The Miraculous Complexity of the Human Brain: A Journey into Connectomics

Most people don’t think about just how miraculous the human brain is. This organ contains about 80 billion neurons, each of which is connected to as many as 10,000 other neurons. Mapping the neurons themselves is a challenging endeavor, but trying to understand the connections between them is nothing short of a herculean task.

Thomas Uram, Data Science and Workflows Team Lead
Credit: Argonne National Laboratory

Pioneering Brain Mapping at Argonne National Laboratory

While fully mapping the human brain will take many more years of hard work, scientists at Argonne National Laboratory are laying the foundation for future explorations. The project is led by Argonne\’s Nicola Ferrier, Senior Computer Scientist in the Mathematics and Computer Science Division.

To learn more about this amazing work, we spoke with Thomas Uram, a Computer Scientist in the Argonne Leadership Computing Facility, who is also working on the project. Uram stated that the brain is “one of the most complex things on the planet” and emphasized the importance of reconstructing its structure and connectivity to uncover the unknown.

Understanding Connectomics: A Cubic Millimeter of Brain

Research that maps the connections within an organism’s nervous system falls under the umbrella of connectomics. Given the complexity of the brain\’s structure, Uram and his colleagues focus on samples of brain tissue just a cubic millimeter in size. These samples are prepared by taking thousands of 30-nanometer-thick slices of residual human brain tissue removed during surgery.

The slices are then mounted on tape for imaging by an electron microscope. Each section is imaged individually as a collection of tiles and reassembled as a larger section. Once fully reconstructed, the sections are aligned with their neighbors, and a neural network is employed to trace objects within that stack of images. The team utilizes a neural network developed by Google called Flood-Filling Network (FNN) to assist in this complex reconstruction process.

With 80 billion neurons, each having as many as 10,000 connections to other neurons, mapping the connectomics of the brain is an extremely difficult task.

Data Challenges in Mapping the Brain

Even with a relatively small sample, studying every connection poses a significant computational challenge. A cubic millimeter of tissue imaged at a lateral resolution of four nanometers generates about two petabytes of data. Using the current neural networks available, Uram and his team could segment this amount of tissue in just a few days, utilizing all of Aurora’s computing power. However, as the scale of research increases, so do the computational requirements.

Uram pointed out that reconstructing a whole mouse brain would generate 1,000 times more data, leading to an estimated 3,000 days on Aurora. In comparison, mapping an entire human brain would demand an astronomical 3 million days of processing time—far beyond current capabilities.

The Future of Brain Mapping: Innovations and Challenges Ahead

Building machines that are 3 million times more powerful isn’t a feasible solution. Instead, Uram emphasizes the need for significant advancements in technology. If the speed of neural network segmentation improves, existing machines could handle much more complex tasks. He noted that current segmentation processes often involve errors similar to those seen in AI models like ChatGPT, necessitating extensive human proofreading to validate results.

For instance, researchers who mapped a fly\’s brain estimated that human correction of segmentation issues took thousands of hours. Beyond the challenge of human oversight, there is also a pressing storage problem. The data generated from cubic millimeter samples currently reaches petabytes—but for larger projects, storage needs could surpass exabytes, demanding new innovations in data management.

Uram remains optimistic about tackling these daunting obstacles. “I have always been interested in the big questions of life,” he said. The quest to understand how the brain works is indeed a complex and vexing question—a great unknown that researchers like Uram are keen to explore.