“Hearst Magazines and Yahoo may earn commissions or revenue from certain merchandise through these links.”
As you read this story, you will learn the following:
-
In a new study, scientists successfully trained brain organoids derived from mouse stem cells to solve an engineering benchmark known as the “stem problem.”
-
By applying weak or strong electrical signals to reinforce certain behaviors through an adaptive AI training algorithm, the researchers found that the performance of this small patch of brain tissue increased dramatically, from a 4.5% success rate with random training to more than 46% with adaptive training and reinforcement learning.
-
A biologist unaffiliated with the study said the work shows that “the ability to perform adaptive computation is intrinsic to the cortical tissue itself, separate from all the scaffolding we normally think of as necessary.”
While creating organs may sound like the stuff of science fiction, scientists have actually been experimenting with the idea for more than a century. In 1907, for example, American biologist Henry Van Peters Wilson demonstrated the basic principles of laboratory-grown organs, or “organoids,” by showing how isolated cells from a sponge could self-organize and regenerate outside the body. This exploration continued in various animals for decades, until finally, in 2009, scientists created the first 3D organoids using intestinal stem cells from mice.
In the following decades, Organoids are becoming increasingly complex It provides scientists with new ways to study how cells, tissues and organs work in the body, while also providing a cheaper and more ethical platform for developing potential treatments. One of the most compelling use cases for organoids is our never-ending study of the brain, the most complex of all organs. In a new study published in diary cell reportA team of scientists from the University of California (UC) Santa Cruz has successfully trained brain organoids developed from mouse stem cells to solve an engineering benchmark known as the “stem problem.” Because these organoids lack a body, any sense of biological purpose, or sensory experience, this learning behavior suggests that adaptive computation may be intrinsic to the cortical tissue itself.
“We are trying to understand the fundamentals of how neurons adapt adaptively to solve problems,” said UC Santa Cruz Ph.D. student Ash Robbins, lead author of the study, said in a press statement. “If we can identify the driving factors in dishes, it gives us new ways to study how neurological diseases affect the brain’s ability to learn.”
The brain organoid, which is less than the size of a peppercorn but packed with millions of neurons, is affixed to a special chip that allows scientists to observe and control the firing of certain neurons. The researchers used an electrophysiology system that uses electrical stimulation to send and receive information from neurons. To test the system, they relied on the classic “car pole problem,” a balance task similar to keeping a vertical ruler upright on the hand while adjusting for motion and gravity. The researchers “taught” the organoids to balance computer-simulated rods using weak or strong electrical signals. If the organoid fails to maintain equilibrium for longer than the average time limit, it undergoes “reinforcement learning” through an artificial intelligence algorithm that selects which neurons to train.
“You can think of it like a human coach saying, ‘You’re doing it wrong, tweak it a little bit this way,'” Robbins said in a press statement. “We’re learning how best to give it these guidance signals.”
Scientists found that the success rate of random training was only 4.5%, while adaptive training greatly improved the success rate, reaching 46%. However, this brief learning almost disappeared once the organoids rested for more than 45 minutes. Because the organoid does not have multiple brain regions like humans (or mice), it cannot retain long-term learning capabilities.
This isn’t the first time lab-grown brains have been trained to solve specific problems. Back in 2022, scientists trained a synthetic brain learn how to play table tennisand more recently, other researchers have developed neuromorphic platforms Using human brain cells. For UC Santa Cruz researchers, the goal is to develop a brain organoid platform that could help scientists treat neurological diseases. But as research becomes more advanced—especially when human stem cells are involved—experts are pondering an equally important ethical question: How live Are these organoids at all?
“We want to be clear that our goal is to advance brain research and treatments for neurological diseases, not to replace robot controllers and other types of computers with lab-grown animal brain tissue,” study co-author David Haussler of the University of California, Santa Cruz, said in a press release. “The latter may be considered cool, but it raises serious ethical issues, especially if human brain organoids are used.”
You may also like
