Information theoretic optimization of cochlear implant electrode usage probabilities

Artificial Visual Information Produced by Retinal Prostheses

Numerous retinal prosthetic systems have demonstrated somewhat useful vision can be restored to individuals who had lost their sight due to outer retinal degenerative diseases. Earlier prosthetic studies have mostly focused on the confinement of electrical stimulation for improved spatial resolution and/or the biased stimulation of specific retinal ganglion cell (RGC) types for selective activation of retinal ON/OFF pathway for enhanced visual percepts. To better replicate normal vision, it would be also crucial to consider information transmission by spiking activities arising in the RGC population since an incredible amount of visual information is transferred from the eye to the brain. In previous studies, however, it has not been well explored how much artificial visual information is created in response to electrical stimuli delivered by microelectrodes. In the present work, we discuss the importance of the neural information for high-quality artificial vision. First, we summarize the previous literatures which have computed information transmission rates from spiking activities of RGCs in response to visual stimuli. Second, we exemplify a couple of studies which computed the neural information from electrically evoked responses. Third, we briefly introduce how information rates can be computed in the representative two ways - direct method and reconstruction method. Fourth, we introduce in silico approaches modeling artificial retinal neural networks to explore the relationship between amount of information and the spiking patterns. Lastly, we conclude our review with clinical implications to emphasize the necessity of considering visual information transmission for further improvement of retinal prosthetics.

Unifying information theory and machine learning in a model of electrode discrimination in cochlear implants

Despite the development and success of cochlear implants over several decades, wide inter-subject variability in speech perception is reported. This suggests that cochlear implant user-dependent factors limit speech perception at the individual level. Clinical studies have demonstrated the importance of the number, placement, and insertion depths of electrodes on speech recognition abilities. However, these do not account for all inter-subject variability and to what extent these factors affect speech recognition abilities has not been studied. In this paper, an information theoretic method and machine learning technique are unified in a model to investigate the extent to which key factors limit cochlear implant electrode discrimination. The framework uses a neural network classifier to predict which electrode is stimulated for a given simulated activation pattern of the auditory nerve, and mutual information is then estimated between the actual stimulated electrode and predicted ones. We also investigate how and to what extent the choices of parameters affect the performance of the model. The advantages of this framework include i) electrode discrimination ability is quantified using information theory, ii) it provides a flexible framework that may be used to investigate the key factors that limit the performance of cochlear implant users, and iii) it provides insights for future modeling studies of other types of neural prostheses.