SIENA

Overview

Quantitative measurement of change in brain size and shape (for example, in order to estimate atrophy) is an important current area of research. New methods of change analysis attempt to improve robustness, accuracy and extent of automation. SIENA is a fully automated method, which is accurate (around 0.2% brain volume change error) and achieves high robustness (no failures in over a thousand analyses over a range of different data sets). The input is two head MR images taken at different points in time, and the output is a change image, along with an estimate of percentage brain volume change. SIENA automatically segments brain from non-brain in each image, and also estimates the external surface of the skull in each. Next, the two brain images are registered, whilst using the skull images to constrain scaling and skew; this corrects for changes in imaging geometry over time. Brain surface points (internal and external) are then found using the registered brain images to subvoxel accuracy, and the surface motion estimated on the basis of these points. The mean perpendicular edge motion across the entire brain surface can then be converted into a percentage brain volume change estimate.

SIENA has also been extended to a cross-sectional method (SIENAX) which uses segmentation to find brain volume and brain & skull-based registration (similar to that used by SIENA) to normalise to standard space, to reduce inter-subject variability.

Papers

• See a slideshow on SIENA/SIENAX.
• A paper on SIENA has been published in JCAT; also see a related technical report (PDF).
• A second paper, on SIENAX and improvements to SIENA has been published in NeuroImage; also see a related technical report (PDF).
• We have recently extended SIENA ("SIENAr") to allow voxelwise statistical analysis of atrophy across subjects; see ISMRM04 and HBM04.

Software

SIENA/SIENAX is available as part of the FMRIB Software Library.