Fast Two-photon Microscopy by Neuroimaging with Oblong Random Acquisition (NORA)

Area of Science:

• Neuroscience
• Computational Imaging
• Optical Microscopy

Background:

• Neural imaging advances allow studying large neural populations for perception, behavior, and cognition.
• Raster-scanning multi-photon imaging faces speed, field of view, and resolution trade-offs, limiting deep brain imaging.
• Computational imaging integrates optics and algorithms to overcome these limitations.

Purpose of the Study:

• Introduce Neuroimaging with Oblong Random Acquisition (NORA) for raster-scanning two-photon imaging.
• Overcome the fundamental trade-off between imaging speed, field of view, and resolution.
• Enable faster and more comprehensive imaging of neural circuits.

Main Methods:

• NORA subsamples fast scanning lines, integrating fluorescence from neighboring lines in the slow-scan direction.
• Randomly selected lines are imaged per frame, diversifying information for video-level reconstruction.
• Full video sequences are recovered using nuclear-norm minimization (matrix completion).

Main Results:

• Simulations using the NAOMi suite demonstrated NORA's capabilities.
• NORA accurately recovered 400 μm X 400 μm fields of view at 20X subsampling rates.
• Reconstruction remained accurate under realistic noise and motion conditions.

Conclusions:

• NORA offers a promising approach for fast neural circuit imaging.
• Minimal modifications to existing microscopy systems are required for NORA implementation.
• NORA enhances the scope of neural imaging by addressing key trade-offs.