Project Summary Photoacoustic imaging is emerging as a valuable research tool with several preclinical commercial systems now available. The technology relies on a pulsed laser which induces absorbers to emit broadband acoustic waves. The waves are typically detected with piezoelectric transducers and images are formed from the time-of- flight of the acquired signals. The high cost of current analog-to-digital conversion technology, however, restricts the number of individual data channels which can be simultaneously acquired. The end result is a tradeoff between cost, the field-of-view, and the temporal resolution. Fabry-Pérot etalons provide an alternative to piezoelectric transducers. They rely on the optical interference of a beam resonating between two parallel optical reflectors to produce sensitive interferometric detection of ultrasound waves. The sensors are inherently broadband, which leads to improved photoacoustic image quality. Current implementations of Fabry-Pérot etalons raster scan a single sensing beam throughout the two- dimensional etalon to acquire a single measurement at a time. Thus, the photoacoustic signal-generating laser must be pulsed hundreds to thousands of times to acquire a single image. In this two-year technology development proposal, we seek to combine the benefits of Fabry-Pérot etalons with new concepts in compressed imaging to acquire an entire volumetric photoacoustic image with a single- shot of the laser. This could dramatically improve the temporal resolution of the imaging technique while simultaneously decrease the cost of the system. Rather than detect the interference from a single point on the Fabry-Pérot etalon at a time, we will instead acquire an image of the entire sensor with a camera. Then, the light from the sensor will be swept quickly across the camera to sample in time. In the absence of additional optics, this would lead to a blurring of the camera image and the photoacoustic image would not be recoverable. Therefore, we will use either a digital micro-mirror device reate a mask which restricts the light from reaching the camera. This will provide structure to the acquired data to enable reconstruction of the photoacoustic image. In conjunction with building the system for optical acquisition of photoacoustic images, we will simultaneously develop computational imaging techniques to reconstruct the images. We will first build a forward model which maps a given photoacoustic image to the acquired camera image. This will be done via simulations and measured experimentally. The model will be underdetermined (i.e., there are more voxels in the 3-D photoacoustic image than there are pixels in the acquired data). Thus, there must be some additional support applied to the reconstruction. We will incorporate a sparsity constraint on the final image (i.e., either the photoacoustic image is sparse or its edges are) in order to converge to the correct result.