Inspenet, December 18, 2023.
Scientists have developed a hybrid biocomputer that fuses lab-grown human brain tissue with conventional electronic circuits, managing to carry out tasks such as voice recognition.
[post_relacionado]This technological advance, detailed in Nature Electronics , could eventually be integrated into artificial intelligence (AI) systems or serve as the basis for improved brain models in neuroscience research.
Brainoware: the hybrid biocomputer
Called Brainoware , this system uses brain organoids, which are collections of human cells designed to mimic tissues and are used in research to model organs. Organoids are made up of stem cells with the ability to specialize into different cell types, in this case, they transformed into neurons, similar to those present in our brain.
The purpose of the research is to establish “a link between artificial intelligence and organoids,” according to Feng Guo, a co-author of the study and a bioengineer at Indiana University Bloomington. Some artificial intelligence systems are based on a network of interconnected nodes, known as a neural network, similar to the operation of the brain. ““We wanted to ask if we can harness the biological neural network within the brain organoid for computing ,” he says.
In creating Brainoware, researchers placed a single organoid on a plate equipped with thousands of electrodes to establish the connection between brain tissue and electrical circuits. Subsequently, they transformed the input information into a pattern of electrical pulses, which were delivered to the organoid. The tissue reaction was recorded by a sensor and decoded by a machine learning algorithm.
Brainoware put to the test
To test Brainoware’s capabilities, the team used the technique in speech recognition tasks, training the system with 240 recordings of eight people speaking. The organoid exhibited distinct patterns of neural activity in response to each voice, and the artificial intelligence learned to interpret these responses with 78% accuracy.
Although more research is needed, the study supports certain essential theoretical ideas that could pave the way toward the viability of a biological computer, according to Lena Smirnova, a developmental neuroscientist at Johns Hopkins University in Baltimore, Maryland. Previous research has shown that only two-dimensional (2D) cultures of neuronal cells are capable of carrying out similar computational tasks, but this study marks the first time it has been demonstrated in a three-dimensional (3D) brain organoid.
Future vision
Combining organoids and circuits could give researchers the opportunity to harness the speed and energy efficiency inherent in human brains to boost artificial intelligence, Guo says. Additionally, Arti Ahluwalia, a biomedical engineer at the University of Pisa in Italy, points out that this technology could be used to investigate the brain, since brain organoids can replicate the architecture and function of a functioning brain in ways that simple cell cultures can. they do not achieve
Likewise, there is the possibility of using Brainoware to model and study neurological disorders and it could also be used to evaluate the effects and toxicities of different treatments. “ That’s where the promise is; use them to hopefully one day replace animal models of the brain ,” says Ahluwalia.
However, the use of living cells in computing presents significant challenges. A crucial problem lies in how to keep the organoids alive, since the cells must be cultured and preserved in incubators, a task that becomes more complicated as the organoids increase in size. According to Smirnova, more complex tasks will require larger “brains.”
To fully leverage Brainoware ‘s capabilities, Guo maintains that future steps will include investigating the adaptability of brain organoids to carry out more complex tasks and how to design them so that they are more stable and reliable than today. This aspect will be crucial if they are to be incorporated into the silicon microchips currently used in artificial intelligence computing.
Don’t miss any of our posts and follow us on social media!
Inspenet.com YouTube LinkedIn Facebook Instagram
Source: nature.com