Full-Field Brillouin Microscopy with a Scanning Fabry-Perot Interferometer

This paper demonstrates that a standard multi-pass tandem Fabry-Perot interferometer, when operated in a spectral filtering mode and combined with light-sheet illumination, enables rapid, full-field Brillouin microscopy with millisecond-scale acquisition times, overcoming the historical speed limitations of FPI-based systems for practical imaging applications.

The Gist

Imagine you have a magical pair of glasses that can see how "squishy" or "stiff" things are, not by touching them, but just by looking at them. This is the promise of Brillouin microscopy. It's a super-powerful tool for scientists studying living cells, tissues, and materials, allowing them to map out mechanical properties without damaging the sample or using dyes.

However, there's a catch: until now, this "magic" was incredibly slow. It was like trying to paint a masterpiece by dipping a single brush into paint, making one tiny dot, waiting for it to dry, moving the brush an inch, and repeating. To get a full picture, it could take hours.

This paper introduces a clever new way to speed things up, turning that slow, single-dot painting into a fast, full-field snapshot. Here is how they did it, explained simply:

1. The Problem: The "Slow Scanner"

Traditionally, these microscopes used a device called a Fabry–Perot Interferometer (FPI). Think of this device as a super-precise musical filter. It only lets through light that matches a very specific "note" (frequency).

• The Old Way: To see the whole picture, the machine had to tune itself to one note, take a photo of one tiny dot, tune to the next note, take another photo, and so on. It was a slow, point-by-point process.
• The Misconception: Scientists thought this device was too slow to ever take a full picture of a whole area at once. They assumed it was only good for single points.

2. The Solution: The "Spectral Camera"

The author, Mikolaj Pochylski, realized they could repurpose this "slow" device to work like a camera.

• The Analogy: Imagine a room full of people singing different notes. The old way was to ask one person to sing, listen, then ask the next person. The new way is to have a special window (the FPI) that only lets through the sound of one specific note. If you stand in front of that window, you can see everyone singing that specific note at the same time.
• The Trick: Instead of scanning the whole range of notes, they only scan the tiny range where the "Brillouin note" (the mechanical signal) exists. By doing this, they can capture a whole 2D image of that specific note in a fraction of a second.

3. The Setup: The "Light Sheet"

To make this work, they combined the FPI with a Light Sheet.

• The Metaphor: Imagine shining a flashlight beam sideways through a glass of water to see dust particles floating in a thin slice. That's a light sheet.
• Why it helps: Instead of blasting the whole sample with light (which can hurt living cells), they slice the sample with a thin sheet of light. This illuminates the whole area evenly and gently. The camera then captures the light bouncing off that entire slice at once.

4. The Hurdle: The "Distorted Lens"

When they tried to take a picture of a whole field at once, they hit a snag. Because the FPI was designed for a single point, looking at a whole area caused "optical distortion."

• The Analogy: It's like looking at a flat map through a funhouse mirror. The center looks right, but the edges are stretched or tilted. In their data, the "stiffness" measurements looked different depending on where you were in the image, even if the sample was perfectly uniform.
• The Fix: They created a mathematical "corrective lens." By using water (which has a known, perfect stiffness) as a reference, they mapped out exactly how the mirror distorted the image. Then, they applied this map to all future images to "un-distort" the data, making the measurements accurate everywhere.

5. The Results: Fast, Clear, and Gentle

With this new system, they achieved some amazing things:

• Speed: They went from taking hours to get a picture to doing it in under a minute.
• Resolution: They can see details as small as a few micrometers (about the width of a human hair's thickness).
• Safety: Because they use a light sheet and don't need to scan point-by-point, the sample gets much less light exposure, keeping delicate living cells happy.
• Versatility: They tested it on everything from a cat's hair (showing the different layers inside) to plant cells and onion skin. They could even see the difference between water and a plastic bead just by their "stiffness."

The Big Picture

This paper is like finding a way to use an old, slow typewriter to print a whole newspaper page at once. By changing how the machine is used (scanning the frequency instead of the space) and adding a little bit of software magic (the correction algorithm), they turned a slow, single-point tool into a fast, full-field imaging system.

This means scientists can now quickly map the mechanical health of tissues, potentially helping to detect diseases like cancer (which often changes tissue stiffness) much faster and more easily than before, all without hurting the sample.

Technical Summary

1. Problem Statement

Brillouin microscopy is a powerful, label-free technique for mapping the mechanical properties (viscoelasticity) of biological tissues at submicron resolution. However, its widespread adoption is hindered by slow acquisition speeds.

• Fundamental Limit: The spontaneous Brillouin scattering cross-section is extremely low, requiring long integration times.
• Methodological Limit: Most high-contrast systems rely on scanning Fabry–Perot Interferometers (FPIs). Traditionally, these are used in point-scanning mode, which is too slow for practical imaging.
• Alternative Limitations: While other full-field approaches exist (e.g., VIPA-based spectrometers or Fourier-transform interferometers), they often suffer from low spectral contrast, requiring additional notch filters to suppress the intense elastic background, or lack the ability to perform frequency-selective imaging.
• The Gap: There is a need for a system that combines the exceptional spectral resolution and contrast of multi-pass FPIs with full-field (2D) spatial multiplexing to achieve high-speed, high-contrast mechanical imaging.

