Mapping the progression of non-carious cervical lesions

Non-carious cervical lesion' (NCCL) is a term used to describe loss of hard tissue in the cervical one-third of teeth. NCCLs are commonly encountered in dental practice and frequently require clinical intervention. There are few reports of high quality mapping and monitoring of NCCLs, and dental practitioners currently have no reliable method for measuring and monitoring their progress. Improved understanding of the underlying causes and progression of NCCLs will rely on reliable and practical methods of mapping their topography and monitoring their progress. NCCLs on the facial surfaces of teeth can be easily imaged, either directly or on cast replicas, and therefore may be well suited to mapping using a photogrammetric approach. A preliminary investigation led to the development of a casting material that incorporated optical texture so that photogrammetric image-matching techniques could be utilized. Three-dimensional coordinate data for tooth replicas was successfully generated using the casting material, convergent stereoscopic photography and commercial digital photogrammetric software. Imaging was performed initially using a semi-metric 35 mm film camera, then a high-resolution digital SLR camera, and finally a fixed-base digital SLR stereo camera. The quality of the surface data and the capacity to align tooth surfaces was investigated. Two specific examples of NCCLs were mapped and monitored at baseline, 12 month and 24 month time periods. For the film camera and the single digital camera, the photogrammetric solutions were not highly stable, with systematic height errors of up to 80 ˜í¬¿m attributed to unstable exterior orientation. However, for the fixed-base stereo camera, model precision was shown to be in the order of 13 ˜í¬¿m and the accuracy of surfaces derived from automatic measurement was approximately 3 ˜í¬¿m. The error associated with aligning independent measurements of tooth surfaces was approximately 17 ˜í¬¿m. Change detection of the two NCCL surfaces was sensitive to 30 ˜í¬¿m, with change ranging from 30 to 320 ˜í¬¿m per annum for one surface and 30 to 70 ˜í¬¿m per annum for the second surface. Different rates of change were clearly evident in different areas of the same surfaces. The replication, stereoimagery, photogrammetric processing, and detection of changes to the surfaces were shown to be reliable and convenient. The results of this investigation show that stereo-photogrammetric techniques can be applied to the mapping of NCCLs, and that the surfaces can be mapped at sufficient accuracy to enable change to be monitored. The two examples suggest that annual change detection studies will provide a clearer picture of the rate of progression and the geometry of progression and, in combination with other analytical techniques, a more detailed explanation of the natural history of non carious cervical lesions.