||Department of Engineering
F-AHG-1: Registering surfaces for bone analysis
This project should appeal to 3G4 students, since it encompasses medical imaging (CT), shape analysis and graphical rendering. The motivation is to establish how structural properties of an individual's femur might predispose that person to hip fracture. This is a hugely important health and wealth issue: hip fracture amongst the elderly causes significant physical and emotional distress, and the cost to health services is considerable, with annual cases forecast to rise to 6.3 million worldwide by 2050. High risk individuals could receive targeted therapy (drugs, exercise), but current screening methods fail to spot the majority of individuals who go on to fracture.
Our recent pioneering research demonstrates the power of cortical thickness maps. These are coloured renderings of the bone surface that show the thickness of the stiff, outer shell (cortex) which is responsible for most of the bone's strength. However, in order to answer the important questions (what type of cortical thickness distribution might predispose an individual to fracture, does any particular therapy increase cortical thickness in the right regions?) it is necessary to compare cortical thickness maps across many individuals. And before we can compare one individual with another, we must first register (spatially align) their cortical thickness maps.
Registration is a key topic in medical imaging, and we have already implemented an algorithm that does a good job of aligning two femur surfaces using an affine transformation followed by a nonrigid B-spline free-form deformation. However, the alignment is based entirely on the surface's shape, and this raises a potential problem with the subsequent thickness analysis. Suppose we observe that individuals in Group A have different thickness distributions to those in Group B. Can we be sure that this is a genuine thickness difference? Maybe the two groups had different femur shapes, and these aligned in systematically different ways, so that what we are actually seeing are different alignments of the same thickness distribution, not different thickness distributions. To answer this question, this project will investigate a range of alternative registration algorithms, which attempt to align thickness as well as shape, and assess their performance.
The project will be supported by data and advice from the Bone
Research Group at Addenbrooke's Hospital.
Our recent research has shown how to extract cortical thickness maps
(top right, pink is thin, blue is thick) from CT data (top
left) by deconvolution (bottom). The result is a detailed, accurate
map of an individual bone. |
For cohort studies, we need to register individual femurs onto a
common morphology. We currently do this using a global affine
transformation (left, first nine frames) followed by a nonrigid
B-spline free-form deformation (left, last eight frames). While this
aligns the surfaces nicely, it does not necessarily align the
thickness maps. |