Session B05: Advanced Physics for Medical Imaging and Therapy

Accelerator Technology for Medical Applications

Accelerators for medical applications range from electron linacs in the meter scale to 70 meter circumference synchrotrons, from compact fixed field ion rings to laser-driven electron acceleration systems, from energies in a few MeVs to several 100s of MeV capability. Technologies are linked to the nature of the particle required which includes photons, electrons, protons, light ions, and some additional more exotic particles. Technologies used are linked as well to the irradiation method: passive, bunch-to-pixel, 3-D conformal, based on Bragg peak or not, on variable energy or not, ultra-short irradiation or slow-scan, high or low beam delivery repetition rate, radiography capabilities, etc.  Appropriate technologies also depend on required irradiation dose, with some categories of accelerators for instance mostly barred from high dose rates and/or ultra-short irradiation. 

This presentation will review the types of particle accelerators found in the field of medical application, it will also be an opportunity for a brief review of evolution in the field in the recent past, a present status, and prospects.    

Principles of physics and advanced computational analyses and the future of cancer care

Physics-driven methods to characterize malignant tumors due to cancer as well as their response to therapy have traditionally taken a back-seat to selecting treatments based on biological sub-type cellular analyses. Newer opportunities to use physics principles combined with machine learning and other computational approaches, however, imply an exciting future for the application of physical science principles together with advanced mathematical methods in oncology. Two key applications include applying energy budget principles as a basis for simulation models of tumor response and applying least-action and continuity principles to fluid flow to better characterize tumor vasculature and blood oxygen characteristics. At the same time, new artificial intelligence / machine learning approaches enable, for the first time, useful quantifications of tumor changes in response to therapy. Another opportunity is that AI/ML methods can be used to provide a quantitative understanding of how multiple physical imaging modalities relate to each other, and furthermore how this can be used to define initial conditions and corrections for the prediction of treatment response. The near future of application of physical thinking in cancer research and therapy, when put together with advanced computational methods, is indeed bright.   

Hyperpolarized Noble Gas Imaging at NewYork-Presbyterian/Columbia University Irving Medical Center

Hyperpolarized noble gas magnetic resonance imaging (MRI) was established by the pioneering work of Albert et. al. in the mid 1990’s, shortly after the development of high-powered diode lasers and high-density polarized helium-3 gas targets for electron scattering experiments at various DOE laboratories.  Over the past three decades, research groups in North America and Europe have developed novel biomarkers for an array of respiratory diseases.  However, despite substantial progress, there is still only at most a handful of large clinical trials using hyperpolarized MRI, and presently no hospitals employ the technique for clinical diagnosis.  We report on lung imaging research being performed at the NewYork Presbyterian/Columbia University Irvine Medical Center (NYP-CUIMC).  This research originated with a ~60 subject hyperpolarized helium-3 study of chronic obstructive pulmonary disease and is presently gearing up for a 100-subject study of long Covid using hyperpolarized xenon-129 MRI.  Measurements for the Covid-19 study will include regional ventilation, alterations in acinar microstructure and xenon gas transport from alveoli to blood.