Evaluation of computer-aided detection and dual energy software in detection of peripheral pulmonary embolism on dual-energy pulmonary CT angiography

Choong Wook Lee, Joon Beom Seo, Jae Woo Song, Mi Young Kim, Ha Young Lee, Yang Shin Park, Eun Jin Chae, Yu Mi Jang, Namkug Kim, Bernard Krauss

Research output: Contribution to journalArticlepeer-review

69 Citations (Scopus)

Abstract

Purpose: To evaluate the sensitivity of computer-aided detection(CAD) and dual-energy software('Lung PBV', 'Lung Vessels') for detecting peripheral pulmonary embolism(PE). Materials and methods: Between Jan-2007 and Jan-2008, 309 patients underwent dual-energy CT angiography(DECTA) for the evaluation of suspected PE. Among them, 37 patients were retrospectively selected; 21 with PE at segmental-or-below levels and 16 without PE according to clinical reports. A standard computer assisted detection (CAD) package and two new types of software('Lung PBV', 'Lung Vessels') were applied on a dedicated workstation. This resulted in four alternative tests for detecting PE: DECTA alone and DECTA with CAD, 'Lung Vessels' and 'Lung PBV'. Two radiologists independently read all cases at different reading sessions. Two thoracic radiologists set the reference standard by combining all information from DECTA and software. The sensitivity of detection for all, segmental and subsegmental-or-below PE were assessed. Results: The reference standard contained 136 PE(segmental 65, subsegmental-or-below 71). With DECTA alone, the sensitivity of detection for all, segmental and subsegmental-or-below pulmonary arteries was 54.5%/73.7%/34.4%; DECTA with CAD, 57.8%/76.8%/37.9%; DECTA with 'Lung PBV', 61.1%/79.9%/41.4%; DECTA with 'Lung Vessels', 64.0%/78.3%/48.5% respectively. Conclusion: The use of CAD, Lung Vessels and Lung PBV shows improved capability to detect peripheral PE.

Original languageEnglish
Pages (from-to)54-62
Number of pages9
JournalEuropean Radiology
Volume21
Issue number1
DOIs
Publication statusPublished - 2011 Jan
Externally publishedYes

Keywords

  • CT
  • Computer-aided diagnosis
  • Diagnosis
  • Dual-energy CT
  • Pulmonary embolism

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging

Fingerprint

Dive into the research topics of 'Evaluation of computer-aided detection and dual energy software in detection of peripheral pulmonary embolism on dual-energy pulmonary CT angiography'. Together they form a unique fingerprint.

Cite this