Machine Learning-Driven Dynamic Four-Dimensional X-Ray Computed Tomography Reconstruction

Hyojin Kim | 20-FS-010

Executive Summary

We are developing a machine learning technique to demonstrate the feasibility of using dynamic x-ray computed tomography to reconstruct time-sequential images representing motion and deformation during the non-destructive evaluation of objects. If successful, this research will be broadly applicable across many national missions, including nuclear weapons inspection and the study of material deformation and dynamics in geoscience and additive manufacturing.