Computed tomography colonography: automated diameter and volume measurement of colonic polyps compared with a manual technique--in vitro study

Polyp size measurement at CT colonography: what do we know and what do we need to know?

Polyp size is a critical biomarker for clinical management. Larger polyps have a greater likelihood of being or of becoming an adenocarcinoma. To balance the referral rate for polypectomy against the risk of leaving potential cancers in situ, sizes of 6 and 10 mm are increasingly being discussed as critical thresholds for clinical decision making (immediate polypectomy versus polyp surveillance) and have been incorporated into the consensus CT Colonography Reporting and Data System (C-RADS). Polyp size measurement at optical colonoscopy, pathologic examination, and computed tomographic (CT) colonography has been studied extensively but the reported precision, accuracy, and relative sizes have been highly variable. Sizes measured at CT colonography tend to lie between those measured at optical colonoscopy and pathologic evaluation. The size measurements are subject to a variety of sources of error associated with image acquisition, display, and interpretation, such as partial volume averaging, two- versus three-dimensional displays, and observer variability. This review summarizes current best practices for polyp size measurement, describes the role of automated size measurement software, discusses how to manage the measurement uncertainties, and identifies areas requiring further research.

Computed tomographic colonography: automated tool for polyp measurement delivering on patient risk stratification

OBJECTIVE

We evaluated an automated polyp size measurement tool in computed tomographic colonography for its accuracy and value for patient risk stratification.

METHODS

A simulation program generated a raw data phantom with sessile and pedunculated polyps of known sizes using 120 to 140 kV and 50, 40, 20, 15, and 10 mAs. All polyps were measured by clicking on the polyp surface. Comparison of the calculated size with the known polyp sizes allowed calculation of reproducibility and accuracy. For patients with proven polyps, we also compared automated measurements with manual and endoscopic measurements to evaluate the effect on patient risk stratification.

RESULTS

The automated measurement tool allowed accurate measurements. In the patient study, assignment to the correct size group was not significantly different from the radiologist's results. However, it slightly improved patient risk stratification by reducing both failed and unnecessary colonoscopy referral.

CONCLUSIONS

An automated tool for polyp measurement in patients facilitates patient risk stratification.

Protrusion method for automated estimation of polyp size on CT colonography

OBJECTIVE

The purpose of this study was to assess the accuracy and measurement variability of automated lesion measurement on CT colonography in comparison with manual 2D and 3D techniques under varying scanning conditions.

MATERIALS AND METHODS

The study included phantoms (23 phantom objects) and patients (16 polyps). Measurement with sliding calipers served as the reference for the phantom data. The mean of two independent colonoscopic measurements was the reference for the polyps. The automated measurement was developed for a computer-aided detection scheme, and the size of any detected object was obtained from measurement of its largest diameter. The automated measurement was compared with manual 2D and 3D measurements by two experienced observers.

RESULTS

For phantom data, the measurement variability of the automated method was significantly less than that of the two observers (p < 0.05), except for the 3D measurement by observer 1, as follows: automated, 0.86 mm; observer 1, 1.76 mm (2D), 0.96 (3D); observer 2, 1.34 mm (2D), 1.45 mm (3D). The variability of the automated method did not differ significantly from that of manual methods in measurement with patient data. The automated method had a systematic error for phantom data (1.9 mm).

CONCLUSION

For phantoms, the automated method has less measurement variability than manual 2D and 3D techniques. For true polyps, the measurement variability of the automated method is comparable with that of manual methods. The automated method does not suffer from intraobserver variability. Because systematic error can be calibrated, automated size measurement may contribute to a practical evaluation strategy.

Computer-aided detection for CT colonography: update 2007

Computed tomographic colonography (CTC) is an emerging technique for polyp detection in the colon. However, lesion detection can be challenging due to insufficient patient preparation, chosen CT technique or reader imperfection. The primary goal of computer-aided detection (CAD) for CTC is locating possible polyps, and presenting the reader with these polyp candidates. Other goals are sensitivity improvement and reduction of reading time and inter-observer variability. The multistep CAD procedure typically consists of segmentation of the colonic wall (e.g. region growing); selection of intermediate polyp candidates (curvature analysis, sphere fitting, normal analysis, slope density function ...); classification of final candidates for detection and listing suspicious polyps (location, size and volume). Remaining task for the radiologist is the validation or rejection of the polyp candidates. State-of-the-art CAD systems should require minimal or even no user interaction for the extraction of the colonic wall, offer a computation time less than 10-20 min and high sensitivity and specificity for different polyp sizes and shapes, with a low number of false positives. These systems have the potential to increase radiologist's performance and to decrease inter-reader variability. Besides CAD key techniques we also discuss new developments in CAD and describe recent applications facilitating CTC.

