Hua P, Webster J G, Tompkins W J
Department of Electrical and Computer Engineering, University of Wisconsin 53706.
Clin Phys Physiol Meas. 1988;9 Suppl A:137-41. doi: 10.1088/0143-0815/9/4a/023.
The Newton-Raphson algorithm is an effective reconstruction method. However, it exhibits degraded performance for ill-conditioned problems, especially in the presence of measurement error. This paper uses the regularisation method to improve the system's conditioning, which stabilises the reconstruction method. It shows that the [[ p'' ]] penalty form yields superior images, as compared to the [[ p ]] penalty form.
