University of Pittsburgh researchers have identified an ultra-fast high-density emitter localization algorithm designed for high throughput super resolution localization microscopy. This innovative algorithm significantly reduces data analysis time, achieving a throughput three orders of magnitude higher than conventional algorithms while maintaining comparable localization accuracy and emitter recall rate.
Description
Portable nanoscope for single-molecule level imaging. Our miniSTORM is a small, self-contained, portable nanoscope that is composed of cutting- edge optical elements, sophisticated photoelectric devices and optimized analysis platforms to achieve the state-of-the-art single-molecule level imaging results. Its compact design and high efficient active anti-vibration system can guarantee its resolution up to 10 nm without the requirement of dedicated room or optical table. More important, its cost is only $35,000, over 10 times lower than existing options.
Applications
- Super resolution localization microscopy
- High throughput imaging
- Data analysis in biomedical research
- Real-time imaging applications
Advantages
This algorithm offers a single iteration, algebraic approach that is three orders of magnitude faster than conventional methods. It provides high localization accuracy and emitter recall rate, making it an ideal solution for high throughput imaging. The efficient GPU implementation further enhances its performance, enabling rapid data analysis and real-time imaging capabilities.
Invention Readiness
The technology is currently at the prototype stage. It has been developed and tested, demonstrating significant improvements in data analysis speed and throughput. The algorithm was implemented on a GPU device and tested using high-density emitter datasets. The results showed that the algorithm could process data at a throughput three orders of magnitude higher than conventional algorithms like DAOSTORM, while maintaining comparable localization accuracy and emitter recall rate. Further experimental validation involved comparing the performance of the algorithm with existing methods in terms of speed, accuracy, and recall rate. The next steps include optimizing the algorithm for different types of imaging data and validating its robustness and reliability in various high throughput imaging applications.
IP Status
https://patents.google.com/patent/US20240319097A1
