渲染(计算机图形)
体绘制
计算机科学
计算机视觉
三维渲染
人工智能
感知
可视化
锥束ct
计算机图形学(图像)
计算机断层摄影术
心理学
医学
放射科
神经科学
作者
Tobias Steffen,Sebastian Winklhofer,Felicitas Starz,Daniel B. Wiedemeier,Uzeyir Ahmadli,Bernd Stadlinger
标识
DOI:10.1016/j.aanat.2022.151905
摘要
The aim of this exploratory study is to analyse whether three-dimensional cinematic rendering image reconstructions offer advantages over conventional volume rendering in the visualisation of cone beam computed tomography (CBCT) and computed tomography (CT) images of the facial skeleton. This is of interest, as some information gets lost during the rendering process. This especially applies to structures in the background of the image and some surface information which can be lost. The commonly applied two-dimensional representation of CBCT or CT images in three different axes requires experience for interpretation. Cinematic rendering is a new three-dimensional post processing reconstruction technique, creating photo realistic visualisations, thus possibly enabling an easier interpretation of the images. In this study, ten investigators assessed ten separate patient cases of the orofacial skeleton. For each case, a conventional volume rendering image reconstruction and a cinematic rendering reconstruction of the same area was created. A specially designed questionnaire assessed both objective and subjective criteria of image perception. Objective criteria were assessed by predefined questions on the visual perception of anatomical image characteristics, showing the two reconstruction types of each case randomly to the investigators in two sessions. Subjective criteria were assessed via a visual analogue scale, showing both reconstructions simultaneously in a third session. The results show that cinematic rendering offers advantages especially in the evaluation of depth perception and three-dimensionality. Volume rendering shows advantages in surface sharpness. Cinematic Rendering was subjectively rated higher for almost all reconstructions. The cinematic rendering process however may cause loss of information and blurring of surfaces compared to volume rendering. With respect to the subjective impression, cinematic rendering scored better than volume rendering. The visualisation is perceived as being very close to reality.
科研通智能强力驱动
Strongly Powered by AbleSci AI