With deep learning being a paradigm shifter, a lot of innovations have also been made in medicine, ranging from medical image understanding to health record analysis. In this talk, recent advancement in medical image reconstruction and analysis carried by our group at RPI will be presented. The involved tasks include low-dose CT image denoising, compressive macroscopic fluorescence lifetime image reconstruction, medical imaging registration, and image-based cancer detection. Motivated by different clinical applications, we have developed several deep learning techniques for reconstructing medical images with higher quality and much faster speed and learning new similarity metrics for improved multi-modality image registration. The proposed methods and achieved experimental results will be given in the talk.

Speaker(s): Pingkun Yan,

Location:
Room: 114
Bldg: MEC
Boise State University
Boise, Idaho
83725