Quantitative assessment of colorectal cancer tumor vascular parameters by using perfusion CT: influence of tumor region of interest

PURPOSE

To prospectively determine whether position and size of tumor region of interest (ROI) influence estimates of colorectal cancer vascular parameters at computed tomography (CT).

MATERIALS AND METHODS

After institutional review board approval and informed consent, 25 men and 22 women (mean age, 65.8 years) with colorectal adenocarcinoma underwent 65-second CT perfusion study. Blood volume, blood flow, and permeability-surface area product were determined for 40- or 120-mm(2) circular ROIs placed at the tumor edge and center and around (outlining) visible tumor. ROI analysis was repeated by two observers in different subsets of patients to assess intra- and interobserver variation. Measurements were compared by using analysis of variance; a difference with P = .002 was significant.

RESULTS

Blood volume, blood flow, and permeability-surface area product measurements were substantially higher at the edge than at the center for both 40- and 120-mm(2) ROIs. For 40-mm(2) ROI, means of the three measurements were 6.9 mL/100 g (standard deviation [SD], 1.4), 108.7 mL/100 g per minute (SD, 39.2), and 16.9 mL/100 g per minute (SD, 4.2), respectively, at the edge versus 5.1 mL/100 g (SD, 1.5), 56.3 mL/100 g per minute (SD, 33.1), and 13.9 mL/100 g per minute (SD, 4.6), respectively, at the center. For 120-mm(2) ROI, means of the three measurements were 6.6 mL/100 g (SD, 1.3), 96.7 mL/100 g per minute (SD, 42.5), and 16.3 mL/100 g per minute (SD, 5.6), respectively, at the edge versus 5.1 mL/100 g (SD, 1.4), 58.3 mL/100 g per minute (SD, 32.5), and 13.4 mL/100 g per minute (SD, 4.3) at the center (P < .0001). Measurements varied substantially depending on the ROI size; values for the ROI for outlined tumor were intermediate between those at the tumor edge and center. Inter- and intraobserver agreement was poor for both 40- and 120-mm(2) ROIs.

CONCLUSION

Position and size of tumor ROI and observer variation substantially influence ultimate perfusion values. ROI for outlined entire tumor is more reliable for perfusion measurements and more appropriate clinically than use of arbitrarily determined smaller ROIs.

Manual and automated polyp measurement comparison of CT colonography with optical colonoscopy

RATIONALE AND OBJECTIVES

The purpose of this study was to assess (1) the agreement of two-dimensional (2D) and three-dimensional (3D) manual and automated polyp linear diameter measurements at CT colonography (CTC), with optical colonoscopic equivalents and (2) intraobserver and interobserver agreement of the CTC measurements.

MATERIALS AND METHODS

Using the same CTC system, two radiologists independently measured the maximum linear diameter of 44 polyps (reference size 3-15 mm) matched on CTC and optical colonoscopy: manual 2D optimized multiplanar reformatted planes with standard window settings (level 1500 HU, width -200 HU), manual 3D measurement with software calipers and automated 3D measurement with software. After 2 weeks, polyps were measured again. Compatibility of CTC measurement with that of optical colonoscopy and measurement reproducibility was assessed statistically.

RESULTS

In the manual measurement, 44 polyps were analyzed and 41 in automated measurement; three polyps could not be extracted. Although the measurement difference was noted for automated, manual 3D, and manual 2D measurements, statistically supported agreement with optical colonoscopic measurement was noted only with manual 2D measurement for both observers. However, 95% limits of agreement were wide for all the measurement methods. When categorized according to the optical colonoscopic measurement, manual 2D, 3D, and automated measurements showed "good" agreement. Although intraobserver and interobserver agreement was good with manual measurement, intraobserver and interobserver agreement was excellent with automated measurement.

CONCLUSION

Manual 2D measurements demonstrated trends of better approximation to optical colonoscopy measurements than manual 3D or automated measurements. And automated measurement eliminated intraobserver and interobserver variability. For noninvasive CTC surveillance, manual 2D measurements are expected to measure medium-sized polyps with sufficient agreement with optical colonoscopic measurements and excellent intraobserver and interobserver variability, especially if combined with automated measurement.

