Project Summary Large-field-of-view high-throughput two-photon endoscope to image neuronal activity Development of miniaturized optical endoscopes have enabled visualization and recording of neural activity in freely-behaving animals. Two-photon endoscopes have excellent signal-to-background ratio, and can image deep into the tissue. It has a good optical sectioning and low phototoxicity. However, two-photon endoscopes have a very limited field of view and imaging throughput, and cannot typically perform axial refocusing for 3D imaging. These technological bottlenecks have so far prevented two-photon endoscopes’ broad deployment in neuroscience studies, particularly in the investigation of large scale neuronal circuits in freely-behaving animals. Here, we propose a two-photon endoscope that can image neural activity over a large field of view (~1.32x1.32 mm2) with high spatiotemporal resolution (>10 Hz, cellular resolution). This represents >35x increase of field of view, and >25x increase of image throughput than the state-of-the-art two-photon endoscopes. Furthermore, the proposed endoscope could perform fast refocusing, enabling multiplane 3D imaging. Our optical and mechanical design also ensures the compactness and lightweight of the endoscope. Such a two-photon endoscope will revolutionize the design principle of multiphoton endoscopes, and could play an important role in the investigation of large scale neuronal circuits in freely-behaving animals. This proposal overcomes the current technological bottlenecks in two-photon endoscopes by multiple innovations: (1) To achieve a large field of view with fine spatial resolution, the dimensions of the optics are typically large. We will custom design and optimize all the optical components, and make the whole package to be <~1.6x1x1 cm3. (2) To maintain a high spatial resolution, the imaging speed could be very low in conventional laser scanning microscopy as a diffraction limited spot is used as the sampling pattern. Here, we will increase the sampling spot size to a large square, and use computation algorithms to achieve super resolution (to cellular resolution). We meanwhile use temporal focusing technique to achieve a tight axial confinement. (3) We will employ a tunable liquid lens for fast axial refocusing. (4) While we do not foresee a large weight of the endoscope, we will use an active feedback mechanical design to greatly reduce its weight, and thus reduce the load burden on the animal. The proposed endoscope will greatly benefit the neuroscience community, be deployed to many research labs, and enable new research that were previously not possible. While we will design and test it for mice, its application could also be extended to rats and non-human primates in the future.