Making advanced characterisation of tissue microstructure clinically practical: a data-driven approach to efficient microstructural MRI
MRI provides information on the structure of the microscopic building blocks of living tissue. Combining different MRI techniques is crucial to achieve a comprehensive quantitative picture, but as the dimensionality of the MRI acquisition space increases, acquisition times and analysis-complexity become prohibitive.
I will design methods to collect the most relevant MRI measurements in the shortest possible time to make quantitative microstructural MRI clinically-viable. I will employ a ‘top-down and bottom-up’ strategy. The top-down approach considers that it is known which tissue parameters are relevant and how they relate to measured signals; the acquisition can then be optimised to maximise precision per unit acquisition time. The bottom-up approach considers that it is unknown how many and which features are relevant; this may differ from assumptions in mathematical models. Starting with rich multicontrast data, I will develop data-driven approaches to characterise the measurement of information-content and use this to select the most relevant measurements.