Roentgen Stereophotogrammetry Applied to the Cleft Maxilla of Infants: I. Implant Technique

Abstract

A new roentgen stereophotogrammetric method was developed for the study of the kinematics of the skeletal system. Based upon the use of metallic implants, this method determines the relationship between the implants as well as the change in position between skeletal segments provided with implants. The object and its positional changes or motion are reconstructed from two roentgenograms with the aid of photogrammetric principles and computer technique. The method requires insertion of three or more implants in each skeletal segment. When the implants remain in a stable interrelationship during the observation period a rigid-body model is obtained. In the analysis this represents the segment and motion is computed according to the kinematics of a rigid body. The present study is concerned with early establishment of rigid-body models in the segments of a cleft maxilla (part I) and registration of postoperative motion between the segments (part II). The experimental set-up consists of two X-ray tubes (0.6×0.6 mm focus) and a Plexiglas cage, in which the object is placed. The focus-film distances are 900 to 1 000 mm and do not have to be identical from time to time as the positions of the two foci are calculated in the photogrammetric reconstruction for each examination. Two films are exposed simultaneously. The cage is made up of four flat Plexiglas plates oriented at right angles to each other and provided with 9 indicators (balls of tantalum 0.5 mm in diameter) in each plate, which are recorded on the same films as the object. In the analysis of the roentgenograms the center of the images of all implants and indicators is focused in a coordinatograph and the X -and Y-coordinates are recorded. After perspective transformation of the images to the laboratory coordinate system and computation of the positions of the two foci the positions of the implants in the laboratory system are determined as the X-, Y- and Z-coordinates of the intersection of reconstructed lines between the foci and the transformed images of the object. The configuration of the rigid-body model is reconstructed and the mean error of rigid-body fitting computed. Motion between the models is calculated and described in terms of translation and rotation along and about the body-fixed X-, Y- and Z-axes. Three infants are examined, who at surgery have been provided with implants (tantalum pins, 1.5×0.5 mm) representing 6 maxillary segments. A.B., complete bilateral clefts of the lip and palate primary lip repair at 3.5 months of age; A. L., similar defects primary palate repair at 12.5 months of age and T. L., complete leftsided clefts of the lip and palate, primary palate repair at 9.5 months of age. Of 21 inserted implants 17 have remained in the bone during observation periods of 898, 878 and 577 days, suggesting that in the examined children the amount of surface resorption in these areas has been moderate. The mean error of the six rigid-body models obtained in the precision test varies between 14 and 73 micrometers (μm). The mean value is 36 μm. In the accuracy test, the mean error for the six rigid-body models varies between 9 and 105 μm, the mean value being 52 μm.