Radiotranslucent foam isolation blocks improve regional tissue accuracy in whole body DXA scans

Introduction DXA regional body composition accuracy and precision are critically important in many clinical applications including monitoring lymphedema, performance training, and injury recovery. While total body composition measurements are highly precise, regional measurements exhibit lower precision partly due to inconsistencies in subject positioning. In particular, tissue overlap between body regions (such as breast tissue from the trunk overlying the arms, or hands overlapping the hip) limits the ability to follow regional changes in fat and lean masses, as well as symmetry measures. We propose a modified DXA scan protocol that includes radiolucent isolation blocks to perfect positioning and regional accuracy. Objective To develop a protocol for whole body DXA scanning that optimizes the trueness and precision of arm, trunk, and leg regions by preventing tissue overlap. Methods Four candidate materials were evaluated for X-ray measured density, stiffness, and cost. Sheets of the selected material were cut and tapered into wedges to fit between the arms and trunk, and between the legs. A pilot recruitment of 15 healthy adults underwent whole body DXA scans on a Hologic Discovery/W. Density of each material was assessed by scanning with the Hologic Whole Body Phantom. Each participant had 2 scans with the foam positioners (with repositioning), and 1 without. All scans were analyzed by a single ISCD-certified technologist. Custom subregions were defined to isolate the arms from the trunk during analysis. Differences in regional masses with and without the foam positioners were calculated. Results Of the four materials tested, 1”-thick closed-cell polyethylene foam sheet was optimal due to low cost ( 2 ), low density (0.03 g/cm 3 ), and high comfort. Scans of the foam with the phantom confirmed its radiolucent properties and agreement between DXA-measured density and physical density. Currently, 5 female participants have finished the protocol. We found that special analysis was still needed to insure arm and trunk separation. A typical example is shown in Figure 1 . Compared to baseline scans, arm and trunk region mass measurements were significantly different with the use of foam positioners. Arm fat and lean changes with the positioners were observed as large as 120 g and 180 g, respectively. Further characterization is underway to assess precision of this technique. Conclusions Polyethylene foam positioning blocks can be used to ensure clear, repeatable delineation of soft tissue between anatomical regions. This technique may improve DXA regional body composition accuracy and precision, improving the ability of DXA to monitor changes in individual limb masses.