Development and Evaluation of Novel MRI methods
Brief Description of Project:
The overall objectives are to develop and use MRI to report specific characteristics of tissue, such as axon diameter or myelin thickness in white matter, myofiber size and density in skeletal muscle, or pore and bound water concentrations in bone. To this end, we develop and use a variety of experimental NMR/MRI protocols, signal analysis methods, and computational modeling. For the summer of 2020, we aim to develop and evaluate machine learning methods with the long term aim of predicting bone fracture risk from low-cost MR measurements.
knowledge of MRI physics, computer programming skills (MATLAB, C, CUDA, Python, any of the above).
Nature of Supervision:
The student will work under the supervision of Mark Does, and with the assistance of other staff and trainees in the lab.
A Brief Research Plan (period is for 10 weeks):
The plan will be to learn the signal equations that describe 1H NMR signal from bone, use these to perform preliminary ML training to extract specific signal characteristics from fast, low-cost measurements, and then evaluate the ML performance using cadaver bone specimens.
Number of Slots: 1
Mark D. Does, Ph.D.
Professor of Biomedical Engineering, Radiology & Radiological Sciences
Vanderbilt University Institute of Imaging Science
1161 21 Ave South
Nashville, TN 37232-2675