🔬 optics

Two-photon light-sheet live imaging at kilohertz frame rate using birefringence-based pulse splitting

This paper introduces a compact, birefringence-based pulse splitting scheme that optimizes two-photon light-sheet microscopy by enabling precise control over temporal excitation profiles, thereby achieving kilohertz frame rates and high pixel rates with minimal photodamage in live zebrafish embryos.

Original authors: Lei Zhu, Dale Gottlieb, Vincent Maioli, Antoine Hubert, Frédéric Druon, Pierre Mahou, Emmanuel Beaurepaire, Willy Supatto

Published 2026-06-30

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Lei Zhu, Dale Gottlieb, Vincent Maioli, Antoine Hubert, Frédéric Druon, Pierre Mahou, Emmanuel Beaurepaire, Willy Supatto

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to take a high-speed video of a hummingbird’s wings or a fish’s beating heart. To see the details clearly, you need to take thousands of photos every second. But here is the problem: the light used to take those photos is like a hot spotlight. If you shine it too brightly or for too long, you cook the subject (photodamage) or bleach out the colors (photobleaching), ruining the image and hurting the living tissue.

This paper introduces a clever optical trick to solve this problem, allowing scientists to film live zebrafish embryos at incredibly high speeds without hurting them.

The Core Problem: The "Slow" Laser

Standard lasers used in microscopy fire pulses of light very fast—like a machine gun shooting bullets. However, even these "fast" lasers aren’t fast enough for the highest-speed imaging needed to see rapid biological processes. If you try to speed up the camera to match the action, you either get a dark, grainy image, or you have to crank up the laser power, which damages the sample.

The Solution: The "Pulse Splitter"

The researchers created a compact, low-cost device they call a pulse splitter. Think of it like a magic prism that doesn’t just bend light, but clones it in time.

Here is the analogy:

Normal Laser: Imagine a drummer hitting a drum once every millisecond. Boom... boom... boom.
The Pulse Splitter: The researchers put special crystals (made of Yttrium Vanadate) in the path of the laser beam. These crystals act like a time-delay mirror. When the single laser pulse hits them, it splits into a tight cluster of smaller pulses that arrive almost instantly after one another.
The Result: Instead of one boom, you get a rapid-fire brrrrt (a burst of 2 or 4 pulses) for every single original pulse.

This effectively doubles or quadruples the "frame rate" of the light hitting the sample without needing a new, expensive, or more powerful laser. It’s like getting four free shots for the price of one, all packed into a tiny fraction of a second.

Why This Works Better

You might ask, "If you hit the sample with more pulses, won’t that damage it more?"

The paper shows that it actually doesn’t, thanks to two key optimizations:

The Right Color (Wavelength): They switched the laser color from 1030 nm to 1070 nm. Why? Because water absorbs less heat at 1070 nm. It’s like switching from a microwave that heats water efficiently to one that lets the water stay cooler. This reduces the "cooking" effect on the fish.
The Right Timing: By splitting the pulses, they can keep the peak intensity of the light lower while still getting enough total signal. The paper proves that this "burst" method causes no more damage than if the pulses were spread out evenly. In fact, the threshold for damage stays the same, but the image quality improves.

The Results: Filming Life in Super-Slow Motion

Using this setup, the team filmed live zebrafish embryos. Here is what they achieved:

Speed: They reached imaging speeds of over 1,000 frames per second (kilohertz range). This is like watching a bullet in slow motion.
Detail: They could see the beating heart of the fish in real-time, capturing the movement of the atrio-ventricular canal.
Brain Activity: They also filmed calcium signals in the fish’s brain using red fluorescent markers. This allowed them to see neurons firing in real-time.
Safety: Despite the high speed, the fish remained healthy. The heat generated was less than 1°C, and the light didn’t bleach out the fluorescent colors significantly.

In Simple Terms

The researchers built a "time-cloning" lens for their microscope. By splitting each laser pulse into a tiny burst of multiple pulses, they could take photos much faster than before. By also tuning the laser to a specific color that doesn’t heat up water as much, they were able to film the rapid beating of a fish’s heart and the firing of its brain cells in stunning detail, without harming the tiny creature.

It’s a simple, cheap, and effective way to make microscopes faster and gentler, opening the door to seeing fast biological processes that were previously too quick or too delicate to capture.

