Advertisement for orthosearch.org.uk
Orthopaedic Proceedings Logo

Receive monthly Table of Contents alerts from Orthopaedic Proceedings

Comprehensive article alerts can be set up and managed through your account settings

View my account settings

Visit Orthopaedic Proceedings at:

Loading...

Loading...

Full Access

3D VISUALISATION OF MUSCLES FROM MR IMAGES



Abstract

Introduction: An estimated 80% of all adults will experience back pain at sometime during their life. To aid in the understanding of how the spine functions as a mechanical system and assist clinicians in their diagnosis this study produced 3D models of the muscles in the lumbar spine region. The models show selected muscles at rest and during controlled activities.

Methods: The images were acquired on a Siemens Sonata 1.5T System using breathhold FISP sequences. Twenty slices of thickness 5mm and zero separation were acquired using an in-plane resolution of .68mm and Fast-Fourier-Transformed to 512 x 512. Single acquisitions were acquired per slice. Imaging time per posture (rest, extension, left rotation and right rotation) was approximately 17–20 seconds. All image series conformed to the DICOM Standard.

The code developed for this study was written in Interactive Data Language (IDL) Version 5.5 from Research Systems Inc (RSI).

Each slice from an image series was displayed to an Operator, who roughly selected the muscle(s) boundary. The user-selected points were then compared with the 24-neighbouring pixels, and the vertices moved to the minimum value in the 5x5 area, which corresponds to the muscle boundary. The adjusted region of interest was then displayed to the user for verification. Once the Operator had completed selection of the regions of interest in all slices, spatial smoothing was performed on the data, and 3D models of the muscles constructed.

Results: This analysis produces 3D images of the muscles in the lower back. The visualisation of the data enables different combinations of muscle and posture to be displayed. Typically, a muscle at rest is overlaid with one of the three controlled activities – extension, left or right extension. The 3D models can be displayed as either a meshed or solid object.

The 3D model is displayed in a window that enables an operator using a mouse to rotate, scale and/or translate the model.

To aid visualisation, the volume of each muscle of interest is calculated using the number of pixels within the region of interest, pixel spacing and slice thickness. The result, in mm3, is displayed alongside the 3D model.

Discussion: The refinement of MR Imaging techniques for subjects in a variety of postures, and the development of post processing techniques provides a useful tool for all in the understanding of the mechanics of the lumbar spine. It is envisaged that this tool with further analysis will assist in determining if there is a link between muscle volume during movement and lower back pain.

The abstracts were prepared by Dr Robert Moore. Correspondence should be addressed to him at Spine Society of Australia, c/o the Adelaide Centre for Spinal Research, Institute of Medical and Veterinary Science, PO Box 14, Rundle Mall, Adelaide SA 500, Australia.