Linear polyp measurement at CT colonography: 3D endoluminal measurement with optimized surface-rendering threshold value and automated measurement

PURPOSE

To determine the optimal surface-rendering threshold value for three-dimensional (3D) endoluminal computed tomographic (CT) colonographic images for accurate manual polyp measurement, with direct measurement of simulated polyps as the reference standard, and to assess the agreement between manual 3D measurements and automated measurements.

MATERIALS AND METHODS

Institutional review board approval was not required for the experimental study with pig colons obtained at an abattoir but was obtained for the use of patient data, with waiver of informed consent. Eighty-six simulated polyps (reference size, 3-15 mm) and 14 human polyps (approximate size, 5-20 mm) were included. Automated polyp measurements and manual measurements with endoluminal views that were surface rendered at threshold values of -800, -700, -600, and -500 HU were performed by one observer. Agreement between CT colonographic measurements and reference sizes and between manual and automated measurements were assessed by using the Bland-Altman method.

RESULTS

For simulated polyps, mean measurement difference between the observed size and reference size was 0.86 mm (95% limits of agreement: -0.52 mm, 2.24 mm), 0.55 mm (95% limits of agreement: -0.75 mm, 1.85 mm), 0.20 mm (95% limits of agreement: -1.11 mm, 1.50 mm), and -0.08 mm (95% limits of agreement: -1.43 mm, 1.27 mm) for -800, -700, -600, and -500 HU, respectively. Mean measurement difference was 0.09 mm (95% limits of agreement: -1.49 mm, 1.67 mm) for automated measurement. Manual polyp size at -500 HU (P = .277) and automated polyp size (P = .288) were not significantly different from reference size. For human polyps, 10 polyps, excluding four lesions that were large, lobulated, or located adjacent to an edge of the haustral fold, showed accurate automated demarcation of lesion boundaries. Automated measurements of the 10 polyps showed the closest agreement with manual measurements at -500 HU.

CONCLUSION

The optimal surface-rendering threshold value for accurate polyp measurement is approximately -500 HU. Automated measurements agree closely with manual measurements at the optimal threshold value for well-circumscribed smooth rounded polyps.

Automated polyp measurement with CT colonography: preliminary observations in a phantom colon model

OBJECTIVE

The purpose of this study was to evaluate the accuracy and precision of polyp measurements obtained with an automated tool in a colon phantom containing polyps of multiple sizes, morphologic types, and locations.

MATERIALS AND METHODS

A colon phantom was scanned at 12, 25, 50, and 100 mA with standard CT colonographic acquisition parameters. Four reviewers using manual 2D methods and an automated polyp measurement tool measured 24 polyps of varying sizes and morphologic types, some at a haustral fold tip and some not at a fold tip. The accuracy (difference from true value) of manual and automated methods was compared across polyp sizes, morphologic types, locations, and doses. Precision (closeness of different measures) was compared for intraobserver and interobserver measurements.

RESULTS

The accuracy of automated polyp measurement was dependent on morphologic type (p < or = 0.02), size (for three of four reviewers, p < or = 0.05), and location of polyps with respect to haustral folds (two of four reviewers, p < or = 0.01). For two of four reviewers, automated measures were less accurate for 5-mm polyps, flat polyps, and polyps at the tips of folds (p < or = 0.04). Intraobserver precision was high, two automated measurements being within 0.1 mm of each other 82-93% of the time. Interobserver precision values for automated measures were more similar 85% of the time (82/96; p < 0.001).

CONCLUSION

Accuracy of automated polyp measurements depends on polyp size, morphologic type, and location. When using an automated tool, radiologists should visually inspect automated polyp measurements, particularly for small and flat polyps and those located on folds, because manual measurements may be more accurate in this setting. Automated polyp measurements are more precise than manual measurements.

