Neural networks based neutron emissivity tomography at JET with real-time capabilities

Tomographic reconstruction techniques typically require computationally intensive algorithms which are not suitable for real-time application. This paper describes a framework to perform neutron emissivity tomography at the Joint European Torus (JET) using neural networks with successful results over a broad range of magnetic configurations, heating and fueling schemes. Application times in the μ s time scale allows for real-time applicability of the method.