Research Projects

We are developing several powerful MR technologies to enhance PET/MR for physiological imaging.

Our aims are to:

• Develop a free-running, three-dimensional, multi-parametric, cardiac MRI method for motion-corrected whole-heart membrane potential imaging using PET/MR
• Develop high-resolution MRSI for multi-modality, multi-molecular imaging using PET/MR
• Develop ultra-fast, multi-component MR relaxometry using deep learning and MR physics modeling

Selected results

We are developing an innovative physics-informed AI-based pipeline of data processing in PET/MR, including image reconstruction, transfer learning, and uncertainty estimation.

Our aims are to

• Develop a deep-learning-based image reconstruction that incorporates anatomic MR image and motion correction for both cross-sectional and longitudinal studies.
• Develop domain adaptation (DA) and DA few shot learning (DA-FSL) in PET/MR.
• Develop a DL-based approach to estimate the posterior distribution of PET image.

Selected results

We are developing radiotracers and PET/MR imaging methods for measuring cellular membrane potential in the myocardium and mitochondrial membrane potential outside the heart, demyelination in the brain, and immune function throughout the body.

Our aims are to

• Quantitatively map membrane potential
• Image potassium channels as a novel biomarker for demyelination
• Develop a nanoparticle-based platform for specific imaging of the immune response

Selected results