“Quantitative Musculoskeletal MR imaging: from basic research to clinical translation”

Dr. Riccardo Lattanzi, a Professor from NYU, visited UCSF and showed us an inspiring variety of cutting edge quantitative MRI for MSK applications.

The quantitative MRI metrics his team has used includes T1, T2, T1rho, and magnetization transfer (MT) methods, which now have a strong set of studies showing their value in assessing joints, particularly in the composition and degeneration of connective tissues.

One exciting improvement for qMRI in the knee has been the development and applicaiton of MR fingerprinting (MRF) methods, which have been primarily been applied to the brain. Prof. Lattanzi showed compelling results using these techniques which are much faster than conventional qMRI methods, with unique innovations in quantifying MT.

They have used these techniques extensively in the hip, and also integrated them with radiomics. Furthermore, he showed preliminary results attempting to predict histology-level characterizations of cartilage based on qMRI and machine learning with a unique dataset from collaborators in Finland.

Another approach that caught my eye was dynamic wrist imaging with flexible coils and fast imaging to measure joint kinematics. Dynamic MRI has greatly improved in recent years (they used the GRASP-style techniques) as has the introduction of highly flexible RF coils. This could enable new types of kinematics research and protocols.

I also was intrigued by Prof. Lattanzi’s description of an emerging platform from DICOM “orchestration” - Mercure. This is a flexible, open-source platform aimed to easily integrate custom algorithms (e.g. AI models). This type of integration is key to many of our translational imaging research, so I think this is a good tool to keep an eye on. It is free, open-source, and seems to have some good momentum. https://mercure-imaging.org

Dr. Lattanzi’s publications: https://scholar.google.com/citations?user=TA_roHEAAAAJ