0
0
In a groundbreaking study published in the journal Science, researchers from Charité-Universitätsmedizin Berlin have unveiled a significant revelation about the wiring of nerve cells in the human neocortex, challenging previous assumptions and shedding light on the brain’s remarkable efficiency in processing information.
Contrary to earlier beliefs drawn from animal models such as mice, the study demonstrates that human neurons in the neocortex communicate in a unidirectional manner, rather than through recurrent loops. This unique neural architecture significantly enhances the brain’s capacity for processing data, offering new insights that could advance the development of artificial neural networks.
The neocortex, a critical hub for human intelligence, boasts a complex structure less than five millimeters thick, housing approximately 20 billion neurons responsible for processing diverse sensory inputs and orchestrating conscious activities. Understanding how these neurons interconnect is pivotal to deciphering the brain’s intricate workings.
Prof. Jörg Geiger, Director of the Institute for Neurophysiology at Charité, elucidates, “Our previous understanding of neural architecture in the neocortex is based primarily on findings from animal models such as mice. In those models, the neighboring neurons frequently communicate with each other as if they are in dialogue. One neuron signals another, and then that one sends a signal back. That means the information often flows in recurrent loops.”
However, the human neocortex, with its greater complexity and thickness, deviates from this paradigm. Through an innovative methodology leveraging rare tissue samples from neurosurgery patients, the research team developed an enhanced version of the “multipatch” technique to observe neuronal communications.
The study involved meticulously analyzing nearly 1,170 neurons and approximately 7,200 potential connections, revealing that only a fraction of neurons engage in reciprocal dialogue. Instead, information predominantly flows in a forward-directed manner, with minimal return to the starting point or through cycles.
Dr. Yangfan Peng, first author of the study, underscores the significance of this directional flow, stating, “In humans, the information tends to flow in one direction instead. It seldom returns to the starting point either directly or via cycles.”
Furthermore, the researchers employed computer simulations to demonstrate that this directed network architecture confers advantages in data processing efficiency. The artificial neural network model, mimicking human structures, outperformed its mouse-modeled counterpart in a speech recognition task, underscoring the human brain’s superior computational prowess.
Prof. Geiger emphasizes the potential implications of these findings for artificial intelligence (AI) development, noting, “These insights into cost-efficient information processing in the human neocortex could provide further inspiration for refining AI networks.”
The study, a collaborative effort between Charité’s basic research and clinical departments, underscores the invaluable synergy between medical practice and scientific inquiry, offering a glimpse into the intricate workings of the human brain and its implications for AI advancement.
Source: Yangfan Peng et al, Directed and acyclic synaptic connectivity in the human layer 2-3 cortical microcircuit, Science (2024).
