Keywords:
Cardiovascular system, CT-Angiography, Computer Applications-3D, Aneurysms, Arteriosclerosis, Dissection
Authors:
M. Finazzo, F. Finazzo, C. Duranti, A. Dell'Era, F. Pinto, D. Messana
DOI:
10.26044/ecr2023/C-16007
Conclusion
GIR has been widely used in movies, video games, lighting engineering and flight simulators but has only recently been adopted in medical imaging. GIR is superior to the conventional volume rendering (VR) technique. Actually, in VR each pixel is formed by one light ray. Conversely, in GIR each pixel results from the simulation of the interaction of billions of light rays - propagating from multiple light sources - with a scanned object. In other words, GIR mimics the real-life propagation, absorption, scattering of light and eventual color transmission. Thus, it provides highly photorealistic 3D images, with enhanced depth perception and better visualization of fine details. Moreover, GIR allows for immediate perception of the shapes, contours, locations and relationships of various objects. What’s more, shadowing effects are improved and segmentation processes can be used. A minor limitation of GIR is that shadows can darken areas that would otherwise be well lit when using VRT. In our experience GIR representation of coronary arteries and thoraco-abdominal aorta provided more sculptural quality and more natural shape and depth perception than VR images . These properties were striking in the depiction of saccular aneurisms of aorta and coronary arteries
. As much as VR - GIR allowed for detection of coronary stenosis caused by a soft plaque -
- not by a calcific plaque, because the latter overshadowed stenosis. Moreover, GIR could not provide any information about coronary stent patency
. Thus, the assessment of coronary artery stenosis and stent patency will continue to rely predominantly on axial images and multiplanar 2D reconstructions. However, GIR provided a comprehensive view of the coronary artery tree, with an immediate visualization of the extent and location of disease. It was also invaluable in the assessment of congenital anomalies and in the preprocedural and postprocedural evaluation of interventions (e.g. aortic aneurysms and dissection coronary artery bypass grafting). GIR can be very helpful in patient engagement and trainee education and in some instances may obviate the need for 3D printing for pre-surgical planning. Finally, it can improve communication during clinical consultation or multidisciplinary meetings. This work has at least three limitations. It was basically focused on the differences in image appearance between GIR and VR. No analysis of diagnostic performance was carried out. Finally, radiologists score assignment was subjective because no objective way was found to assess 3-D rendering of CT volumetric data.