Automatic digitization and analysis of moiré topograms on a personal computer for clinical use

Moiré topography is a non-invasive optical technique which may be used for the detection and follow up of spinal deformities such as scoliosis. The moiré technique produces interference patterns (moiré fringes) which are very good at showing up differences in symmetry on the patient's back. The moiré fringes follow lines of constant height and can therefore be considered analogous to contour lines on an ordinance survey map. The moiré technique is already being used as a tool for qualitative diagnosis but widespread use will wait until the task of interpreting the moiré patterns quantitatively has been automated. An automated process means extracting the third dimension from the moiré topogram to produce 3D representations from which various deformity indices may be calculated. Image processing of moiré topograms is not trivial due to the difficulty of producing distinct good contrast fringes. The contrast problem is addressed with an interactive contrast enhancing algorithm. The approach to extracting 3D information from the fringes is to tackle 'real world' problems head on; fringes are first detected using edge detection techniques and assigning heights to the fringes is accomplished with algorithms designed to cope with noisy and broken fringes.