A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2022; you can also visit <a rel="external noopener" href="https://os.zhdk.cloud.switch.ch/tind-tmp-epfl/717172a9-e706-4677-aea8-0e149341c6c5?response-content-disposition=attachment%3B%20filename*%3DUTF-8''EPFL_TH8379.pdf&response-content-type=application%2Fpdf&AWSAccessKeyId=ded3589a13b4450889b2f728d54861a6&Expires=1641737493&Signature=fH0YuMmVXeAuiVfr68uqJb5C3as%3D">the original URL</a>. The file type is <code>application/pdf</code>.
Thanks to recent technological advances in microelectronics and bioengineering, it is now possible to restore lost or impaired sensory modalities by interfering the nervous system with electronic devices and artificially reproducing the electrical encoding of the neural signals of the missing or defective sensory organs or tissues, as evidenced by the successful development of cochlear implants allow the restoration of hearing in deaf patients. This makes the treatment of blindness, a very<span class="external-identifiers"> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.5075/epfl-thesis-8379">doi:10.5075/epfl-thesis-8379</a> <a target="_blank" rel="external noopener" href="https://fatcat.wiki/release/wohihwreqfhq5b6ijzk3jknmey">fatcat:wohihwreqfhq5b6ijzk3jknmey</a> </span>
more »... ring condition that significantly decreases the quality of life of the person it affects, a plausible possibility today, and many groups are currently attempting to apply the same principle of cochlear implants to sight restoration. So far, the preferred approach has been retinal implants due to their proximity to the defective cells, the photoreceptors, which minimizes the amount of processing of the neural signal that needs to be reproduced, and due to the reasonable number of stimulating sites that the retina can accommodate. Several devices have already demonstrated the possibility of restoring some degree of functional vision, albeit still very rudimentary. However, unlike hearing, vision is an extremely complex sensory modality, involving several cell types forming elaborate circuitry, which makes the advantages of retinal implants less definitive, and other alternative approaches such as optic nerve of cortical prosthesis still worth exploring. We believe the optic nerve stimulation is an attractive approach for several reasons; firstly, it does not require optical transparency. Secondly, it can address specific conditions such as severe eye trauma or retinal detachment where retinal implants cannot be employed. Finally, thanks to the organization of the optic nerve axonal fibers, it may likely result in the activation of axons originating from the same region of the visual field. This is not the case with epiretinal implants, which have been reported to stimulate undesired axons of passage in addition to the somas of their cellular targets, resulting in curved and elongated phosphenes that limit the implant's achievable resolution. In this thesis, we introduce a novel intraneural electrode design, the OpticSELINE, which aims at improving the selectivity and stability of the extra-neural cuff-electrodes previously used in optic nerve stimulation. We first characterized the device's improved mechanical and electrical properties and then proceeded to evaluate its stimulation capacity in vivo in the rabbit. In parallel to demonstrating the ability to modulate the magnitude of the cortical activation through the increase or decrease of the stimulation current, we implemented a hybrid finite element computational model to estimate the dimensions of the nerve portions activated by our electrical stimulation protocols. Using independent component analysis (ICA), we also developed a workflow that allows isolating components of the signal selectively related to a single or a pair of stimulating electrodes. We then focused on the development of new and improved algorithms to analyze the cortical activation patterns. We first developed a support machine vector (SVM)-based classifier that allowed us to accurately identify both the visual stimulus or the stimulating electrode that elicited a given cortical activation pattern amongst ten and eight total potential stimuli. Finally, we developed a regression model that also allowed us to accurately predict both the visual stimulus or the stimulating electrode, without having encountered the cortical activation patterns it was tested on during its training phase, resulting in an improved extrapolation capacity compared to the classifier.
<a target="_blank" rel="noopener" href="https://web.archive.org/web/20220108141133/https://os.zhdk.cloud.switch.ch/tind-tmp-epfl/717172a9-e706-4677-aea8-0e149341c6c5?response-content-disposition=attachment%3B%20filename*%3DUTF-8''EPFL_TH8379.pdf&response-content-type=application%2Fpdf&AWSAccessKeyId=ded3589a13b4450889b2f728d54861a6&Expires=1641737493&Signature=fH0YuMmVXeAuiVfr68uqJb5C3as%3D" title="fulltext PDF download" data-goatcounter-click="serp-fulltext" data-goatcounter-title="serp-fulltext"> <button class="ui simple right pointing dropdown compact black labeled icon button serp-button"> <i class="icon ia-icon"></i> Web Archive [PDF] <div class="menu fulltext-thumbnail"> <img src="https://blobs.fatcat.wiki/thumbnail/pdf/0c/98/0c98641a86ab3847f566a72f1ef907f03ac56c8c.180px.jpg" alt="fulltext thumbnail" loading="lazy"> </div> </button> </a> <a target="_blank" rel="external noopener noreferrer" href="https://doi.org/10.5075/epfl-thesis-8379"> <button class="ui left aligned compact blue labeled icon button serp-button"> <i class="external alternate icon"></i> Publisher / doi.org </button> </a>