Improved methods for quantitative radiographic evaluation with particular reference to total-hip arthroplasty.

The range of magnification factors in 40'' and 72'' radiographs were calibrated using three simple mathods. The relationship between X-ray source to cassette distance, object to cassette distance and magnification factors was determined experimentally for both 40'' and 72'' radiographs. The resulting nomogram showed that: (1) there was no significant magnification difference between centered (AP-hip) and offset (AP-pelvis) radiographs; (2) magnification factors of 40'' films were twice those of 72'' films for the same object to cassette distances; and (3) due to the smaller magnification range of 72'' films (1 to 13%) compared to 40'' films (1 to 26%) there was less chance of error in determining magnification factors for the larger film distances. The skin radiation dosage for the patient was similar in both procedures. Magnification factors were also calculated from 414 A-P radiogrphs of total-hip prosthesis, using the femoral ball as a calibration standard. This provided magnifications of 20 per cent +/- 6(2SD) for 40'' films and by extrapolation on the nomogram gave corresponding values of 10 per cent +/- 3 for the 72'' films. The associated hip to cassette distances (L), also obtained from the nomogram, varied from 125 to 205 mm, average 165 mm. In subsequent clinical trials using 72'' films, it was possible to represent the patients as either small (L less than 145 mm) medium (145 less than or equal to L less than or equal to 185 mm) or large (L greater than 185 mm) with corresponding magnification factors of 8, 10 and 12 per cent, respectively. The error with this simplification was still within +/-21/2 per cent. For 40'' films the corresponding magnification factors for small, medium or large patient categories were 16, 20 and 24 per cent respectively with likely errors within +/-5 per cent. Magnification factors were also calculated from the radiographic image of a 100 mm long rod which had been attached to the patien's thigh. The overall accuracy with this method was comparable to the techniques described above and had the advantage that every radiograph featured its own scale, a particularly important feature for preoperative films which have no implant for calibration purposes. As a result of such simple calibration procedures, it was possible to identify the overall magnification variations and specify particular magnification factors within +/-21/2 per cent (72'' films) or +/-5 per cent (40'' films).