Scattering Correction through Fourier-Domain Intensity Coupling in Two-Photon Microscopy

Daniel Zepeda, Yucheng Li, Yi Xue*,
University of California, Davis,
Paper arXiv

Abstract

Light penetration depth in biological tissue is limited by tissue scattering. Correcting scattering becomes particularly challenging in scenarios with limited photon availability and when access to the transmission side of the scattering tissue is not possible.

Here, we introduce a new two-photon microscopy system with Fourier-domain intensity coupling for scattering correction (2P-FOCUS). 2P-FOCUS corrects scattering by intensity modulation in the Fourier domain, leveraging the nonlinearity of multiple-beam interference and two-photon excitation, eliminating the need for a guide star, iterative optimization, or measuring transmission or reflection matrices. 2P-FOCUS can also correct scattering beyond the limitation of the memory effect by automatically customizing correction masks for each subregion in a large field-of-view.

We provide several proof-of-principle demonstrations here, including focusing and imaging through a bone sample, and imaging neurons and cerebral blood vessels in the mouse brain ex vivo. 2P-FOCUS significantly enhances two-photon fluorescence signals by several tens of folds compared to cases without scattering correction at the same excitation power. 2P-FOCUS can also correct tissue scattering over a 230 x 230 x 500 µm3 volume, which is beyond the memory effect range. 2P-FOCUS is able to measure, calculate, and correct scattering within a few seconds, effectively delivering more light deep into the scattering tissue. 2P-FOCUS could be broadly adopted for deep tissue imaging owing to its powerful combination of effectiveness, speed, and cost.

Overview

In 2P-FOCUS, we use the digital micromirror device (DMD) to modulate the intensity at the back aperture of the two-photon microscope to segment the incident collimated beam into multiple narrow beams, which generate multiple-beam interference within the scattering tissue.


The process of generating a mask on DMD for scattering correction consists three steps:

   1) Random intensity patterns are projected in the Fourier domain while monitoring the fluorescence intensity excited under this modulation.

   2) The random masks that generate bright fluorescence are selected by thresholding the fluorescence intensity detected by the PMT. Then, a gray-scale correction mask is generated by summing these selected random patterns.

   3) The final binary correction mask for DMD projection is generated by thresholding the intensity of the grayscale correction mask.


2P-FOCUS Principle and schematic diagram

a, Without correction, the incident light is scattered and cannot form a tight focus. b, Three steps of the process of generating a correction mask. c, With correction, the correction results in a brighter focus compared to the case before correction. d, Optical schematic diagram of 2P-FOCUS.

Results

We demonstrate the performance of 2P-FOCUS with proof-of-principle experiments, including focusing and imaging through approximately 200 μm-thick bone, and imaging fluorescence-labeled neurons and blood vessels up to 500 μm deep in the mouse brain ex vivo. 2P-FOCUS achieves a 3-60 fold enhancement in fluorescence intensity compared to standard two-photon microscopy.

Focusing through a 200 μm-thick bone with 2P-FOCUS

We first demonstrate 2P-FOCUS experimentally by focusing through a chicken bone onto a homogeneous fluorescence slide.

a, Representative random patterns with a sparsity of 0.4 and a super-pixel radius of 8 pixels. b, Fluorescence intensity corresponding to the 2,500 patterns detected by the PMT is shown by the blue line. Random patterns contributing to the top 10% of fluorescence intensity are selected, as indicated by measurements above the red line. c, The sum of the selected random binary patterns forms the grayscale correction mask. d, A binary correction mask is generated by binarizing the grayscale correction mask. e, Photo of the bone sample and the fluorescence sample. f, Zoomed-in image of the focus before correction, where no focus is identified. g, Zoomed-in image of the focus after correction, showing a clear focus with a peak value of 629. h, Comparison of the intensity profile along the x-axis before (blue line) and after (red line) correction. i, Fluorescence intensity before (blue dashed line) and after (blue solid line) correction, as well as the intensity ratio (red line), as functions of the laser power on the sample. j, Fluorescence intensity before (blue dashed line) and after (red line) correction as a function of the sparsity of random patterns used for measurements. k, Fluorescence intensity before (blue) and after (red) correction as a function of the size of super-pixels. l, Fluorescence intensity before (blue dashed line) and after (red line) correction as a function of the number of measurements.




Imaging fluorescence beads through a bone beyond the memory effect range

To image beyond the memory effect range, we developed a new strategy called “subregion correction” that synchronizes the scanner and the DMD to project different masks to different region during the scanning process. We experimentally compared global correction and subre- gion correction side-by-side by imaging red fluorescent beads through a piece of chicken bone.

a, Schematic diagram of global correction. b, Schematic diagram of subregion correction. c, Schematic diagram of the sample used in this experiment. d-f, d-f Imaging fluorescent beads through highly scattering bone across a 230 × 230 μm2 field-of-view. d, The image taken without scattering correction. e, The image taken under global correction. f, The image taken under subregion correction. g, The intensity profile of the two clusters of beads indicated by arrows in d-f.

Imaging neurons deep in the brain


Video of fluorescence-labeled neurons at different axial planes


Imaging fluorescence-labeled neurons deep in the mouse brain using 2P-FOCUS.

a, The volumetric view of parvalbumin (PV) interneurons expressing cell-fill tdTomato imaged by 2P-FOCUS ex vivo. b, The top plane of the image stack is below the surface of the brain. c, The bottom plane of the image stack, 450 µm deep from b, shows some cell bodies. d, Schematic diagram illustrating global correction and the correction mask used for e-f. e-f, Images of the same region at 250 µm depth e before and f after global correction. g, Schematic diagram for subregion correction and the four subregions in h-i. h-i, Images of the same region at 400 µm depth h before and i after subregion correction. j, The correction masks for four subregions used to capture image i. k, Comparison of the intensity profile of representative neurons (pointed by the blue arrow in e and the red arrow in f) at 250 µm depth before (blue line) and after (red line) correction. l, Comparison of the intensity profile of representative neurons (indicated by the blue arrow in h and the red arrow in i) at 400 µm depth before (blue line) and after correction (red line).




Imaging cerebral blood vessels deep in the mouse brain


Video of blood vessels with fluorophore injection at different axial planes


Imaging blood vessels with intravascular fluorophore injection deep in the mouse brain using 2P-FOCUS.

a, Volumetric view of cerebral blood vessels with intravascular FITC-dextran injection, imaged by 2P-FOCUS ex vivo. b, Maximum intensity projection (MIP) of the top 100 μm-thick volume along the z axis. c, MIP of the bottom 100 μm-thick volume along the z axis. d, Two-photon image of blood vessels at 340 μm depth without correction. e, Image of the same region as in d after subregion correction. f-g, Zoomed-in view of the regions in the dashed box in d-e h, The location of the 4 subregions on the image plane. i, The correction masks used in the experiment when acquiring image e. j, Comparison of the fluorescence intensity profile before (yellow line) and after (red line) correction along the dashed line in f-g.

BibTeX

@misc{zepeda2024scatteringcompensationfourierdomainopenchannel,
      title={Scattering compensation through Fourier-domain open-channel coupling in two-photon microscopy}, 
      author={Daniel Zepeda and Yucheng Li and Yi Xue},
      year={2024},
      eprint={2401.15192},
      archivePrefix={arXiv},
      primaryClass={physics.optics},
      url={https://arxiv.org/abs/2401.15192}, 
}