Quantitative image analysis of optical sensor surfaces for movement monitoring

The use in clinical situations of a near real-time prototype opto-electronic dynamic 3D surface sensor has been investigated. Variations within optical sensor system height-maps show that noise amounts to 100 microns on snapshots of smooth body surfaces. It is not increased by image processing software removing triangulation spot artefacts. Presence of hairs on a body surface degrades the image quality, but still within tolerance (800 microns, typically the size of in-slice CT pixel). The accuracy of the system is 150 microns, the repeatibility 300 microns; `off-surface´ points should however be masked. Feasibility of dynamic 3D imaging of a body surface is demonstrated. Visualisation and analysis of the entire chest surface movement during breathing is achieved using the optical sensor system. Such monitoring and quantification of patient movement is directly relevant to gated radiotherapy and post-treatment verifications. Analysis of surface movement (height variations over time) in terms of mean and standard deviation should also be included in patient set-up procedures before radiotherapy treatment