CAD in CT colonography without and with oral contrast agents: progress and challenges

Computed tomographic colonography (CTC), also known as virtual colonoscopy, is an emerging alternative technique for screening of colon cancers. CTC uses CT to provide a series of cross-sectional images of the colon for detection of polyps and masses. Fecal tagging is a means of labeling of residual feces by an oral contrast agent for improving the accuracy in the detection of polyps. Computer-aided diagnosis (CAD) for CTC automatically determines the locations of suspicious polyps and masses in CTC and presents them to radiologists, typically as a second opinion. Despite its relatively short history, CAD has become one of the mainstream techniques that could make CTC prime time for screening of colorectal cancer. Rapid technical developments have advanced CAD substantially during the last several years, and a fundamental scheme for the detection of polyps has been established, in which sophisticated 3D image processing, analysis, and display techniques play a pivotal role. The latest CAD systems indicate a clinically acceptable high sensitivity and a low false-positive rate, and observer studies have demonstrated the benefits of these systems in improving radiologists' detection performance. Some technical and clinical challenges, however, remain unresolved before CAD can become a truly useful tool for clinical practice. Also, new challenges are facing CAD as the methods for bowel preparation and image acquisition, such as tagging of fecal residue with oral contrast agents, and interpretation of CTC images evolve. This article reviews the current status and future challenges in CAD for CTC without and with fecal tagging.

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.

Polyps: linear and volumetric measurement at CT colonography

PURPOSE

To retrospectively determine which of several computed tomographic (CT) colonography-based polyp measurements is most compatible with the linear measurement at optical colonoscopy and which is best for assessing change in polyp size.

MATERIALS AND METHODS

This HIPAA-compliant study had institutional review board approval; informed consent was obtained. Prone and supine CT colonography with same-day optical colonoscopy was performed in 216 patients (147 men and 69 women; age range, 46-79 years; mean age, 59.2 years) with 338 polyps detected at CT colonography. Polyp size was measured with three linear measurements and two volume measurements. One linear measurement and one volume measurement were performed by using automated segmentation; remaining measurements were performed manually. Compatibility with linear size at optical colonoscopy and measurement reproducibility were assessed three ways: variation from size measurement at optical colonoscopy, change between prone and supine scans, and variability between observers. Confidence analysis assessed the ability of each measurement to identify polyps with an optical colonoscopy measurement of 1 cm or greater.

RESULTS

Two hundred fifty-one segmentable polyps were present on both supine and prone scans. Linear polyp diameter manually measured on a three-dimensional endoluminally viewed surface (L(M3D)) indicated with 95% confidence that a polyp measured as 0.8 cm or smaller was less than 1.0 cm at optical colonoscopy. Prone and supine polyp size difference was smallest for L(M3D) and the linear diameter computed from manual and automated volume measurements, with interquartile ranges smaller than or equal to 0.3, 0.2, and 0.5 cm, respectively. Interobserver and intraobserver variability was smallest for linear polyp diameter measurements on a two-dimensional display, with a mean percentage difference of 2.8% (95% Bland-Altman limits of agreement: -17.8%, 23.4%) and 5.0% (95% Bland-Altman limits of agreement: -28.3%, 38.3%), respectively.

CONCLUSION

L(M3D) best approximated polyp size measurements at optical colonoscopy. Linear diameter calculated from automated volume measurements showed the smallest variation between supine and prone scans while avoiding observer variability and may be best for assessing polyp size changes with serial examinations.

Polyp Measurement with CT Colonography: Multiple-Reader, Multiple-Workstation Comparison

OBJECTIVE

The risk of invasive colorectal cancer in colorectal polyps correlates with lesion size. Our purpose was to define the most accurate methods for measuring polyp size at CT colonography (CTC) using three models of workstations and multiple observers.

MATERIALS AND METHODS

Six reviewers measured 24 unique polyps of known size (5, 7, 10, and 12 mm), shape (sessile, flat, and pedunculated), and location (straight or curved bowel segment) using CTC data sets obtained at two doses (5 mAs and 65 mAs) and a previously described colonic phantom model. Reviewers measured the largest diameter of polyps on three proprietary workstations. Each polyp was measured with lung and soft-tissue windows on axial, 2D multiplanar reconstruction (MPR), and 3D images.

RESULTS

There were significant differences among measurements obtained at various settings within each workstation (p < 0.0001). Measurements on 2D images were more accurate with lung window than with soft-tissue window settings (p < 0.0001). For the 65-mAs data set, the most accurate measurements were obtained in analysis of axial images with lung window, 2D MPR images with lung window, and 3D tissue cube images for Wizard, Advantage, and Vitrea workstations, respectively, without significant differences in accuracy among techniques (0.11 < p < 0.59). The mean absolute error values for these optimal settings were 0.48 mm, 0.61 mm, and 0.76 mm, respectively, for the three workstations. Within the ultralow-dose 5-mAs data set the best methods for Wizard, Advantage, and Vitrea were axial with lung window, 2D MPR with lung window, and 2D MPR with lung window, respectively. Use of nearly all measurement methods, except for the Vitrea 3D tissue cube and the Wizard 2D MPR with lung window, resulted in undermeasurement of the true size of the polyps.

