ABSTRACT
BACKGROUND: With the aid of scanning laser tomography, feasible with the Heidelberg Retina Tomograph (HRT), refined structures can be measured three-dimensionally. Pictures are built up from scanned layers which are represented as two-dimensional topographical or reflectivity pictures by the HRT software. The kind of information that is provided by the third dimension can be exploited much better by real spatial three-dimensional presentation. The autostereoscopic Dresden 3D display makes such a spatial presentation feasible as add-on to the HRT. METHODS: Seventeen patients (9 women, 8 men) were chosen on the basis of a long duration of observation (mean 42.41 months) and a large number of follow-up examinations (mean 5.29). These patients were examined with the HRT as well as with the Dresden 3D display. The results were compared regarding their correlation pairs. RESULTS: Comparing the correlation pairs (r>0.7) the Dresden 3D display provided a higher Pearson correlation coefficient with 8 out of 10 pairs. It was evident that the evaluations of the two devices referred to the same parallel structures. DISCUSSION: The Dresden 3D display facilitates, with its form of presentation, a better evaluation of the measurements of the HRT. More precise evaluation of retinal structures by virtue of the autostereoscopic display presents an improvement.
Subject(s)
Glaucoma/pathology , Lasers , Retina/pathology , Tomography/methods , Feasibility Studies , Female , Humans , Male , Observer Variation , Optic Disk/pathology , Reproducibility of Results , Severity of Illness Index , SoftwareABSTRACT
BACKGROUND: For years the Heidelberg Retina Tomograph has been an established method to diagnose early glaucomatous damages at the optic nerve head. The major difficulty consists in defining the outlayer of the optic nerve head in a 2-dimensional reflective or topographic picture. A 3-dimensional presentation of the ocular fundus could ease the defining very much. The Dresden 3D-display tested provides a true 3-dimensional presentation of the HRT-values measured. METHODS: For the study 5 groups of prediagnosed follow-up examinations were formed, which in their course showed various progression. The examination firstly was carried out by manual defining of the papilla contour on the Dresden 3D-display and secondly by the new automatic contour-finding of the 3D-software. RESULTS: Through the 3-dimensional presentation a more correct position of the contour can be found when drawn manually. As a result some diagnoses of the 2-dimensional HRT-pictures had to be revised. The results proved considerable differences between the automatic outlayer-finding and the probably real position of the papilla borders. CONCLUSION: The 3-dimensional presentation of the HRT-pictures represents an subjective improvement regarding the accuracy of manual contour definition. In particular, the fact that it is feasible to make the pictures rotate around the axes offers completely new insights to the morphology of the disc.
