2016 Winter Project Week/Projects/ShapeAnalysis

• 

  Projects List

Key Investigators

• Miaomiao Zhang (MIT)
• Polina Golland (MIT)

Project Description

Statistical shape analysis develops methods for the geometric study of objects. A common way to represent shapes for a group of images is the geometric transformation between each individual and the mean image. Analyzing diffeomorphic shape changes can be linked to disease processes and changes in cognitive and behavioral measures. One challenge of shape variability quantification is 'the curse of dimensionality', for instance, the transformation grid 128x128x128 as a shape descriptor for a 3D brain image. This makes the inference procedure computationally complicate and time-consuming. An efficient method needs to be developed to handle this complex dataset.

Objective	Approach and Plan	Progress and Next Steps
Develop a low-dimensional statistical shape analysis method on the manifold of diffeomorphic transformations.	Talk to Miaomiao	Derived mathematical framework Try to test the method and solve real problems

Background and References

Bayesian Principal Geodesic Analysis for Estimating Intrinsic Diffeomorphic Image Variability, Miaomiao Zhang and P. T. Fletcher, MICCAI, 2014.