In a bid to unravel the complexities of the human brain and shed light on the origins of various mental illnesses and neurological disorders, Google Research, in collaboration with a consortium of esteemed institutions, has embarked on an ambitious five-year, $33 million project. The venture aims to pioneer advancements in the field of connectomics, which seeks to chart the intricate network of cellular connections within the brain. This groundbreaking initiative, supported by the Brain Research Through Advancing Innovative Neurotechnologies (BRAIN) Initiative at the National Institutes of Health (NIH), is led by researchers from Harvard University, with contributions from the Allen Institute, MIT, Cambridge University, Princeton University, and Johns Hopkins University, along with guidance from advisers at HHMI’s Janelia Research Campus.
The primary challenge before the scientific community lies in comprehending the workings of the human brain, a marvel of computational complexity composed of billions of cells. Mental illnesses and neurodegenerative diseases, including dementia, remain in mystery due to a lack of comprehensive understanding of the brain’s intricate network. The nascent field of connectomics holds immense promise in unraveling these mysteries, potentially offering avenues for effective treatments.
The current venture will focus on mapping a minute portion (2-3%) of the mouse brain, honing in on the hippocampal region. This area is pivotal in memory encoding, attention, and spatial navigation. By charting the connectivity of this crucial segment, the researchers aspire to lay the groundwork for future endeavors targeting a comprehensive understanding of the entire mouse brain and, by extension, the human brain.
In 2021, Google Research and its partners at Harvard achieved a significant milestone by mapping one cubic millimeter of the human brain, a dataset known as the H01 dataset. However, mapping the entire human brain’s connectome necessitates an unprecedented amount of data – up to a zettabyte, currently beyond the capabilities of existing technology. Focusing on the mouse brain, which offers a more manageable scope, presents an invaluable opportunity to glean insights relevant to human brain function and dysfunction.
To undertake this endeavor, the project will amass an extraordinary dataset, encompassing approximately 25,000 terabytes or 25 petabytes of brain data. This dataset, slated to be one of the largest in the realm of biology, eclipses the scale of previous connectome mapping projects. In perspective, the data collected from mapping the mouse hippocampal region alone is equivalent to more than 48,800 Pixel phones stacked as high as the Empire State Building.
The venture also encompasses the development of tools and technologies tailored to managing massive connectomic datasets. These include flood-filling networks, which utilize deep learning to trace neurons’ paths in three-dimensional brain volumes, and self-supervised learning technology, SegCLR, to extract vital insights from segmented volumes automatically. Core connectomics infrastructure will be fortified to handle the unprecedented scale of data.
In conclusion, Google Research’s audacious endeavor to map the mouse brain represents a pivotal step toward unraveling the enigma of neurological disorders. Through the collaborative efforts of leading research institutions and the application of cutting-edge technologies, this project promises to deepen our understanding of the brain and holds the potential to revolutionize treatments for a spectrum of neurological conditions. As the initiative gains momentum, the scientific community eagerly anticipates the revelations that peering into a mouse’s mind may unveil about our own.
Check out the Google Blog. All Credit For This Research Goes To the Researchers on This Project. Also, don’t forget to join our 30k+ ML SubReddit, 40k+ Facebook Community, Discord Channel, and Email Newsletter, where we share the latest AI research news, cool AI projects, and more.
Niharika is a Technical consulting intern at Marktechpost. She is a third year undergraduate, currently pursuing her B.Tech from Indian Institute of Technology(IIT), Kharagpur. She is a highly enthusiastic individual with a keen interest in Machine learning, Data science and AI and an avid reader of the latest developments in these fields.