Scientists have developed an artificial intelligence algorithm capable of identifying different types of neurons based solely on brain activity recordings. This new method achieves high accuracy, specifically 95%, without the need for complex genetic tools to label specific cell types. The researchers state this breakthrough addresses a century-old challenge in neuroscience and opens the door to a better understanding of how different neurons contribute to behavior and disease processes. The developed tool holds potential for future applications, including improving neural implants and helping to decode neurological disorders such as epilepsy.
Overcoming a Blind Spot in Neuroscience
Brains are composed of many different types of neurons, each thought to play distinct roles in processing information. For a long time, scientists have used electrodes to record neuronal activity by detecting the electrical ‘spikes’ they generate. While recording these spikes has been invaluable for monitoring the activity of individual neurons deep in the brain, the method was ‘blind’ to the specific type of neuron being recorded.
This made it impossible to identify how different neuron types contribute to the brain’s overall operation. In a new study published in Cell, a research team reports overcoming this problem by identifying the distinct ‘electrical signatures’ of different neuron types in the mouse brain, using brief pulses of blue light to trigger spikes in specific cell types (optogenetics). They created a library of these signatures to train the AI.
AI Training and Validation
The AI algorithm was trained on the library of electrical signatures. It is capable of automatically recognizing five different types of neurons with 95% accuracy without requiring further genetic tools for identification. The algorithm was also validated on brain recording data from monkeys, demonstrating its applicability across different species. Dr. Maxime Beau, co-first author of the study from the UCL Wolfson Institute for Biomedical Research, commented on the long-standing problem: “For decades, neuroscientists have struggled with the fundamental problem of reliably identifying the many different types of neurons that are simultaneously active during behavior.”
New Capabilities and Comparisons
Dr. Beau highlighted the accuracy achieved by the new approach. He was quoted saying, “Our approach now enables us to identify neuron types with over 95% accuracy in mice and in monkeys.” Comparing neurons to computational units, he stated, “This advance will enable researchers to record brain circuits as they perform complex behaviors such as movement. Like logic gates on a computer chip, neurons in the brain are elementary computing units that come in several types.” Beau added, “Our method provides a tool to identify many of the brain’s logic gates in action at the same time. Before, it could only be done one at a time and at much greater cost.”
Short-Term Benefits and Long-Term Aims
In the short term, the new technique means researchers can study the brain using any normal animal without complex genetic engineering to identify neuron types. This allows for studying how different neurons interact and generate behavior. One ultimate aim is to study neurological and neuropsychiatric disorders like epilepsy, autism, and dementia, many thought to involve changes in how cell types interact.
Professor Beverley Clark, a senior author from UCL Wolfson Institute, likened this to studying an orchestra: “Just as many different instruments in an orchestra contribute to the sound of a symphony, the brain relies on many distinct neuron types to create the complex behavior that humans and other animals exhibit.” Quote her: “Our work is analogous to learning the sound that each instrument makes and then teaching an algorithm to recognize the contribution of each of them to a symphony.”
Deciphering the Neural Symphony
Professor Clark also commented on the historical challenge: “Being able to observe this ‘neural symphony’ of the brain in action has been a fundamental challenge in neuroscience for over 100 years, and we now have a method for reliably doing this.” She acknowledged the road ahead: “Although the technology is a long way from being able to be used to study neurological conditions such as epilepsy, we’ve now overcome a major hurdle to reaching that goal.” She added the technique could study recordings of human brain activity (already recorded during surgery) to understand how brains work, first in health, then in disease.
Potential for Neural Implants
The new technique also holds potential for improving human brain-to-computer interfaces or neural implants. Researchers say it could help implants by more accurately recording which cell types are involved in specific actions, allowing implants to recognize signals and generate responses. Key to this is understanding the healthy brain to compensate for damage (e.g., after a stroke).
Convergence of Innovations and Team Synergy
Professor Michael Häusser, a senior author from UCL Division of Medicine and the University of Hong Kong, attributed the project’s success to the convergence of three innovations: molecular biology tagging with light, silicon probe recording technology, and fast AI deep learning improvements. He stated, “This project came to life thanks to the convergence of three critical innovations: using molecular biology to successfully ‘tag’ different neuron types using light, developments in silicon probe recording technology, and of course the fast-paced improvements in deep learning.
” He emphasized team synergy: “Crucially, the synergy in our team was absolutely instrumental. The partner labs at UCL, Baylor, Duke, and Bar Ilan University have all contributed critical pieces to the puzzle. Just like the brain, the whole is larger than the sum of its parts.”
Open Resources and Funding
The database gathered by the team is freely available, and the algorithm is open source, allowing scientists worldwide to use them for neurological research. Funding for this research was provided by Wellcome, the National Institutes of Health (NIH), the European Research Council (ERC), and the European Union’s Horizon 2020 research and innovation program.
