Gamers helping to piece together the brain’s complex circuitry
Crowd-sourced science moves to new level with the help of artificial intelligence
Billions of cells known as neurons, connected through trillions of synapses, make up the human brain. Mapping neurons helps find answers on how they connect and respond to a stimulus.
The brain is the all-powerful driver of the nervous system, made up of a vast number of specialised cells transmitting nerve impulses, known as neurons – which allow us to process information and to respond rapidly. And yet there is so much about how the organ is structured and functions that is still unknown, according to US neuroscientist Amy Sterling.
Billions of neurons, connected through trillions of synapses, make up the most complex thing in the Universe: the human brain. But we still don’t know how many neurons there are; what types exist, and how they are connected, she explains in advance of her lecture at the Science Gallery in Trinity College, Dublin.
“Despite great advances, we don’t know how many neurons are in our own heads,” she adds. Mapping neurons helps to begin the process of finding answers but it’s painstaking work, normally “requiring 100 hours of expertise to map just one neuron”, she adds.
Crowd-sourcing citizen scientists around the world and adding artificial intelligence (AI) capacity moves the mapping process to a new level. Those elements were combined to make EyeWire, a game to map the brain co-ordinated from the Seung Lab at Princeton University. This computation game challenges players to map neurons in the retina.
Five years on, with the help of more than 250,000 players from 160 different countries, it has been a remarkable success from a scientific discovery perspective, according to Sterling, who is EyeWire’s executive director. She describes the puzzle as “connecting of jagged tree branches” – it’s helping to construct the circuitry of a very small cube of tissue taken from a mouse’s retina.
By solving a puzzle it helps researchers chart previously unknown cells and circuits in the brain. When combined with brain activity data, they can begin to model the information processing capabilities of neurons. It’s all about connectedness.
To create neural wiring diagrams by hand, researchers need to start with cross-sectional images of ultrathin brain slices. Each image features cross sections of individual axons and dendrites, which the scientist/gamer traces and tracks through the stack of images to painstakingly reconstruct circuits. While computers can reconstruct much of the structure, the ability of humans to see patterns is much better – players colour in all the branches of an individual neuron.
The field of “connectomics” aims to map these billions of neurons and trillions of synaptic connections, because knowing the architecture of our connectomes – essentially wiring diagrams – is crucial for understanding how thoughts, emotions, and consciousness originate in our brains.
The output has already helped identify a new cell type in the retina of the eye and has shown how some retinal neurons respond to movement direction. “The development of graphics processing units in parallel with more advanced artificial AI algorithms have further revolutionised connectomics, resulting in software that can reconstruct some brain areas with super human accuracy.”
A welcome “by-product” of EyeWire and research stemming from it, she adds, is the ability to generate the 3-dimensional maps for visualizations of the brain, detailed to a minute scale.
Her quest is to understand the brain began not because she wanted to chart highly-detailed maps of neuron circuits, but because she is deeply curious how a brain makes a mind. “EyeWire stands at the intersection of AI and human insight, paralleling the inevitable rise of AI and the restructuring of human endeavours as we know it,” she told the New York Times in the project’s early years.
Tech du jour
Its success led to Fast Company crediting her with “making neuroscience into a playground for today’s hot tech du jour”, while she was named among the Forbes “30 Under 30” in 2015.
Driven by EyeWire’s success, ambition is about to be scaled up later this year with the birth of a new game called NEO. It will reconstruct a 1,000-fold larger volume of brain matter arising from the cerebral cortex, which is composed of folded grey matter and plays an important role in consciousness.
Backed by much better AI, participating gamers will help map 70,000 neurons and 1 billion synapses by way of engaging in a challenging puzzle, adds Sterling whose career has broadened into science communication and policy.
She has advised the White House’s Office of Science and Technology Policy and the US Congress on crowd-sourcing methods in STEM. Her lecture was part of TCD’s “Neurohumanities Public Talks” bringing STEM researchers and those working in the humanities together to reflect, discuss and think together as well as providing a platform for public engagement.
Any notion that citizen science is a worthy but marginal exercise while real scientists get on with their much more important endeavour is belied by EyeWire.
“Citizen neuroscience” is a platform that has brought people into the lab remotely, enabled them to contribute to advanced science, she believes, and helps “shake off the exclusive aura of science”. It genuinely helps researchers, Sterling says, “and makes a meaningful contribution to their work”.