We recently discovered a phenomenon that the digitally summed Fourier spectrum of two images acquired from two-angle illumination exhibits interference-like fringe modulation when the sample is out-of-focus. These digital fringes correlate directly with defocus through a physics-based relation. Based on this principle, we developed an automatic, efficient, and generalizable defocus detection method termed digital defocus aberration interference (DAbI).

We propose Fourier Ptychotomography (AFP), a new computational microscopy technique that analytically reconstructs aberration-free, complex-valued 3D RI distributions without iterative optimization or axial scanning. AFP incorporates a new concept named finite sample thickness (FST) prior, thereby simplifying the inverse scattering problem into solving linear equations.

Single-shot volumetric fluorescence (SVF) imaging captures biological processes with high temporal resolution and a large field of view, unlike traditional methods requiring multiple axial plane scans. Existing SVF methods often face limitations due to large, complex point spread functions (PSFs), affecting signal-to-noise ratio, resolution, and field of view. The paper introduces a QuadraPol PSF combined with neural fields, using a compact custom polarizer and a polarization camera to detect fluorescence and encode the 3D scene within a compact PSF without depth ambiguity.

Brain metastases can occur in nearly half of patients with early and locally advanced (stage I–III) non-small cell lung cancer (NSCLC). There are no reliable histopathologic or molecular means to identify those who are likely to develop brain metastases. We sought to determine if deep learning could be applied to routine H&E-stained primary tumor tissue sections from stage I–III NSCLC patients to predict the development of brain metastasis.

Fourier ptychographic microscope images the biological samples with high-resolution and large field-of-view simultaneously. However, this microscope faces challenges with long image stack reconstruction time and huge data volumes. We designed physics-based neural signal representations to tackle these challenges and showed potential in facilitating remote diagnosis, digital pathology, and efficient clinical data packaging.