PubMed Health⌕ Search
Home/Archive/PubMed

PubMed · 10887263

Automated polyp detector for CT colonography: feasibility study.

R M Summers·C F Beaulieu·L M Pusanik·J D Malley·R B Jeffrey·D I Glazer·S Napel
View original source ↗Export metadata ↓

Abstract

An abdominal computed tomographic scan was modified by inserting 10 simulated colonic polyps with use of methods that closely mimic the attenuation, noise, and polyp-colon wall interface of naturally occurring polyps. A shape-based polyp detector successfully located six of the 10 polyps. When settings that enhanced the edge profile of polyps were chosen, eight of 10 polyps were detected. There were no false-positive detections. Shape analysis is technically feasible and is a promising approach to automated polyp detection.

Explore related subjects

ColonColonic PolypsComputer GraphicsComputer SimulationFeasibility StudiesHumansImage Processing, Computer-AssistedMaleMiddle AgedRadiographic Image Interpretation, Computer-AssistedSensitivity and SpecificityTomography, X-Ray ComputedJournal ArticleResearch Support, Non-U.S. Gov'tResearch Support, U.S. Gov't, P.H.S.

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

R M Summers, C F Beaulieu, L M Pusanik, J D Malley, R B Jeffrey, D I Glazer, S Napel. 2000. Automated polyp detector for CT colonography: feasibility study.. https://doi.org/10.1148/radiology.216.1.r00jl43284

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related citations

PubMed2007

Expression of the mitotic checkpoint gene MAD2L2 has prognostic significance in colon cancer.

Aneuploidy and genetic instability are a hallmark of colorectal cancer and other solid tumors, and they are thought to enhance tumor progression. The gene MAD2L2 (mitotic arrest deficient 2-like 2) encodes the spindle checkpoint protein MAD2L2 (or MAD2B), a key component of a surveillance system that delays anaphase until all chromosomes are correctly oriented. Defects in this mitotic checkpoint are known to contribute to genetic instability, i.e., numerical and structural aberrations of chromosomes. We have previously identified MAD2L2 as significantly upregulated in locally restricted colorectal tumors by gene expression profiling. So far, MAD2L2 has not been reported to play a major role in human cancer in contrast to its homologue MAD2. To address this question, 118 histologically confirmed colorectal lesions were analyzed by quantitative real-time PCR for expression of MAD2L2, and compared to normal colon tissue from 11 patients. Twenty-five out of 118 tumor samples (21%) showed MAD2L2 overexpression of 3-fold or more compared to normal colon, and the fraction of overexpressing tumors increased with tumor stage. Correspondingly, protein levels of MAD2L2 were found to be significantly upregulated in tumors as compared to matched normal tissue. Tumors with upregulated MAD2L2 expression had significantly higher numbers of aberrant mitotic figures (anaphase bridges), an indication of chromosomal instability. Elevated expression of MAD2L2 was significantly correlated with reduced patient survival. By multivariate analysis, MAD2L2 expression was retained as an independent prognostic parameter for patient survival. Thus, our results demonstrate that overexpression of MAD2L2 correlates with bad prognosis in colorectal cancer.

Colon↗
PubMed2007

CT colonography: automatic measurement of polyp diameter compared with manual assessment - an in-vivo study.

AIM: To investigate whether automated diameter assessment was feasible for CT colonography. MATERIALS AND METHODS: Two experienced observers independently measured the maximum diameter of 50 polyps (colonoscopic reference size range 5-12 mm) from colonography datasets using conventionally placed software callipers and a variety of two-dimensional (2D) computed tomography (CT) window settings (colon, abdominal, bone, lung), and also three-dimensional (3D) perspective rendering. Polyps were also measured using automated polyp-segmentation software. Agreement between observers and with the colonoscopic reference measurement was determined using Bland-Altman, Wilcoxon, and Mann-Whitney U analyses. RESULTS: Inter-observer agreement was similar for all window displays: mean difference in millimetres (SD difference; 95% limits of agreement) ranged from 0 (1.7, -3.3, 3.3) for 2D colon to -1.1mm (1.6, -4.3, 2.0) for 3D, compared with -0.5 (2.09, -4.6, 3.6) for automated measurement. When compared to colonoscopy, the largest discrepancy occurred using the 3D display (mean difference 1.3mm, 2.5mm for each observer). There was also a significant difference between estimates and reference size when using the 2D abdominal and 3D displays (p=0.03, <0.001). CONCLUSION: Automated polyp measurement is possible in vivo. Automated and conventional methods have comparable inter-observer agreement. The greatest measurement error is encountered when using a 3D display for estimates of diameter.

Colon↗