2. Methodology

The authors repurposed a standard commercial Sandercock-type multi-pass tandem Fabry–Perot Interferometer (TFPI) for full-field imaging by operating it in a spectral filtering mode rather than a scanning spectroscopy mode.

• Optical Setup:

  • Illumination: A light-sheet geometry is used (532 nm CW laser) to provide uniform excitation, optical sectioning, and minimal photodose.
  • Detection: Orthogonally scattered light is collected and directed into the TFPI.
  • Interferometer: The TFPI operates in a six-pass configuration with two slightly detuned cavities. It acts as a narrowband spectral filter. Instead of scanning the entire spectrum for a single point, the system captures a 2D image at a specific frequency shift, then steps the mirror separation to capture the next frequency.
  • Acquisition: A stack of ~30–40 spectral images (covering Stokes and Anti-Stokes shifts) is acquired sequentially. Each image corresponds to a discrete frequency.

• Data Processing & Correction:

  • Angular Artifacts: Because the TFPI was designed for on-axis single-point spectroscopy, off-axis rays in full-field imaging cause spectral distortions (shifts, broadening, and vignetting) due to the finite numerical aperture (NA) of the objective and internal apertures of the FPI.
  • Calibration: The authors developed a calibration protocol using water (a standard with known mechanical properties). They modeled the scattering angle distribution $p(\theta)$ as a Gaussian function for each pixel.
  • Deconvolution: By fitting the measured spectra to an angularly averaged Damped Harmonic Oscillator (DHO) model, they extracted pixel-specific angular parameters ($\theta_m$ and $\sigma_\theta$). These parameters were then used to correct the mechanical maps of unknown samples, removing field-dependent distortions.

3. Key Contributions

1. Repurposing FPI for Full-Field Imaging: Demonstrated that a standard multi-pass tandem FPI, previously considered too slow for imaging, can achieve high-speed full-field Brillouin microscopy when operated as a tunable spectral filter.
2. Frequency-Selective Emission Imaging: Uniquely, this system allows for the acquisition of Brillouin emission images at selected frequency shifts. This enables the visualization of specific mechanical components within a heterogeneous sample without needing to reconstruct the full spectrum first.
3. Correction Algorithm: Developed a robust angular deconvolution method to correct for field-dependent spectral distortions inherent to using a high-NA objective with a collimated FPI, enabling accurate mechanical mapping across the field of view.
4. High-Speed, Low-Dose Imaging: Achieved acquisition times of 30–80 seconds for a full spectral stack (millisecond-scale pixel dwell times) using low-power light-sheet illumination, orders of magnitude faster than point-scanning FPIs.

4. Results

• Performance Metrics:

  • Speed: Full 2D images acquired in ~1 minute; pixel dwell times in the millisecond range.
  • Spatial Resolution: Estimated at ~3 µm (based on intensity transitions at material interfaces) and ~7 µm (based on DHO amplitude maps), limited by light-sheet thickness and FPI optics.
  • Spectral Contrast: The TFPI provided >150 dB suppression of the elastic background, eliminating the need for external notch filters even in highly scattering samples.
  • Precision: Spectral precision of 15–50 MHz.

• Demonstrations on Diverse Samples:

  • Water: Validated the correction algorithm, showing uniform mechanical maps after deconvolution.
  • Cat Hair (Felis catus): Resolved distinct mechanical layers (water, cortex, cuticle) corresponding to ~5, 7, and 12 GHz peaks. Frequency-selective imaging clearly delineated these structures.
  • Dextran Microspheres: Detected subtle mechanical differences between microspheres and water, demonstrating sensitivity to low-contrast variations.
  • Plant Cells (BY-2): Label-free imaging of subcellular structures (cytoplasm, vacuole, nucleus) based on mechanical contrast.
  • Onion Tissue: Successfully imaged optically dense, scattering tissue, proving robustness in challenging biological environments.

5. Significance

This work bridges a critical gap in Brillouin microscopy by proving that high-contrast, high-resolution FPIs can be used for rapid, full-field imaging.

• Accessibility: Since multi-pass FPIs are already common in many physics and materials science laboratories, this approach offers a practical pathway to upgrade existing hardware for biological imaging without building new, expensive spectrometers.
• Unique Capabilities: The ability to perform frequency-selective imaging (visualizing specific mechanical components instantly) is a capability unique to this FPI-based approach among spontaneous Brillouin techniques.
• Future Outlook: While current limitations include a restricted field of view (due to FPI apertures) and sensitivity of viscosity measurements to alignment drifts, the study outlines clear paths for optimization (e.g., telecentric designs, real-time monitoring). This paves the way for dedicated, high-throughput FPI instruments for mechanobiology.