An implanted nuclear powered anything is a bad idea. You should research into nuclear power so you understand the physics and the fact it will never provide a viable source of implanted power.
Secondly the issue with a fully implantable cochear implant is that it is restricted to the fact that current microphone, battery and sound processing technology is not mature enough to be implantable. Remember that infants are being implanted with the expectation that it will operate for their whole life. Battery technology is not there yet (and no...nuclear power that is small enough to implant is not the answer).
"Cochlear implant companies are NEVER going to release a commercial totally implantable cochlear implant"....... "Man will never walk on the moon", "No one will ever need more than 640K of RAM in their PC".
Replay available now: A handful of emerging network technologies are competing to be the preferred wide-area connection for the Internet of Things. All claim lower costs and power use than cellular but none have wide deployment yet. Listen in as proponents of leading contenders make their case to be the metro or national IoT network of the future. Rick Merritt, EE Times Silicon Valley Bureau Chief, moderators this discussion. Join in and ask his guests questions.