alt FUW
logo UW
other language
webmail
search
menu

Seminarium Optyczne

sala 0.06, ul. Pasteura 5
2026-03-19 (10:15) Calendar icon
Piotr Zdańkowski (Faculty of Mechatronics, Technical University of Warsaw)

Computational Microscopy: New Approaches to Label-Free Quantitative Imaging

Traditional optical microscopy frequently struggles against the physical barriers of the Abbe diffraction limit and the inevitable trade-off between sample invasiveness and image contrast. To overcome these hurdles, the emerging field of computational microscopy merges innovative optical architectures with powerful numerical reconstruction algorithms. At the Quantitative Computational Imaging Lab (QCI Lab), we focus on developing stable, label-free imaging methods, primarily Quantitative Phase Imaging (QPI) and Fourier Ptychographic Microscopy (FPM). Our QPI systems are based on common-path interferometry, a robust approach that drastically reduces environmental phase noise by using amplitude and liquid-crystal polarization gratings as beam splitters. Unlike classic two-beam Mach-Zehnder setups, these naturally achromatic configurations allow both the reference and object beams to travel the same optical path. This unique architecture enables the use of low-coherence illumination, virtually eliminating parasitic speckle noise. I will showcase novel computational frameworks we have developed to maximize the potential of these optical setups. Furthermore, I will demonstrate how our polarization-guided holotomographic techniques enable precise, label-free biochemical differentiation of lipid droplets based purely on subtle differences in their refractive indices. Another method that we develop at QCI Lab is Fourier Ptychographic Microscopy (FPM). This computational technique synthesizes a high numerical aperture in the spatial frequency domain through sequential angle-varied LED illumination, effectively breaking the traditional compromise between a wide field of view and high spatial resolution. To make FPM more accessible and resistant to experimental errors, I will introduce our open-source "FPM app" for advanced phase and amplitude reconstruction. Alongside this software, I will highlight a robust, two-step automated hardware calibration method that precisely corrects LED translatory and rotational misalignments without the need for any specialized calibration targets. Lastly, we have recently built a novel FPM system based on micro LED array together with the new framework for the ptychographic reconstruction that fully models illuminating wave sphericity,

Wróć

Wersja desktopowa Stopka redakcyjna