Simulation of the Radiography Formation Process from CT Patient Volume

The aim of this work is to develop an algorithm to simulate the radiographic image formation process using volumetric anatomical data of the patient, obtained from 3D diagnostic CT images. Many applications, including radiographic driven surgery, virtual reality in medicine and radiologist teaching and training, may take advantage of such technique. The designed algorithm has been developed to simulate a generic radiographic equipment, whatever oriented respect to the patient. The simulated radiography is obtained considering a discrete number of X-rays paths departing from the focus, passing trough the patient volume and reaching the radiographic plane. To evaluate a generic pixel of the simulated radiography the cumulative absorption along the correspondent X-ray is computed. To estimate X-ray absorption in a generic point of the patient volume, 3D interpolation of CT data has been adopted. The proposed technique is quite similar to those employed in Ray Tracing. A computer designed test volume has been used to assess the reliability of the radiography simulation algorithm as a measuring tool. From the errors analysis emerges that the accuracy achieved by the radiographic simulation algorithm is largely confined within the sampling step of the CT volume