Diesel2p mesoscope with dual independent scan engines for flexible capture of dynamics in distributed neural circuitry

Che‐Hang Yu, Jeffrey N. Stirman, Yiyi Yu, Riichiro Hira, Spencer L. Smith

TL;DR

Researchers have developed "Diesel2p," an open-source two-photon mesoscope that overcomes previous limitations in imaging neural activity across large brain areas. By employing dual, independently controlled scan engines and advanced optical techniques, Diesel2p achieves subcellular resolution over an unprecedented ~25 mm² field of view, enabling simultaneous, flexible imaging of distributed neural circuits.

What Problem Does This Paper Address?

Imaging neural activity deep within the brain, especially across distributed networks spanning millimeters, has been a significant challenge. Traditional two-photon microscopes are limited to small fields of view (around 0.5 mm diameter) due to optical constraints. While some systems have attempted to increase throughput with multi-beam scanning, these often force all beams to use the same scanning parameters, limiting flexibility. This inflexibility prevents researchers from optimizing imaging for different neural circuits or experimental needs, hindering the study of complex, widespread brain dynamics. The paper addresses this by proposing a system that offers both a large field of view and independent control over multiple scanning regions.

Methods

The core innovation of Diesel2p lies in its Dual Independent Enhanced Scan Engines (Diesel2p). Key methodological components include:

  • Large Field-of-View (FOV) Optics: The system is designed to achieve a ~25 mm² FOV (approximately 5x5 mm square) while maintaining subcellular resolution. This is achieved through custom optical design, including specialized relay lenses, scan lenses, tube lenses, and objectives, optimized for a wide range of scan angles.
  • Dual Independent Scan Engines: Instead of a single scanning mechanism, Diesel2p utilizes two completely independent scan engines. Each engine comprises a resonant scanner for rapid line scanning and galvo scanners for transverse scanning. This independence allows for arbitrary positioning and distinct scanning parameters (e.g., frame rate, scan region size) for each of the two imaging beams.
  • Temporal Multiplexing: The two beams are temporally multiplexed, meaning their pulses are interleaved in time. This allows for simultaneous imaging from both pathways without the dead time associated with sequential scanning or beam switching. A delay line of 1.87 meters introduces a 6.25 ns delay between the two beams, enabling their independent detection.
  • Adaptive Optics (AO): Each pathway is equipped with a deformable mirror. This AO system is used for two primary purposes:
  • Aberration Correction: It corrects optical aberrations to maintain subcellular resolution across the large FOV.
  • Remote Focusing: It allows for rapid axial adjustment of the focal plane, enabling imaging at different depths without physically moving the objective.
  • Polarization Optics for Beam Recombination: The two beams are combined using polarization beam splitters. This method is crucial as it allows the input lasers to originate from the same or different sources (enabling multi-wavelength imaging) and allows each arm to use multiple sources simultaneously (e.g., for imaging and photoactivation).

Datasets

The paper does not describe training datasets. Instead, it focuses on the performance of the Diesel2p system using:

  • Subresolution Fluorescent Beads (0.2 µm): Used to measure the system's optical resolution (lateral and axial) across the FOV and at various depths.
  • Structured Fluorescent Sample: Used to verify the imaging field of view and flatness.
  • In Vivo Calcium Imaging Data:
    • GCaMP6s expressing neurons in transgenic mice: Used to demonstrate subcellular resolution in living brain tissue at depths up to 345 µm and to record neural activity across large cortical areas.
  • Thy1-GFP mouse: Used to visualize and resolve dendritic spines in vivo.

The paper introduces new benchmarks for large-FOV, independent multi-region imaging, demonstrating its capabilities in capturing distributed neural activity.

Key Insights and Ideas

  • True Independence is Key: The most significant insight is that truly independent scan engines, rather than yoked or sequentially scanned beams, unlock unprecedented flexibility. This allows researchers to tailor imaging parameters to specific neural populations or experimental questions, opening up new avenues of research.
  • Large FOV + Subcellular Resolution = New Possibilities: Combining a ~25 mm² FOV with subcellular resolution is a major leap. This means researchers can now simultaneously observe activity in multiple cortical areas or track the dynamics of widely distributed neuronal ensembles, which was previously impossible.
  • AO for Enhanced Performance: The integration of adaptive optics is critical for maintaining high resolution across such a large and deep imaging volume, compensating for optical imperfections and enabling remote focusing.
  • Open-Source Design: Making the system open-source is a crucial contribution, allowing other labs to replicate, adapt, and build upon this technology, accelerating neuroscience research.

Key Results

  • Field of View: Achieved a ~25 mm² (5x5 mm) field of view with subcellular resolution.
  • Depth: Capable of imaging up to 7 mm deep in scattering tissue.
  • Resolution: Maintained nearly constant subcellular resolution across the entire FOV, with lateral FWHM ~1 µm and axial FWHM ~8 µm.
  • Flexibility Demonstrated:
    • Simultaneously imaged two regions 4.36 mm apart with independent parameters.
  • Imaged overlapping regions.
  • Performed "quad imaging" of four distinct regions.
  • Combined raster scanning with random-access scanning for targeted cell imaging.
  • In Vivo Performance: Successfully imaged calcium signals from over 5,800 neurons across a 15 mm² area at 3.84 frames/s, demonstrating robust spike inference and correlation analysis over large distances.

Limitations and Future Work

  • AO Limitations: While effective, the current AO system's defocusing range is limited by the deformable mirror's stroke. The authors note that other AO systems could offer greater range.
  • Laser Power: The paper mentions imaging was performed with <100 mW out of the objective, and no damage was observed. However, for deeper imaging or faster acquisition, higher laser powers might be required, necessitating careful consideration of phototoxicity.
  • Future Directions: The authors suggest that Diesel2p is compatible with various multiphoton imaging extensions, such as Bessel-beam scanning and reverberation microscopy, and can be combined with optogenetics for simultaneous imaging and perturbation. The system's achromatic nature at 910 nm and 1040 nm also makes it ideal for dual-wavelength imaging of different molecules simultaneously. The open-source nature encourages further development and application to study neural interactions across brain areas.

Summary updated

Abstract

Imaging the activity of neurons that are widely distributed across brain regions deep in scattering tissue at high speed remains challenging. Here, we introduce an open-source system with Dual Independent Enhanced Scan Engines for Large field-of-view Two-Photon imaging (Diesel2p). Combining optical design, adaptive optics, and temporal multiplexing, the system offers subcellular resolution over a large field-of-view of ~25 mm<sup>2</sup>, encompassing distances up to 7 mm, with independent scan engines. We demonstrate the flexibility and various use cases of this system for calcium imaging of neurons in the living brain.
BibTeX
@inproceedings{yu2021diesel,
  author = {Che‐Hang Yu and Jeffrey N. Stirman and Yiyi Yu and Riichiro Hira and Spencer L. Smith},
  title = {{Diesel2p mesoscope with dual independent scan engines for flexible capture of dynamics in distributed neural circuitry}},
  booktitle = {Nature Communications},
  year = {2021},
  doi = {10.1038/s41467-021-26736-4},
}
Figure