Application of neural networks techniques for 2-D emission tomography
I. Eliash, Eitan Wacholder, Ezra Elias · 2002
This paper describes a computational neural-network architecture for the simultaneous solution of the location and intensity of a single radiation source distributed uniformly over a small section of a two-dimensional box. The possibility of locating two distinct sources in a box using noisy external signals is also discussed. A brief description of the neural-network method is followed by a detailed discussion of the simulation model and its application.>