Technical Summary: Two-photon light-sheet live imaging at kilohertz frame rate using birefringence-based pulse splitting

Problem Statement
Multiphoton microscopy is a standard tool for imaging live, intact tissues, but its speed is fundamentally limited by fluorophore emission rates and the thresholds for photodamage. While multi-point acquisition schemes, particularly two-photon (2P) light-sheet microscopy, offer a path to high-speed imaging by parallelizing excitation, optimizing the signal-to-photodamage ratio requires precise control over both excitation wavelength and laser pulse frequency. Conventional femtosecond lasers typically do not allow for the independent modulation of these parameters. Specifically, optimizing 2P light-sheet imaging necessitates tuning the wavelength to minimize water absorption-mediated heating and adjusting the pulse frequency to balance signal intensity against nonlinear photodamage. Existing methods for modulating the temporal excitation profile, such as pulse picking or time gating, often involve complex optical setups or significant energy losses. Therefore, there is a need for low-cost, efficient, and robust strategies to modulate the temporal profile of wavelength-tunable pulsed lasers.

Methodology
The authors introduce a compact, passive pulse-splitting scheme based on cascaded birefringent crystals. The system utilizes Yttrium Vanadate ( $YVO_4$ ) crystals, selected for their high birefringence, transparency in the near-infrared range, and low group-velocity dispersion. By propagating a short laser pulse through these crystals with specific thicknesses and orientations, the input pulse is converted into an adjustable sequence of collinear sub-pulses with picosecond-scale delays.

The setup allows for the generation of 1, 2, or 4 collinear sub-pulses per burst by rotating the optical axes of the crystals and a half-wave plate. This effectively increases the average pulse frequency without changing the laser source’s native repetition rate. For instance, a 4 MHz laser source can be converted to an effective 8 MHz (2× splitting) or 16 MHz (2×2× splitting) average frequency. The system is integrated into a 2P light-sheet microscope equipped with a wavelength-tunable laser (operating at 1030 nm or 1070 nm). The authors optimized imaging parameters for red fluorophores (mCherry and jRGECO1b) in zebrafish embryos, selecting 1070 nm to exploit a local minimum in water absorption and maximize mCherry two-photon absorption.

Key Contributions

Development of a Birefringence-Based Pulse Splitter: A straightforward, alignment-free, and nearly lossless method to increase the effective pulse repetition rate of a laser source using cascaded $YVO_4$ crystals.
Optimization of Illumination Parameters: Demonstration that combining wavelength tuning (1070 nm) with pulse splitting (to achieve ~16 MHz average frequency) optimizes the trade-off between signal strength, linear heating, and nonlinear photodamage for red fluorophores.
High-Speed In Vivo Imaging: Achievement of kilohertz frame rates in live zebrafish embryos, enabling the capture of fast biological dynamics such as heartbeats and neuronal calcium signaling.

Results

Pulse Splitting Performance: The $YVO_4$ crystals introduced minimal dispersion (pulse broadening limited to ~60 fs for 30 mm of crystal) and maintained >98% transmission. The pulse delays matched theoretical predictions (approx. 0.7 ps/mm).
Image Quality and Photobleaching: The pulse splitter did not degrade spatial resolution (axial FWHM remained ~3.5 µm) or 2PEF signal intensity compared to native high-frequency lasers. Crucially, photobleaching rates were not increased by the burst-like pulse structure, as the inter-pulse delays (7.6–15.2 ps) are significantly shorter than the fluorescence lifetime.
Thermal and Nonlinear Damage: Increasing the wavelength from 1030 nm to 1070 nm reduced linear heating (measured via heart beat rate changes) by a factor of 0.68. Nonlinear photodamage thresholds followed expected scaling laws with pulse frequency, confirming that the damage threshold is independent of whether pulses are evenly distributed or delivered in bursts.
Imaging Performance: Using 1070 nm illumination and a 2×2×4 MHz (16 MHz average) pulse configuration, the authors achieved:
- Heart Imaging: Volumetric imaging of the beating heart at 1052 Hz frame rate (147 MHz pixel rate) and 465 Hz frame rate (163 MHz pixel rate).
- Neuronal Imaging: Calcium dynamics imaging using jRGECO1b at 465 Hz frame rate (165 MHz pixel rate).
- Signal Levels: Achieved >10 photons per pixel with a signal-to-background ratio of ~10, using a mean laser power of ~100 mW and peak intensity of ~0.1 TW/cm².

Significance
The paper claims that this compact and adjustable pulse-splitting scheme allows for the full exploitation of light-sheet illumination for fast in vivo 2P imaging. It facilitates the optimization of illumination parameters (wavelength and pulse frequency) that are otherwise constrained by commercial laser sources. By decoupling the effective pixel rate from the laser’s native pulse repetition rate, the method enables pixel rates (up to 165 MHz) that are an order of magnitude higher than the laser frequency (16 MHz), surpassing the limits of point-scanning microscopy. The approach is presented as a low-cost, robust solution that minimizes photodamage and heating while maintaining high image quality, thereby advancing the capability of multiphoton microscopy for live biological samples.

The Core Problem: The "Slow" Laser

The Solution: The "Pulse Splitter"

Why This Works Better

The Results: Filming Life in Super-Slow Motion

In Simple Terms

More like this