Digitally Reconstructed Radiograph
A Digitally Reconstructed Radiograph (DRR) is a simulation of a conventional 2D x-ray image, created from computed tomography (CT) data. A radiograph, or conventional x-ray image, is a single 2D view of total x-ray absorption through the body along a given axis. Two objects (say, bones) in front of one another will overlap in the image. By contrast, a 3D CT image gives a volumetric representation. (Earlier CT data sets were better thought of as a set of 2D cross sectional images.) Sometimes one must compare CT data to a classical radiograph, and this can be done by comparing a DRR based on the CT data. An early example of their use is the beams eye view (BEV) as used in radiotherapy planning. In this application, a BEV is created for a specific patient and is used to help plan the treatment.
DRRs are created by summing CT intensities along a ray from each pixel to the simulated x-ray source.
Since 1993, the Visible Human Project (VHP) has made full body CT data available to researchers. This has allowed several universities and commercial companies to try and create DRR's. These have been suggested as useful for training simulations in Radiology and Diagnostic Radiography. It takes a significant number of calculations to create a summative 2D image from a large amount of 3D data. This is an area of medical science and education that has benefited from the advancing of graphics card technology, driven by the computer games industry.
Another novel use of DRR's is in identification of the dead from old radiographic records, by comparing them to DRR's created from CT data.