Voice Recognition is Achieved Using Brain Tissue on a Chip

Voice Recognition is Achieved Using Brain Tissue on a Chip

Aggregates of laboratory-grown brain cells linked to a computer demonstrate basic capabilities in speech recognition and solving mathematical problems.
Brainoware with unsupervised learning for AI computing. a, Schematic of an adaptive reservoir computing framework using Brainoware. b, Schematic of the paradigm of Brainoware setup that mounts a single brain organoid onto a high-density MEA for receiving inputs and sending outputs. c, Whole-mount immunostaining of cortical organoids showing complex three-dimensional neuronal networks with various brain cell identities (for example, mature neuron, MAP2; astrocyte GFAP; neurons of early differentiation stage, TuJ1; neural progenitor cells, SOX2). d, Schematic demonstrating the hypothesized, unsupervised learning of Brainoware by reshaping the BNN during training, and the inhibition of unsupervised learning after synaptic plasticity is blocked. Scale bar, 100 μm. Credit: Nature Electronics (2023). DOI: 10.1038/s41928-023-01069-w

Aggregates of laboratory-grown brain cells linked to a computer demonstrate basic capabilities in speech recognition and solving mathematical problems.

Feng Guo, a bioengineer at Indiana University, Bloomington, and his team have made significant progress in showcasing how brain-inspired computer neural networks can enhance artificial intelligence capabilities.

Brainoware Breakthrough

They cultivated specialized stem cells into neurons, the primary building blocks of the brain, forming an organoid smaller than a nanometer. Named Brainoware, this organoid, connected to a circuit board with electrodes, demonstrated voice recognition and outperformed an artificial network in predicting a Henon map, achieving exciting possibilities for biocomputing in the future.

Biocomputing holds a notable advantage in energy efficiency compared to artificial neural networks, which currently consume millions of watts daily. Brainoware, acting as a bridge between AI and organoids, explores leveraging the biological neural network within brain organoids for computing.

Biocomputing’s Promising Applications in Neurological Disease Research and Sleep Brain Wave Decoding

The proof-of-concept aims to demonstrate the feasibility of this approach, offering potential applications in studying neurological diseases like Alzheimer’s and decoding brain wave activity during sleep.

To conclude, Challenges include maintaining organoid health and addressing neuroethical concerns, highlighting the need for continued research into learning mechanisms, neural development, and cognitive aspects of neurodegenerative diseases.


Read the original article on: TechXplore

Read more: Video-to-Sound Tech Helps Visually Impaired Recognize Faces

Share this post