Rapid wavefront shaping using an optical gradient acquisition Journal: opticajournal Publication type: Research Article
Sagi Monin, 1 Marina Alterman, 1 and Anat Levin 1,* · 2025
Read Rapid wavefront shaping using an optical gradient acquisition Journal: opticajournal Publication type: Research Article by Sagi Monin, 1 Marina Alterman, 1 and Anat Levin 1,* (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-13711.txt.
Introduction
Sagi Monin, Marina Alterman, and Anat Levin speed up wavefront shaping, a method for imaging through cloudy tissue, in this 2025 arXiv paper. The method bends light to cancel the blur caused by uneven tissue. Finding the right bend is slow because the tissue structure is unknown, and most methods tweak one setting at a time.
The authors instead work out the full slope of the quality score, so every setting can be updated from the same measurements at once. They show this slope can be inferred from what the optical system measures, even without knowing the tissue. Cost no longer grows with the number of settings, so very fine corrections become practical. They prove the idea by sharpening images in a confocal microscope.
Catalog record
| CREATOR | Sagi Monin, 1 Marina Alterman, 1 and Anat Levin 1,* |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 42 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-13711
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Rapid wavefront shaping using an optical gradient acquisition Journal: opticajournal Publication type: Research Article Sagi Monin, 1 Marina Alterman, 1 and Anat Levin 1,* Abstract 1 Abstract Wavefront shaping systems enable deep imaging inside scattering tissue by correcting aberrations caused by tissue inhomogeneity. However, estimating optimal modulations remains challenging as such modulations depend on the unknown tissue structure. Most current techniques employ slow coordinate descent algorithms, which sequentially scan all modulation elements and query their values independently. Thus, their complexity scales prohibitively with the number of modulation parameters. We present a rapid wavefront shaping method that transitions from coordinate to gradient descent optimization, simultaneously updating all modulation parameters. Our approach employs a non-invasive, guide-star-free score function to quantify modulation effectiveness. We derive an analytical framework expressing the score’s gradient with respect to all parameters. Although this gradient depends on the unknown tissue structure, we demonstrate how it can be inferred from optical measurements. This method enables rapid computation of high-resolution wavefront corrections, with complexity independent of the number of modulation parameters. High resolution modulations enable superior correction in thicker tissue samples. We demonstrate our framework’s efficacy in correcting aberrations in a coherent confocal microscope. 2 Introduction Optical imaging of tissue is challenging because cells and other tissue components scatter light. As a result, as light propagates deeper into tissue, it becomes heavily aberrated, making it challenging to resolve clear images of deeper structures. Over the decades, several techniques, such as confocal microscopy [ 1 ] and optical coherence tomography (OCT) [ 2 ] , have been developed to achieve deeper imaging by filtering out scattered photons and isolating ballistic photons to form an image. Despite this progress, these techniques are inherently limited to thin layers as the number of ballistic photons decays rapidly within scattering materials. To overcome this challenge, wavefront shaping was introduced. Instead of filtering out the scattered photons, wavefront shaping aims to measure the scattered light and invert the process. Such corrections follow two major approaches: digital and optical. Digital aberration correction methods attempt to illuminate the tissue using a set of wavefronts and measure the scattered light. This data is fitted with a parametric model to estimate the aberration and reconstruct the hidden target [ 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 ] . This approach has led to impressive results, but it is ultimately limited by signal-to-noise ratios (SNR). Even with advanced gating strategies, light scattered by deeper tissue components is weak and can be lost in measurement noise. An alternative class of techniques uses optical wavefront-shaping [ 22 , 23 , 24 , 25 , 26 ] . Such techniques use spatial light modulators (SLMs) to control the incoming and/or outgoing paths of the optical system. These SLMs reshape the incoming and/or outgoing wavefront in a way that is the inverse of the aberration it undergoes inside the tissue. This allows an incoming wavefront to focus to a diffraction-limited spot, and the light emitted from a single point within the tissue can be focused onto a single detector element. Unlike digital aberration corrections, the significant advantage of optical wavefront shaping is that the correction is physical; hence, all light photons emerging from a single target point can be brought into a single sensor point and measured with a significantly higher SNR. Wavefront shaping ideas have found applications in a wide range of imaging modalities, including sound [ 24 , 27 ] and light, coherent imaging and OCT [ 28 ] , and incoherent fluorescence imaging using single-photon [ 29 , 30 , 31 , 32 ] and multi-photon [ 33 , 22 ] excitation. This work showcases a simple aberration correction in a coherent, reflection mode confocal microscope. However, our framework also applies to other popular imaging schemes such as OCT and fluorescent microscopy. The practical application of wavefront shaping is hindered by the complexity of determining the desired modulation correction. This wavefront correction differs between different tissue samples, and vary even within different regions of the same tissue sample. Earlier proof-of-concept demonstrations employed a validation camera behind the tissue to provide feedback for the algorithm [ 34 , 35 , 36 , 37 , 38 , 39 ] . Other approaches utilize a guide star [ 24 , 40 , 33 , 41 , 42 , 43 , 28 , 44 , 45 , 46 , 47 ] , where scattering originates from a strong single point source within the tissue, allowing a wavefront sensor [ 47 , 42 ] to directly measure the scattered wavefront. In the absence of such a guide-star, determining a wavefront shaping correction is a significantly more challenging task. Most approaches define a score function to evaluate the quality of the modulation based on the captured signal, and then optimize the SLM parameters to maximize this score. A core challenge for such optimization algorithms lies in the unknown structure of the tissue. This structure can only be probed by projecting multiple modulation patterns and imaging their scattered output. Thus, most wavefront shaping optimization schemes employ coordinate-descent (CD) approaches, where a single parameter of the modulation is varied at a time. Each optimization step involves selecting a new value for one parameter to improve the modulation score. These algorithms iterate sequentially through all degrees of freedom [ 34 , 48 , 35 , 33 , 39 , 29 , 31 ] . The main drawback of coordinate descent schemes is their time complexity, which scales with the number of free parameters in the modulation. For thick tissue, the modulation should ideally use all pixels on the SLM, usually in the megabyte range. In this research, we derive a fast approach for estimating a wavefront-shaping correction in a non-invasive, guide-star free setting. Our algorithm is capable of simultaneously updating all SLM parameters from the information in one measurement , and explicitly optimize a desired score function. To this end, we start by analytically differentiating the wavefront-shaping score. Although the gradient depends on the unknown tissue structure, we show we can use optical computing to measure it. By simply capturing the back scattered field, the gradient can be evaluate in closed-form. The dimensionality of the gradient vector is equivalent to the number of SLM parameters. With this gradient at hand, we can simultaneously update all SLM parameters, and we can transition from slow coordinate descent to fast gradient descent optimization. Our gradient descent scheme is significantly faster than existing coordinate descent schemes and can also recover better modulations. Coordinate descent methods typically restrict the number of SLM modes they optimize due to computational complexity and since the measurement of high-frequency modes is noise sensitive. However, in thick tissue where the scattering is wide, the optimal modulation requires a large number of degrees of freedom. In contrast, our approach can optimize a large number of modes without an increase in computational complexity. We show that this higher number of modes allows us to find significantly better modulations. Figure 1: System Schematic: (a) Schematic diagram of our system consisting of two SLMs. The