Analisi tecnica degli algoritmi di volume rendering: Applicazione alle strutture a basso contrasto usando come modello la vascolarizzazione epatica

Translated title of the contribution: Technical analysis of volume-rendering algorithms: Application in low-contrast structures using liver vascularisation as a model

Filippo Cademartiri, Giacomo Luccichenti, Giuseppe Runza, Massimo Gualerzi, Lorenzo Brambilla, Paolo Coruzzi, Paolo Soliani, Mario Sianesi, Tommaso Vincenzo Bartolotta, Massimo Midiri

Research output: Contribution to journalArticlepeer-review

Abstract

Aim. To assess the influence of pre-set volume-rendering opacity curves (OC) on image quality and to identify which absolute parameters (density of aorta, hepatic parenchyma and portal vein) affect visualization of portal vascular structures (low-contrast structures). Materials and methods. Twenty-two patients underwent a dual-phase spiral CT with the following parameters: collimation 3 mm, pitch 2, increment 1 mm. Three scans were performed: one without contrast medium and the latter two after the injection of contrast material (conventionally identified as "arterial" and "portal"). The images were sent to a workstation running on an NT platform equipped with post-processing software allowing three-dimensional (3D) reconstructions to generate volume-rendered images of the vascular supply to the liver. Correlation between the absolute values of aorta, liver and portal vein density, OC parameters, and image quality were assessed. Results. 3D images generated using pre-set OC obtained a much lower overall quality score than those produced with OC set by the operator. High contrast between the liver and the portal vein, for example during the portal vascular phase, allows wider windows, thus improving image quality. Conversely, the OC in the parenchymal phase scans must have a high gradient in order to better differentiate between the vascular structures and the surrounding hepatic parenchyma. Conclusions. Image features considered to be of interest by the operator cannot be simplified by the mean of pre-set OC. Due to their strong individual variability automatic 3D algorithms cannot be universally applied: they should be adapted to both image and patient characteristics.

Translated title of the contributionTechnical analysis of volume-rendering algorithms: Application in low-contrast structures using liver vascularisation as a model
Original languageItalian
Pages (from-to)376-384
Number of pages9
JournalRadiologia Medica
Volume109
Issue number4
Publication statusPublished - 2005

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging

Fingerprint Dive into the research topics of 'Technical analysis of volume-rendering algorithms: Application in low-contrast structures using liver vascularisation as a model'. Together they form a unique fingerprint.

Cite this