CONCLUSION

Use of CTC computer workstations facilitates accurate polyp measurement. For routine CTC examinations, polyps should be measured with lung window settings on 2D axial or MPR images (Wizard and Advantage) or 3D images (Vitrea). When these optimal methods are used, these three commercial workstations do not differ significantly in acquisition of accurate polyp measurements at routine dose settings.

CT colonography: automated measurement of colonic polyps compared with manual techniques--human in vitro study

PURPOSE

To prospectively investigate the relative accuracy and reproducibility of manual and automated computer software measurements by using polyps of known size in a human colectomy specimen.

MATERIALS AND METHODS

Institutional review board approval was obtained for the study; written consent for use of the surgical specimen was obtained. A colectomy specimen containing 27 polyps from a 16-year-old male patient with familial adenomatous polyposis was insufflated, submerged in a container with solution, and scanned at four-section multi-detector row computed tomography (CT). A histopathologist measured the maximum dimension of all polyps in the opened specimen. Digital photographs and line drawings were produced to aid CT-histologic measurement correlation. A novice (radiographic technician) and an experienced (radiologist) observer independently estimated polyp diameter with three methods: manual two-dimensional (2D) and manual three-dimensional (3D) measurement with software calipers and automated measurement with software (automatic). Data were analyzed with paired t tests and Bland-Altman limits of agreement.

RESULTS

Seven polyps (<or=6-mm diameter) could not be extracted by using the software; 20 polyps (5-15-mm diameter) remained for analysis. Automated measurement was not significantly different from histologic size for the experienced reader (mean difference, 0.63 mm; P=.06) or novice reader (mean difference, 0.58 mm; P=.12). With manual 2D measurement and manual 3D measurement, the experienced reader (1.21-mm mean difference, P<.001, and 0.68-mm mean difference, P=.03, respectively) and novice reader (1.54-mm mean difference, P<.001, and 0.84-mm mean difference, P=.002, respectively) significantly underestimated polyp size. Interobserver agreement was good and similar for all three methods (95% limits of agreement span, approximately 2.5 mm). Intraobserver agreement was related to reader experience, with differences of up to 2.5 mm within expected limits of agreement.

CONCLUSION

For polyps smaller than 1 cm, measurement differences of up to 2.5 mm are within the expected limits of inter- and intraobserver agreement for all measurement techniques. Automated and manual 3D polyp measurements are more accurate than manual 2D measurements.

CT colonography: effect of colonic distension on polyp measurement accuracy and agreement-in vitro study

RATIONAL AND OBJECTIVES

To investigate the effect of colonic distension on polyp measurement accuracy and reader agreement.

MATERIALS AND METHODS

Institutional review board permission was obtained. A sealed colectomy specimen from a patient with familial adenomatous polyposis was scanned using a four-detector-row computed tomography (CT) after half and full air distension. A histopathologist measured the maximum dimension of all polyps in the opened specimen. Digital photographs and line drawings were used to individually match polyps visible in the CT datasets. Two observers (radiologist, technician) independently estimated the maximum polyp diameter using both two-dimensional (2D) and three-dimensional (3D) surface rendering. Full-distension measurements were repeated 1 week later. Accuracy was analyzed using paired t-test. Observer agreement was assessed using Bland Altman limits of agreement.

RESULTS

Twenty-three polyps (4-15 mm) were identified. 2D measurements were significantly smaller than histologic size at both half distension (radiologist first): mean difference [md] -1.1 mm, md -1.7 mm, and full distension md -1.1 mm, md 1.4 mm (all P < .001). 3D measurements were not significantly different from true size other than after half distension for the technician (md -0.7 mm, P = .01). 95% Bland Altman limits for interobserver agreement were narrower after full distension, and better using 2D (half-distension span of agreement approximately 4.7 mm and 6 mm for 2D and 3D, respectively). 2D intraobserver span of agreement between half and full distension was approximately 3.8 mm and 3.2 mm for the radiologist and technician, respectively, compared with 6.2 mm and 5.5 mm using 3D.

CONCLUSION

3D polyp measurement is more accurate than 2D. However, in the presence of suboptimal distension, inter- and intraobserver agreement is superior using 2D.