Improving the axial and lateral resolution of three-dimensional fluorescence microscopy using random speckle illuminations

Abstract

We consider a fluorescence microscope in which several three-dimensional images of a sample are recorded for different speckle illuminations. We show, on synthetic data, that by summing the positive deconvolution of each speckle image, one obtains a sample reconstruction with axial and transverse resolutions that compare favorably to that of an ideal confocal microscope.

Joint Reconstruction Strategy for Structured Illumination Microscopy With Unknown Illuminations

“…On the one hand, the quadratic penalty in (9) was mostly introduced to ensure that the minimizer defined by (8) is unique (via strict convexity of the criterion). However, because high-frequency components in qm are progressively damped as β increases, the latter parameter can 10 A rat hippocampal neuron in culture labelled with an anti-βIV-spectrin primary and a donkey anti-rabbit Alexa Fluor 647 secondary antibodies, imaged by STORM and processed similarly to [28]. also be adjusted in order to prevent an over-amplification of the instrumental noise.…”

“…Obviously, if the observation model H is also a BCCB matrix built from the discretized OTF, the choice C = H t H in (28) leads to B = (∇ 2 g) −1 for a = 1. Such a preconditioner is expected to bring the fastest asymptotic convergence since it corrects the curvature anisotropies induced by the regular part g in the criterion (10). The PPDS pseudo-code for solving the joint Blind-SIM problem is given in Algorithm 1.…”

Computational structured illumination for high-content fluorescence and phase microscopy

“…Unfortunately, the reconstruction formulation proposed in that work is especially ill-posed due to randomness between the illumination patterns, i.e., if N img raw images are taken, there would be N img + 1 unknown variables to solve for (N img illumination patterns and 1 sample distribution). To better condition this problem, priors based on speckle statistics [46,47,49,50,52] and sample sparsity [48,51] can be introduced, pushing blind SIM to 2× resolution gain. However, to implement high-content microscopy using SIM, we desire a resolution gain of > 2×.…”

“…Our method is related to blind SIM [46]; however, instead of using many random speckle patterns (which restricts resolution gain to ∼1.8×), we translate the speckle laterally, enabling resolution gains beyond that of previous methods [46][47][48][49][50][51][52] (see Appendix D). Previous works also use high-cost spatial-light-modulators (SLM) [53] or galvonemeter/MEMs mirrors [41,54] for precise illumination, as well as expensive objective lenses for aberration correction.…”

Numerical approach for reducing out-of-focus light in bright-field fluorescence microscopy and superresolution speckle microscopy

“…The minimization is performed following a standard gradient algorithm as detailed in Ref. [7]. More precisely, at iteration i, ξ is modified along…”