Hyperspectral fiber photometry for dissecting neural circuit dynamics
Project Description
(Conference abstract submission) - Conference Poster
The brain is made up of many cell types that communicate through cell firing and the release of a variety of neurotransmitters to control brain function and behavior. Genetically encoded fluorescent indicators are widely used to measure cellular activity and neurotransmitter release on specific locations and cell types to dissect neural circuits. However, due to technical challenges, prior work has largely been limited to studying at most two signals at once in a given brain region. To enable multiplexed fluorescent measurements across many brain areas in vivo, we developed a hyperspectral fiber photometry system. The system records spectrally resolved emission at five excitation wavelengths from optical fibers implanted into the brain. Excitation lasers are combined and shaped to illuminate a linear array of optical fibers that are coupled to different brain areas. Emitted light is dispersed perpendicular to the fiber array using a prism. The emitted spectra are captured on a sCMOS camera. Image pixels are calibrated to wavelengths and images are interleaved according to the excitation laser. These hyperspectral measurements are then unmixed using a custom constrained non-negative matrix factorization algorithm that accounts for bound and unbound indicator states, autofluorescence, hemodynamics, and brain movement. In preliminary experiments, we have performed simultaneous measurements of dopamine, acetylcholine, and calcium in the ventral striatum of mice performing a reward-based decision-making task. Preliminary analysis reveals distinct dynamics of each of these signals modulated by behavioral states and outcomes.