A "biocomputer" powered by human brain cells could be developed in our lifetime, researchers at Johns Hopkins University hope the technology will exponentially expand the capabilities of today's computers and open up new areas of research.
The team outlined their "organoid intelligence" plan today in Frontiers in Science .
“Computing and artificial intelligence have led to a technological revolution, but they have reached the limit,” said Thomas Hartung, professor of environmental health sciences at the Johns Hopkins Bloomberg School of Public Health and the Whiting School of Engineering. "Biocomputing is a major attempt to concentrate computing power and increase its efficiency beyond our current technological limitations."
Image caption: An enlarged image of a brain organoid from Thomas Hartung's lab, colored to show neurons in magenta, cell nuclei in blue, and other support cells in red and green.
Image credit : Jesse Plotkin/Johns Hopkins University.
For nearly two decades, scientists have been using tiny lab-grown tissue organelles that resemble fully grown organs to perform experiments on kidneys, lungs and other organs without testing on humans or animals. More recently, Hartung and his colleagues at Johns Hopkins have been working with brain organoids, pen-sized spheres of neurons and other features that promise to support basic functions like learning and memory.
“This opens up opportunities to study how the human brain works,” Hartung said. "Because you can start manipulating the system, doing things that you can't ethically do to the human brain."
In 2012, Hartung began growing brain cells and assembling functional organelles from them using cells from human skin samples that had been reprogrammed into embryonic-like states. Each organelle contains about 50,000 cells, which is about the size of a fruit fly's nervous system. Now he wants to build a futuristic computer with these brain organelles.
Computers running on this "biological hardware" could start reducing the power consumption of increasingly unstable supercomputers in the next decade, Hartung said. While computers can process numbers and data faster than humans, their brains are much smarter at making complex logical decisions, such as telling a cat from a dog.
“The brain is still no match for today's computers,” said Hartung. “Frontier, a new supercomputer in Kentucky, is a $600 million, 6,800-square-foot facility. Last June, it surpassed the processing power of a single human brain for the first time, but used a million times more energy.”
Thomas Hartung
Professor of Environmental Sciences
According to Hartung, it could take decades for organoid intelligence to power a mouse-like intelligence system. But by increasing the production of brain organoids and training them for artificial intelligence, he envisions a future in which biocomputers will support higher computational speed, processing power, data efficiency and storage capacity.
“It will take decades to get something comparable to any computer,” Hartung said. "But if we don't start creating funding programs for that, it will be more difficult."
Intelligent organoids could revolutionize drug testing research for neurodevelopmental disorders and neurodegeneration, said Lena Smirnova, assistant professor of health and environmental engineering at Johns Hopkins University, who led the study.
“We want to compare brain organoids from normally developing donors with brain organoids from donors with autism,” Smirnova said. “The tools we are developing for biological computing are the same tools that will help us understand the changes in neural networks that are characteristic of autism, without using animals or giving patients access to understand the mechanisms underlying these problems and cognitive impairment in patients."
To assess the ethical implications of working with organoid intelligence, the group brings together a diverse consortium of scientists, bioethicists and members of the public.
Johns Hopkins Contributors: Brian S. Cafo, David H. Gracioso, Qi Huang, Yitzi E. Morales Pantosa, Bohao Tang, Donald J. Jack, Cynthia A. Berlinick, J. Lomax Boyd, Timothy D. Harris, Eric S. . Johnson, Geoffrey Kahn, Burton L. Polhamas, Jesse Plotkin, Alexander S. Zahle, Joshua T. Vogelstein, and Paul F. Worley.