Effect of Scan Angle and Reconstruction Algorithm on Model Observer Performance in Tomosynthesis

摘要: The goal of this work is to develop task-based figures merit for the assessment tomosynthesis imaging. Towards aim, we have computed performance a prewhitening model observer SKE detection task in uniform background, in-focus plane signal reconstructed volume, as well contrast. Test images were computer generated and using both iterative ML-EM filtered backprojection. We found that noiseless images, contrast was substantially higher FBP reconstruction. For reconstruction, decreased with iteration number. At 10 iterations, similar an increase scan angle images. showed slight decrease angle. results study are limited by task, which not realistic because backgrounds do contain structure noise. This will be addressed future.