Title: New Methods in PET Image Reconstruction: Going Beyond the State-of-the-Art

Abstract:

In this talk, we will introduce some of the methods that our group has recently proposed to improve the quality of PET image reconstructions beyond the state-of-the-art. On one hand, the super-iterative method is able to obtain better image resolution by using an initially reconstructed image to improve the acquired data. The improved data can be then reconstructed with improved quality. We also proposed a generalized method of moments as an alternative to using frames for dynamic PET studies. By using moments instead of frames, we obtained results much faster and with lower bias. Finally, we have used deep-learning methods to perform positron range correction in PET images. The tool, trained with GPU-based Monte Carlo simulator MCGPU-PET, has been successfully applied to 68Ga acquisitions.

Short Bio:

Joaquin Lopez Herraiz ( https://www.ucm.es/gfn/jlherraiz ) is an associate professor at Faculty of Physics at Complutense University of Madrid (UCM). He obtained a PhD in Nuclear Physics in 2010 at UCM and was a postdoctoral researcher at MIT. His research topics are Monte-Carlo Simulations, Tomographic Image Reconstruction, High-Performance Computing, Multimodal Imaging, and Artificial Intelligence applied to Medical Imaging. He has more than 40 papers on medical imaging, participated in multiple Spanish, European and USA projects on PET imaging.

Location:

Chadwick Building | UCL Maps Room 102A
________________________________________________________________________________

Microsoft Teams meeting

Join on your computer or mobile app

Click here to join the meeting

Or call in (audio only)

+44 20 3794 0272,,224326417# United Kingdom, London

0800 047 8623,,224326417# United Kingdom (Toll-free)

Phone Conference ID: 224 326 417#

Find a local number | Reset PIN

Learn More | Meeting options