Computer Vision News - October 2022

27 Ashley Bruce morning and think, time to run my ten tabs of Colab before I start my day!” Ashley and Michael have been looking at the optimal arrangement of electrodes in epiretinal prostheses, but their work could be generalized to other prostheses . The field is relatively new, so there are many opportunities to explore. There is lots of potential for this type of research to be used in the community to help patients, but it is still early days. Simulations do not always correlate precisely with real life, and patients have many different ρ and λ values that impact how the implant affects them, so more testing in the real world is needed first. “We’ve shown there is merit in using computational models to design new implants and that the current electrode layout might not be the best.” Michael points out. “In that way, we’re using insights from neuroscience to design better technology. We hope our work can help guide future prototyping of new devices.” has been a labor of love at times. “Running these simulations took a while – I always had ten instances of Google Colab up at once to run every simulation with a bunch of different parameters,” she tells us. “I’d be sitting at my computer just going through all the tabs making sure that none of my Colabs had timed out. It could take more than eight hours to run them, and if I let them time out, I’d have to start over. If I ran them while I was asleep, I’d sometimes wake up in the middle of the night and have to get out of bed to check! It’s funny looking back on it because I would wake up every Michael Beyeler BEST OF MICCAI