The evolution of experimental superresolution microscopy has been accompanied by the development of advanced computational imaging capabilities. Recently introduced, quantum image scanning microscopy (Q-ISM) has successfully harnessed quantum correlations of light to establish an improved viable imaging modality that builds upon the preceding image scanning microscopy (ISM) superresolution method. While offering improved resolution, at present the inherently weak signal demands exhaustively long acquisition periods. Here we exploit the fact that the correlation measurement in Q-ISM is complementary to the standard ISM data, acquired simultaneously, and demonstrate joint sparse recovery from Q-ISM and ISM images. Reconstructions from images of fluorescent quantum dots are validated through correlative electron microscope measurements, and exhibit superior resolution enhancement as compared to Q-ISM images. In addition, the algorithmic fusion facilitates a drastic reduction in the requisite measurement duration, since low signal-to-noise-ratio Q-ISM measurements suffice for augmenting ISM images. Finally, we obtain enhanced superresolved reconstructions from short scans of a biological sample labeled with quantum dots, demonstrating the potential of our method for quantum imaging in life science microscopy.
Abstract In quantum gas microscopy experiments, reconstructing the site-resolved lattice occupation with high fidelity is essential for the accurate extraction of physical observables. For short interatomic separations and limited signal-to-noise ratio, this task becomes increasingly challenging. Common methods rapidly decline in performance as the lattice spacing is decreased below half the imaging resolution. Here, we present an algorithm based on deep convolutional neural networks to reconstruct the site-resolved lattice occupation with high fidelity. The algorithm can be directly trained in an unsupervised fashion with experimental fluorescence images and allows for a fast reconstruction of large images containing several thousand lattice sites. We benchmark its performance using a quantum gas microscope with cesium atoms that utilizes short-spaced optical lattices with lattice constant 383.5 nm and a typical Rayleigh resolution of 850 nm. We obtain promising reconstruction fidelities ≳ 96% across all fillings based on a statistical analysis. We anticipate this algorithm to enable novel experiments with shorter lattice spacing, boost the readout fidelity and speed of lower-resolution imaging systems, and furthermore find application in related experiments such as trapped ions.
Quantum gas microscopy provides single-shot access to many-body configurations, yet has so far relied on pinning atoms to a discrete lattice during fluorescence readout. Here, we realize ultrafast atom-resolved fluorescence microscopy of an ultracold
162Dy gas directly in free space, with all external confinement removed during imaging. Exploiting the large mass and broad optical transition of dysprosium together with a high numerical aperture in-vacuum objective, we detect single atoms within
4μs while maintaining an effective point-spread function with a full width at half maximum of
254.8(3)nm. We achieve a single-atom conditional classification fidelity of
99.72(3)% for identified fluorescence candidates. We use the measured single-atom response to construct a statistical configuration-inference model that recovers atomic positions and multiplicities, coming from overlapping atoms, directly from single fluorescence snapshots. We further benchmark our reconstruction algorithm through end-to-end Monte Carlo simulations with known ground-truth atomic configurations, obtaining a reconstruction fidelity of
98.7%. Our approach, applicable to a wide variety of atomic species, extends quantum gas microscopy beyond predefined detection lattices, enabling atom-resolved measurements of counting statistics and higher-order spatial correlations directly on the intrinsic length scales of continuum many-body systems.