Dr. Xing Chen
Xing’s driving ambition is the creation of a brain implant that allows profoundly blind people to regain a functional form of artificial vision. Originally from Singapore, she won a Trustee (full-tuition) scholarship to study at the University of Southern California in 2004, and graduated with a Bachelor’s in Neuroscience in 2008. Subsequently, she obtained her PhD in Visual Neuroscience at Newcastle University (UK), in the lab of Alexander Thiele, examining how extensive training and improvement on fine visual tasks is accompanied by changes in the primate visual system at the neuronal level. She moved to the Netherlands in 2014, working as a postdoctoral researcher with Pieter Roelfsema at the Netherlands Institute for Neuroscience. In 2016, she played a pivotal role in securing funding from the Netherlands Organisation for Scientific Research (NWO) for a 4-year €1.78M programme on the development of a visual cortical neuroprosthesis, involving five research labs, ten companies, and three organisations for the blind and visually impaired. Together with Blackrock Microsystems, she developed an unprecedentedly high-channel-count, high-resolution neuroprosthesis for chronic recording and electrical stimulation of primate visual cortex, establishing proof-of-concept for the generation of artificial vision in the blind.
Globally, 40 million people across the world are blind, many of whom could potentially regain functional vision through a neuroprosthetic device that interfaces directly with the visual cortex. Electrical stimulation of the visual cortex is known to elicit the percept of a dot of light at a particular location in visual space, known as a 'phosphene.' However, the ability to create coherent artificial percepts, consisting of readily interpretable shapes and contours, has remained elusive and only sparsely documented in the literature (Dobelle 1974). Here, we present results obtained from a high-resolution, high-density, high-channel-count prosthesis for the visual cortex, consisting of 1024 electrodes (sixteen 8x8 Utah arrays) that were chronically implanted in V1 and V4 of the left hemisphere, in each of two monkeys.
We developed customized cranial implants (Chen et al., 2017); new and improved surgical tools and techniques; and a large-scale microstimulation and recording system. Initially, the monkeys performed a simple task in which they made eye movements to the location of a phosphene that was elicited by V1 stimulation via individual electrodes. We found a significant correlation between saccade end points and the neurons' RFs. Using novel methods to remove the microstimulation artifact, we recorded neuronal activity in V4 during stimulation of V1, and observed a correlation between V4 activity levels and behavioural reports of phosphene perception. Next, the monkeys were tested on more complex tasks: a line orientation discrimination task, in which microstimulation was delivered on several electrodes simultaneously, creating the percept of either a horizontally or a vertically oriented line; a direction-of-motion task, in which microstimulation was delivered on a sequence of electrodes; and a letter discrimination task, in which microstimulation was delivered on 10-15 electrodes simultaneously, creating a percept in the form of a letter. They successfully reported the identity of these artificially generated percepts- performing above chance levels even with novel combinations of electrodes. Finally, over a period of several years, we monitored the current thresholds needed to elicit phosphene percepts, as well as electrode impedance levels and the quality of neuronal signals recorded on each of the channels.
This proof-of-concept demonstrates the potential of a visual cortical neuroprosthesis for restoration of functional, life-enhancing vision in the blind.
Meibergdreef 47, 1105BA, Amsterdam