Attenuation Tomography Use index matching liquid and cylidrical corrections (Trifonov paper) Increase angular range: Hexagonal setup with 3 walls of receiver and 3 walls of emitters 1D array but instead of k x 1 do (k/3) x 3 and place each LED slightly offset How to get around ART, by computing 3D coordinates of each ray, and simply assigniging attenuation to each ray. Then doing ART Refraction Tomography Direct method: assuming single unknown deflection angle per voxel (no attenuation), capture 4D LF and see if this can be inverted. Multiplexing Rays (dual of Atcheson) : Each source is a projector, projecting identifyable pattern, e.g. each with different frequency. Then use windowed fourier transform to find local frequency. Assuming monotonicity, find correspondence and then deflection in each view.