Active CT Reconstruction with a Learned Sampling Policy

What is it about?

CT imaging system is an import step in clinical diagnosis, providing important evidence for clinicians. We have reformulated the CT imaging system, in which the raw signal sampling are also optimized to obtain a patient-specific sampling manner. The latter reconstructed CT image would benefit from such optimized sampling and the result provide more anatomy details for the latter diagnosis.

Featured Image

Photo by National Cancer Institute on Unsplash

Photo by National Cancer Institute on Unsplash

Why is it important?

We propose two deep-learning-based sampling policies in the reformulated CT imaging system. Both of the two policies exhibit better imaging performance than the widely used uniform sampling. With additional clinical Region-of-Interest (such as the anatomy to be diagnosed), our policies would interact with such information and provides better body information.