Impact of vertical and horizontal malrotation on measurements of anteroposterior radiographs of the scapula: need for standardized images in modern omometry

BACKGROUND An increasing number of parameters measured on anteroposterior radiographs are used for the evaluation of the bony geometry of the scapula. Inhomogeneous acquisition of images is common because of the lack of standardization in radiographic positioning. Images with malrotation around the horizontal axis of the scapula are particularly frequent. We hypothesized that malrotated images would result in large variations in measured radiographic parameters and that image standardization using qualitative and semiquantitative "omometric" criteria would decrease these variations in measurements. "Omometry" is a newly introduced umbrella term that contains all standardized measurements on plain radiographs of the shoulder, analogous to the term "coxometry," which is widely used for the radiographic assessment of the osseous pelvis and hip. METHODS In this experimental, cadaveric radiographic study, 7 dry-bone human scapula cadaveric specimens from anonymous donors were used to obtain 210 radiographs. We incrementally rotated (steps of 3°) every scapula around its horizontal and vertical axis, with a total range of 42° per each axis. Then, we measured 5 radiographic parameters on every image and observed their change with malrotation. Furthermore, we introduced 4 omometric criteria defining an appropriate (presence of ≥3 criteria) radiographic image to improve standardization of scapular image acquisition. RESULTS Overall, measured values remained stable within a narrow range of ±9° of malrotation. Beyond this range, values of all parameters significantly deviated (>±2°) from the initial value. Measurements on appropriate images were significantly less prone to deviation. Within the appropriate images, those with 4 criteria showed a higher specificity than those with 3 criteria. CONCLUSION There is significant variation in values of measured radiographic parameters on anteroposterior radiographs of the scapula with substantially malrotated images. With the use of the 4 newly introduced semiquantitative and qualitative omometric criteria, which define an appropriate image, reliability of the measured parameters can be significantly improved