Air gap technique is recommended in axiolateral hip radiographs

Abstract Purpose To investigate the replacement of conventional grid by air gap in axiolateral hip radiographs. The optimal air gap distance was studied with respect to radiation dose and image quality using phantom images, as well as 26 patient axiolateral hip radiographs. Methods The CDRAD phantom, along with polymethylmethacrylate slabs with thicknesses of 10.0, 14.6, and 20.0 cm was employed. The inverse image quality index and dose area product (DAP), as well as their combination, so called figure‐of‐merit (FOM) parameter, were evaluated for these images, with air gaps from 20 to 50 cm in increments of 10 cm. Images were compared to those acquired using a conventional grid utilized in hip radiography. Radiation dose was measured and kept constant at the surface of the detector by using a reference dosimeter. Verbal consent was asked from 26 patients to participate to the study. Air gap distances from 20 to 50 cm and tube current‐time products from 8 to 50 mAs were employed. Exposure index, DAP, as well as patient height and weight were recorded. Two radiologists evaluated the image quality of 26 hip axiolateral projection images on a 3‐point nondiagnostic — good/sufficiently good — too good scale. Source‐to‐image distance of 200 cm and peak tube voltage of 90 kVp were used in both studies. Results and conclusion Based on the phantom study, it is possible to reduce radiation dose by replacing conventional grid with air gap without compromising image quality. The optimal air gap distance appears to be 30 cm, based on the FOM analysis. Patient study corroborates this observation, as sufficiently good image quality was found in 24 of 26 patient radiographs, with 7 of 26 images obtained with 30 cm air gap. Thus, air gap method, with an air gap distance of 30 cm, is recommended in axiolateral hip radiography.


| INTRODUCTION
Air gap technique is a well-known method to reduce the amount of scattered x-ray radiation reaching the detector, thus reducing noise and improving image contrast. 1 It is rather commonly utilized instead of a conventional grid in plain radiography. [2][3][4][5][6] Air gap is an additional distance between a patient and an image detector. 7 The gap decreases the likelihood for scattered x-ray radiation to reach the detector, as radiation is partially absorbed and scattered in the air. 8,9 Air gap technique offers advantages over conventional grids, as the latter may, for example, lead to image artifacts, typically related to the misalignment of the grid. In the air gap technique, the object to image distance (OID) is increased compared to the imaging with a conventional grid, which results in magnified image. 8,9 To reduce magnification, source to image distance (SID) can be increased, albeit in some cases imaging geometry may pose limitations for this increase. In recent years, so called virtual grid techniques have also been proposed. 10,11 These techniques model scattered radiation mathematically, and at least partially remove its effects from the image to enhance contrast.
Early studies on air gap methods concentrated on chest radiographs. 12,13 These studies indicate that images acquired with air gap technique can provide contrast similar to images acquired using conventional grids, in particular for the posterior-anterior projection 12 with no increase in the patient exposure. Later, Chan and Fung 14 reported 10 cm to be the optimal air gap distance at pelvic anterior-posterior (AP) examination. They reported that the effective dose was reduced by a factor of roughly two in both computed radiography (CR) and digital radiography (DR) examinations by replacing antiscatter grid with an air gap of 10 cm, while image quality was observed to be diagnostic. On the contrary, Moey et al.
reported that pelvic radiographs obtained with air gap distances <20 cm were not fully acceptable. 15 Trimble 16 assessed image quality of both lateral thoracic spine and chest radiographs using conventional grid, as well as air gap techniques and concluded that image quality was higher with the latter. Bell et al. 5  In horizontal beam lateral or axiolateral hip projection, the air gap technique has been shown to reduce radiation dose, and demonstrates diagnostic image quality, as compared to the use of conventional grid. 18,19 At our institution, examinations of the hip comprise roughly 5% of all conventional radiographic studies. Most examinations involve the axiolateral projection with horizontal x-ray beam, a common protocol for patients with hip trauma. This projection has also been employed to calculate cup anteversion after hip arthroplasty surgery. [20][21][22] The aim of the study was to investigate whether the conventional grid can be replaced by an air gap technique in the hip axiolateral radiographic projection. The optimal air gap distance was sought by acquiring images with a contrast-detail phantom with different air gap distances. Both image quality and dose area product (DAP), as well as their combination, so called figure-of-merit (FOM) parameter were utilized in the analysis. A small sample of patient hip axiolateral radiographs were analyzed in terms of air gap distances to verify the results found in the phantom study.

| MATERIALS AND METHODS
This study was performed in two phases; first phase comprised of phantom measurements, and the second phase comprised of a patient study (institutional review board approved the study, 183/ 2019). All images were obtained with Fuji FDR Acselerate system  For each measurement setting, we exposed CDRAD phantom three times. Inverse image quality parameter IFQ inv was determined 23 as an average of three images. IQF inv is a measure of both image contrast and resolution, and is defined as where C i and D i denote contrast and threshold diameter in column i, respectively. A higher IQF inv indicates higher image quality as more and smaller objects are detected from the CDRAD phantom images.
The DAP values were calculated as averages from the three expo-

| RESULTS
In CDRAD phantom measurements (see Fig. 3 Fig. 3(a)]. On the contrary, in the case of 20.0 cm PMMA slab, IQF inv yielded a maximum value of 6.6 at an air gap of 40 cm. Furthermore, IQF inv decreased from 6.6 to 6.3 when air gap distance increased from 40 to 50 cm. This is might be due to the amount of primary radiation reaching the detector is starting to decrease at the latter distance.  Table 1). The  Table 1), respectively. The calculated interobserver reliability was poor (kappa = −0.026).

| DISCUSSION
The most suitable air gap distance not only involves a reduction in patient dose but also maintains the diagnostic image quality. Thus, a compromise between the two entities is to be sought. Consequently, we employed the FOM parameter to evaluate the image quality with respect to dose. We observed that the image quality, as determined by the IQF inv parameter in the phantom setup, was always higher with any of the air gap distances employed than with the conven-

| CONCLUSIONS
Based on the FOM analysis combining the effects of image quality and radiation dose in the current phantom setup, the optimal air gap distance in the axiolateral hip radiographs appears to be 30 cm. With this air gap distance, all patient hip radiographs were evaluated either too good or good/sufficiently good. A single exception occurs for a hip radiograph of an obese patient with insufficient tube current-time product value. As some of the recorded imaging parameters of the patient study imply, careful, and continuous education on different radiographic techniques in a real clinical environment is essential for diagnostic radiography.

ACKNOWLEDGMENTS
Marianne Haapea is acknowledged for help with the statistical analysis related to the interobserver reliability. We thank the anonymous reviewers for valuable comments during the review process of the manuscript.

AUTHORS' CONTRI BUTIONS
SK was involved in conceptualization, data curation, formal analysis, investigation, methodology, and writingoriginal draft preparation and review/editing; AK was involved in conceptualization, data curation, formal analysis, investigation, methodology, visualization, and writingoriginal draft preparation and review/editing; AH was involved in conceptualization, methodology, and writingreview and editing; TN was involved in data curation, formal analysis, and writingreview and editing; JN was involved in conceptualization, methodology, resources, data curation, formal analysis, and writing review and editing; MTN was involved in conceptualization, methodology, project administration, resources, supervision, and writingreview and editing; MH was involved in conceptualization, data curation, formal analysis, investigation, methodology, project administration, supervision, visualization, and writingoriginal draft preparation and review/editing.

CONFLI CT OF INTEREST
No conflict of interest.