According to the news of the Medicalxpress site, researchers from the Swiss Federal Institute of Technology have found a new algorithm for the artificial neural network model to be used to capture brain dynamics and signals.
With a machine learning algorithm called CEBRA, the researchers were able to decipher the codes in the brain’s nerves.
In the rat experiment, a rat was shown a black and white video clip from the 1960s. In the experiment, in which the newly developed algorithm was used, the brain signals of the mouse were examined and the images watched could be projected onto another screen almost exactly the same as the original.
In the experiment, the brain signals of the mice were measured with electrode probes placed directly in the visual cortex of the brain and optical probes created with the help of genetically modified mice, while CEBRA was run using data on the brain activity of the mice and the movies they watched.
It has been reported that CEBRA is not capable of fully reconstructing what people see, but represents an important step forward in this area.
Mackenzie Mathis, Head of Integrative Neuroscience at Bertarelli, who conducted the research, stated that the study will serve to enable high-performance brain-machine interfaces (BMIs) and to develop algorithms needed in neurotechnology.
Mathis stated that the success of the new algorithm lies in its ability to combine brain signals with movie data.
The research results were published in the journal “Nature”.