Optical imaging of neuronal activity in the mammalian brain at depth and at high spatial and temporal resolution remains a key challenge in neuroscience. This is because tissue scattering eliminates directional information carried by photons, with a characteristic length scale of hundreds of microns. As a result, the remaining unscattered, or ?ballistic? component of light decays exponentially with depth. Since all optical microscopy modalities rely?either in the excitation or the emission?on this ballistic component of light for image formation, the limitations imposed by scattering limit the maximally obtainable imaging depths to ~1.5 mm in the rodent brain. Hereby, we propose a systematic approach for exploring and developing a new platform for in vivo deep tissue calcium imaging that actively exploits the scattering properties of the tissue as a resource to localize, demix and extract neuronal activity traces in the millimeter depth range in the highly scattering rodent brain. Our approach utilizes the fact that the scattering process results in an accumulation of different relative phases between scattered light components, leading to the formation of deterministic but complex interference patterns known as speckles. Emission from different locations of the sample results in speckles with different intensity distributions. Thus, these patterns carry complex information about both the location of the emitter and the properties of the scattering medium. Our current results show that the obtainable contrast for these speckle patterns should be sufficient to identify, localize and demix them even for extended objects and with partially coherent light. We will design and build an array of hardware and computational tools that will allow us to systematically explore the limits within which such demixing and neuronal signal extraction can be achieved in vivo. We will use these insights and develop an optical platform for volumetric calcium imaging in the mouse brain in the deep forward-scattering regime, as well as for transcranial calcium imaging in the mouse brain. Our approach will synergistically combine the expertise on the physics of light propagation in disordered media in the Gigan Lab with the experience in machine learning, computational imaging and in vivo high-speed volumetric neuronal recording in the Vaziri Lab. Our method has the potential of opening up a new paradigm for neuronal imagining and signal extraction in the multiple scattering regime by enabling millimeter-range calcium imaging in the highly scattering rodent brain.
Optical recording of neuronal activity in the mammalian brain at depth with high spatial resolution remains, due to the highly scattering properties of brain tissue, a key challenge in neuroscience. This is because all microscopy modalities rely on the non-scattered ballistic component of light?which is exponentially attenuated with tissue depth?either in the excitation or the emission for image formation. We will overcome this issue by developing a deep-tissue calcium imaging platform that exploits the scattering properties of the tissue to localize, demix and extract neuronal activity traces in the millimeter range in the highly scattering rodent brain.
