{"id":"arxiv/2602.13048","url":"https://arxiv.org/abs/2602.13048","tags":["math.NA"],"score":0,"title":"Data-Driven Filter Design for Flexible and Noise-Robust Tomographic Imaging","domain":"math.NA","source":"arxiv","authors":["Hamid Fathi","Alexander Skorikov","Tristan van Leeuwen"],"reviews":0,"subject":"math.NA","summary":"While filtered back projection (FBP) is still the method of choice for fast tomographic reconstruction, its performance degrades noticeably in the presence of noise, incomplete sampling, or non-standard scan geometries. We propose a data-driven approach for learning FBP filters and projection weights directly from training data, with the goal of improving robustness without sacrificing computational efficiency. The resulting reconstructions adapt naturally to the noise level and acquisition geometry, while retaining the speed and simplicity of classical back-projection. The proposed method…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-02-13","reactions":{}}