New super imaging tool unveils unprecedented brain tissue resolution

Developed by Johann Danzl's group at ISTA (Institute of Science and Technology, Austria), LIONESS, which stands for Live Information Optimized Nanoscopy Enabling Saturated Segmentation, represents a significant leap in imaging brain activity.

"With LIONESS, for the first time, it is possible to get a comprehensive, dense reconstruction of living brain tissue. By imaging the tissue multiple times, LIONESS allows us to observe and measure the dynamic cellular biology in the brain take its course," said first author Philipp Velicky in a press release. 

"The output is a reconstructed image of the cellular arrangements in three dimensions, with time making up the fourth dimension, as the sample can be imaged over minutes, hours, or days."

One of LIONESS's strengths lies in its refined optics and two levels of deep learning— an artificial intelligence (AI) technique. 

Using minimal sample information, the first deep learning step fills in additional information on the brain tissue's structure in a process called Image Restoration. This approach achieves a gentle real-time imaging resolution of around 130 nanometers. 

The second deep learning step analyzes the complex imaging data and automates neuronal structure identification.

The tool successfully integrates functional measurements with structural observations by imaging calcium ion fluxes and measuring cellular electrical activity. It also benefited from the information provided by human cerebral organoids— miniature brain models that mimic human brain development. 

As the brain's structure and activity are highly dynamic, with constant changes occurring as it performs and learns new tasks, understanding this 'plasticity' is crucial to unraveling its mysteries.

Brain imaging techniques

Conventional techniques to get reconstructions of brain tissue include electron microscopy. This captures brain tissue at a few nanometers—a millionth of a millimeter—resolution by interacting with electrons. However, a sample must be fixed in one biological state, sectioning. Hence, no dynamic information can be obtained.

Another known technique is light microscopy, which can observe intact tissue without sectioning but lacks resolution for cellular details. 

That said, Super-Resolution Light Microscopy with SUSHI (Super-resolution Shadow Imaging) and STED Stimulated Emission Depletion) microscopy improved resolution but faces challenges due to potential tissue damage from intense light.

This is where LIONESS excels, as it has been designed to operate under "fast and mild" imaging conditions, according to its creators, while preserving the sample's life.

The complete study was published in Nature Methods and can be found here.