Precompensating for spread of excitation in a cochlear implant coding strategy.

Cochlear implant users' limited ability to understand speech in noisy environments has been linked to the poor spatial resolution and the high degree of spectral smearing associated with the spread of neural excitation. A sound coding algorithm that aims to improve the spectro-temporal representation of the sound signal at the implanted ear by precompensating the electrical stimulation for the spread of excitation is presented in this study. The spread precompensation algorithm was integrated into the standard clinical advanced combination encoder (ACE) strategy and the resulting strategy was called SPACE. SPACE was evaluated acutely with a group of six implant users and was compared to their daily used ACE strategy in terms of preference rating and speech recognition in four-talker babble and stationary speech-shaped noise. While no significant differences in preference rating were observed, speech recognition in four-talker babble was improved by SPACE processing. Analysis of the group results revealed a significant improvement in mean speech reception threshold (SRT) over the ACE strategy of 1.4 dB in four-talker babble, whereas the difference of 0.9 dB in stationary noise did not reach statistical significance. Assessment of individual differences showed that four out of six listeners obtained significant SRT improvements with SPACE and that no subject scored significantly worse compared to ACE. The results suggest that the proposed sound coding strategy has the potential to improve speech perception for cochlear implant users in challenging listening situations.

Auditory inspired machine learning techniques can improve speech intelligibility and quality for hearing-impaired listeners.

Machine-learning based approaches to speech enhancement have recently shown great promise for improving speech intelligibility for hearing-impaired listeners. Here, the performance of three machine-learning algorithms and one classical algorithm, Wiener filtering, was compared. Two algorithms based on neural networks were examined, one using a previously reported feature set and one using a feature set derived from an auditory model. The third machine-learning approach was a dictionary-based sparse-coding algorithm. Speech intelligibility and quality scores were obtained for participants with mild-to-moderate hearing impairments listening to sentences in speech-shaped noise and multi-talker babble following processing with the algorithms. Intelligibility and quality scores were significantly improved by each of the three machine-learning approaches, but not by the classical approach. The largest improvements for both speech intelligibility and quality were found by implementing a neural network using the feature set based on auditory modeling. Furthermore, neural network based techniques appeared more promising than dictionary-based, sparse coding in terms of performance and ease of implementation.

Speech enhancement based on neural networks improves speech intelligibility in noise for cochlear implant users.

Speech understanding in noisy environments is still one of the major challenges for cochlear implant (CI) users in everyday life. We evaluated a speech enhancement algorithm based on neural networks (NNSE) for improving speech intelligibility in noise for CI users. The algorithm decomposes the noisy speech signal into time-frequency units, extracts a set of auditory-inspired features and feeds them to the neural network to produce an estimation of which frequency channels contain more perceptually important information (higher signal-to-noise ratio, SNR). This estimate is used to attenuate noise-dominated and retain speech-dominated CI channels for electrical stimulation, as in traditional n-of-m CI coding strategies. The proposed algorithm was evaluated by measuring the speech-in-noise performance of 14 CI users using three types of background noise. Two NNSE algorithms were compared: a speaker-dependent algorithm, that was trained on the target speaker used for testing, and a speaker-independent algorithm, that was trained on different speakers. Significant improvements in the intelligibility of speech in stationary and fluctuating noises were found relative to the unprocessed condition for the speaker-dependent algorithm in all noise types and for the speaker-independent algorithm in 2 out of 3 noise types. The NNSE algorithms used noise-specific neural networks that generalized to novel segments of the same noise type and worked over a range of SNRs. The proposed algorithm has the potential to improve the intelligibility of speech in noise for CI users while meeting the requirements of low computational complexity and processing delay for application in CI devices.