefficient between the time-frequency representations of the extracted neural response and the stimulus signal envelope, which is a good approximation of the artifact, as a means of quantitatively assessing how much artifact energy remains in the extracted response. A novel technique for extracting the neural response from contaminated CAEPs is then proposed. The new method uses matching pursuit (MP) based feature extraction to represent the contaminated CAEP in a feature space, and support vector machines (SVM) to classify the components as normal hearing (NH) or artifact. The NH components are combined to extract the neural response without artifact energy. The proposed method was applied on two sets of CI CAEPs generated using tone stimuli, and was shown to extract CAEPs from CI data more effectively than current artifact removal techniques, as verified using the evaluation tool. The method was then implemented on a CI CAEP generated by a speech stimulus, which has not been performed by any existing methods, and successfully extracted the neural response. The proposed extraction technique and evaluation tool will allow accurate clinical evaluation of CI patient auditory pathways by removing artifact energy which obscures the important CAEP features. |
