Three-Dimensional Cinematic Rendering to Optimize Visualization of Cerebrovascular Anatomy and Disease in CT Angiography.

Journal Article (Journal Article)

Three-dimensional cinematic rendering (3DCR) is an emerging postprocessing technique for computed tomography (CT) and CT angiography (CTA) that produces photorealistic, volumetric images. In contrast to conventional volume rendering techniques, 3DCR depicts life-like shadowing and surface reflection, which can improve the perception of depth and complex anatomic spatial relationships. This tool allows clinical neuroimagers to study, explore, and teach the complex relational anatomy of the cerebral vessels and skull in a more intuitive manner. The purpose of this report is to introduce the physical and optical principles behind 3DCR and to explore applications of 3DCR in modern cerebrovascular imaging. Using CTA source data, we describe our approach to visualizing cerebrovascular anatomy and disease and introduce three simple, reproducible techniques through a series of case vignettes. First, we show how selective manipulation of rendered models can imitate cadaveric dissection. Next, we discuss surface rendering as a means of recapitulating the neurologic physical exam. Last, we provide a step-by-step method of simulating the operating room perspective in visualizing cerebrovascular disease. In our experience, 3DCR proves most useful for visualizing structures at the vessel-skull interface, which can be difficult to assess with conventional imaging methods. 3DCR, therefore, complements traditional 2-dimensional and 3-dimensional imaging methods and serves as an emerging tool for neuroimagers to communicate with and educate other clinicians.

Full Text

Duke Authors

Cited Authors

  • Caton, MT; Wiggins, WF; Nunez, D

Published Date

  • May 2020

Published In

Volume / Issue

  • 30 / 3

Start / End Page

  • 286 - 296

PubMed ID

  • 32072719

Pubmed Central ID

  • 32072719

Electronic International Standard Serial Number (EISSN)

  • 1552-6569

Digital Object Identifier (DOI)

  • 10.1111/jon.12697

Language

  • eng

Conference Location

  • United States