HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Brown, M. S. Articles by Aberle, D. R. Search for Related Content PubMed PubMed Citation Articles by Brown, M. S. Articles by Aberle, D. R.

Lung Micronodules: Automated Method for Detection at Thin-Section CT—Initial Experience¹

¹ From the Department of Radiology, David Geffen School of Medicine at UCLA, 10833 Le Conte Ave, Los Angeles, CA 90095-1721. Received October 18, 2001; revision requested January 11, 2002; revision received March 11; accepted May 7, 2002. Address correspondence to M.S.B. (e-mail: mbrown@mednet.ucla.edu).

View larger version (20K): [in a new window]   Figure 1. Schematic shows simplified overview of the a priori anatomic model, which contains nodes that describe the lungs and nodules within the lungs. The arcs (arrows) between the nodes represent structural relationships between the respective anatomic entities.

View larger version (12K): [in a new window]   Figure 2. Schematic shows fuzzy membership function that describes the possible range of values for the sphericity of a nodule.

View larger version (141K): [in a new window]   Figure 3A. Automated segmentation and nodule detection results. A, Original transverse thin-section CT image. B, Transverse lung (shaded). C, Peripheral lung (shaded). D, Opacities within the lung parenchyma (gray). E, Two nodules (arrows) detected by the system that were confirmed by radiologists.

View larger version (149K): [in a new window]   Figure 3B. Automated segmentation and nodule detection results. A, Original transverse thin-section CT image. B, Transverse lung (shaded). C, Peripheral lung (shaded). D, Opacities within the lung parenchyma (gray). E, Two nodules (arrows) detected by the system that were confirmed by radiologists.

View larger version (149K): [in a new window]   Figure 3C. Automated segmentation and nodule detection results. A, Original transverse thin-section CT image. B, Transverse lung (shaded). C, Peripheral lung (shaded). D, Opacities within the lung parenchyma (gray). E, Two nodules (arrows) detected by the system that were confirmed by radiologists.

View larger version (140K): [in a new window]   Figure 3D. Automated segmentation and nodule detection results. A, Original transverse thin-section CT image. B, Transverse lung (shaded). C, Peripheral lung (shaded). D, Opacities within the lung parenchyma (gray). E, Two nodules (arrows) detected by the system that were confirmed by radiologists.

View larger version (140K): [in a new window]   Figure 3E. Automated segmentation and nodule detection results. A, Original transverse thin-section CT image. B, Transverse lung (shaded). C, Peripheral lung (shaded). D, Opacities within the lung parenchyma (gray). E, Two nodules (arrows) detected by the system that were confirmed by radiologists.

View larger version (166K): [in a new window]   Figure 4. Transverse CT images with targeted field of view show nodules detected automatically by the system. A, Micronodule (arrow) was initially missed by the radiologist, who used visual inspection, but it was identified and confirmed by that radiologist when system results were made available. B, Larger nodule (lower arrow) was detected by the radiologist without the system, but the micronodule (upper arrow) was initially missed until the system results were available. C, D, Nodules (arrows [confirmed by the truth committee]) were detected by the system but were rejected by the radiologist. It is particularly difficult to distinguish nodules from vessels on a two-dimensional image (as shown here), but the system performs segmentation and object recognition in three dimensions (with use of information from adjacent sections).