Borrowing 'Star-Pattern Matching,' a Finnish Team Lets a Microscope Find Its Way Back to the Original Field of View
Borrowing 'Star-Pattern Matching,' a Finnish Team Lets a Microscope Find Its Way Back to the Original Field of View
People doing microscopy experiments often face this trouble: after finally finding an interesting region in a sample, once they change the magnification, move the sample to another microscope, or subject it to further processing, they cannot find that field of view again. Re-searching by hand is both time-consuming and prone to error.
To address this problem, an international study involving the University of Turku, Finland, has offered a clever solution—"borrowing" an idea from astronomy.
1. Cell-Nuclei Clusters as 'Constellations'
When astronomers recognize the starry sky, they compare an unknown star field with a star chart using the relative positions among several stars—this is "star-pattern matching." The research team transplanted this method under the microscope: treating a cluster of several cell nuclei in a sample as a "constellation" in the night sky and using the arrangement among the nuclei as a "fingerprint." On this basis they developed an open-source tool named NucleiSky to search a larger microscopic image for regions with the same arrangement features. In theory, the relative positions of just 5 cell nuclei are enough to lock onto the correct field of view.
2. Not by 'Looks,' but by 'Arrangement'
NucleiSky's cleverness lies in recognizing the arrangement pattern among cell nuclei rather than the overall appearance of the image. Therefore, even if the magnification changes, the imaging mode changes, or even the microscope itself is different, so long as the arrangement of the nuclei is unchanged, the tool can still find the original region back. This makes it especially useful for "hand-off" observation across different instruments and different imaging conditions.
3. Working with Intelligent Microscopy Systems
The research team also tested NucleiSky in an intelligent microscopy application: first imaging a specific region of a cell sample with one microscope, then transferring the sample to another microscope using a different imaging mode. The second microscope, equipped with an intelligent workflow, first automatically scans the sample at low magnification and recognizes the previous cell-nucleus "constellation," then automatically navigates back to the original observation region and images it at higher resolution. The whole process requires almost no manual intervention.
4. Significance
The idea behind this method is not simply "finding the field of view again." The researchers believe it can reduce dependence on manual searching and make complex microscopy experiments more convenient and more reproducible; in the future it could also be used in intelligent microscopy systems, allowing a microscope to autonomously scan a sample, discover regions of interest, and automatically return for high-precision imaging. The related paper has been published in the Journal of Cell Science.
Applying an astronomical method to the microscope may seem cross-disciplinary, but it points to the same problem: how to reliably "locate" within an image. For increasingly automated microscopy, such a tool is of no small significance.