University of Pittsburgh researchers have developed an algorithm for use in marker-free drift tracking during high-throughput localization microscopy. The algorithm, referred to as Adaptive localization Intersection based Drift correction (AID) is designed to provide fast and precise drift correction for single molecule localization microscopy (SMLM).
Description
For the last decade significant efforts have been made to improve the resolution limit of SMLM. Technological advances have largely improved resolution, although drift error continues to limit the image resolution. This approach aims to compensate for sample drift during image acquisition using a fast and precise algorithm without the need for additional markers or optical components.
Applications
• Improving accuracy in single molecule localization microscopy
• Marker-free microscopy
• High resolution of biological samples
Advantages
A major limiting factor in high-throughput super-resolution SMLM is sample drift. Current marker-free drift correction methods lack the speed and precision needed to correct high frequency drift. These challenges have led many researchers to use marker-assisted approaches to improve precision. However, marker-assisted approaches require complex sample preparation and additional imaging channels, adding to workflow and time requirements.
This approach is the first marker-less algorithm that enables real-time drift-free high-throughput localization imaging. AID overcomes these challenges of traditional marker-free SMLM by reducing drift resolution degradation, and improves the speed more than 10-fold. AID has the potential to improve SMLM resolution instantly by replacing existing marker-less drift correction approaches.
Invention Readiness
A prototype has been developed. Simulation and biological experiments have shown a sub-nanometer (~0.1 nm) precision and sub second (0.2 s) tracking interval, comparable with the best marker-assisted approaches. Testing has also shown that this can be used on consumer-level computers, making this approach readily available to research laboratories without the need for specialist hardware.
IP Status
https://patents.google.com/patent/WO2024097628A2
