Enabling Oral Communication Through Subvocal Recognition For Aphasia

Hi Everyone!

My name is Rishi Ambavanekar and I have started doing an experiment with OpenBCI to create
a communication device for people with Aphasia. Aphasia is a disorder–commonly occurring
after a stroke–in which a patient cannot verbally communicate. My first experience with the
disorder was when my father has a Transient Ischemic Attack, colloquially referred to as a
“mini-stroke”. While he was able to make a full recovery, this prompted my interest in stroke
recovery and Aphasia in particular.

About a year ago, I started working on a device that would use subvocal recognition (read the
inner speech of users) to classify words. This device (which I nicknamed “Cerebro” after a
similar device in X-Men) had an accuracy of 75% on average for 4 words: “food”, “water”,
“get”, and “home”. I attributed the low accuracy to the hardware I had (which I jury-rigged
myself), so I am hoping that with the new OpenBCI equipment I am able to get a 90%+ accuracy
and add more words!

Thanks for taking the time to read this, let me know if you have suggestions or questions at
[email protected]!