New Processing Technique Enhances Super-Resolution Microscope Images

In a study published in Science Advances, a team of bioengineering researchers at the University of Illinois Urbana-Champaign developed an algorithm known as adaptive intersection maximization, or AIM. This algorithm removes high-frequency noise from super-resolution optical microscope data much faster than standard methods, resulting in much higher image resolutions.

When measuring molecular structures with nanoscale precision, every piece of noise is picked up in the data: someone walking past the microscope, minor vibrations in the building, and even traffic outside. A novel processing approach reduces noise from optical microscope data in real-time, allowing scientists to follow individual molecules with nearly ten times the precision previously attainable.

Thanks to the algorithm, scientists will be able to examine chemical and biological processes far more quickly and accurately than they could in the past.

At first, we just wanted to develop a fast algorithm because our lab produces too much data for traditional algorithms to handle, but we found that AIM can also achieve sub-nanometer precision, which is unheard of in our field. In addition, it doesn’t require immense computing power like traditional tools. It can run on a laptop. We want to make this a plug-and-play tool for all microscope users.

Hongqiang Ma, Study Lead Author and Research Professor, University of Illinois Urbana-Champaign

In recent decades, the single-molecule localization microscopy technique has allowed scientists to observe molecular-scale structures, overcoming what was previously regarded as a fundamental limitation of optical microscopes.

However, it is hampered in practice by unpredictable noise, or “drift,” which effectively blurs the images and prevents super-resolution microscopy from attaining its full resolution.

Single-molecule localization actually uses a fairly simple instrument, but the tricky part that really impacts image resolution is drift. Many researchers only remove low-frequency drift. Removing the high-frequency drift–minute vibrations caused by environmental noise–is computationally intensive and requires large amounts of time and resources.

Yang Liu, Project Lead and Professor, University of Illinois Urbana-Champaign

The mathematical correlations between image frames serve as the foundation for standard techniques for eliminating drift. Even with supercomputing resources, Liu states that the amount of image data produced by her lab’s microscopes is so great that image correlation techniques take days.

AIM also compares neighboring frames, but it does so by locating each data point in the center of a circle (determined by localization precision) and searching for points within that circle in other frames. Overlapping locations inside the “radius of intersection” are grouped into a single localization.

The process is then repeated using the condensed points. These stages need minimal computational resources and are faster than the acquisition time of a microscope camera. As a result, drift-corrected images can be generated in real-time.

The researchers tested AIM on simulated data and well-defined DNA origami structures. The algorithm effectively localized the structures, and the level of accuracy, less than 1 nanometer, was far higher than that of standard image correlation approaches, which were around 10 nanometers.

Liu’s group will incorporate AIM into high-throughput microscopy methods being developed to improve disease diagnosis. However, Liu hopes the algorithm will have applications in biology and bioengineering.

Liu added, “It is a fast and easy-to-use tool, and we want to make it widely accessible for the entire community. We are making our software publicly accessible. We want people to get the boost in their image resolution just from this one bit of post-processing.”

Maomao Chen of the University of Pittsburgh and Phuong Nguyen of Illinois also contributed to this study.

Liu is affiliated with the Department of Electrical & Computer Engineering and the Cancer Center in Illinois. Ma and Nguyen are also associated with the Beckman Institute for Advanced Science and Technology at Illinois.

National Institutes of Health supported the study.

Journal Reference:

Ma, H., et. al. (2024) Toward drift-free high-throughput nanoscopy through adaptive intersection maximization. Science Advances. doi:10.1126/sciadv.adm7765

Source: http://illinois.edu/