Computer-aided polyp detection on CT colonography: Comparison of three systems in a high-risk human population

A comparison of computer-assisted detection (CAD) programs for the identification of colorectal polyps: performance and sensitivity analysis, current limitations and practical tips for radiologists

AIM

To directly compare the accuracy and speed of analysis of two commercially available computer-assisted detection (CAD) programs in detecting colorectal polyps.

MATERIALS AND METHOD

In this retrospective single-centre study, patients who had colorectal polyps identified on computed tomography colonography (CTC) and subsequent lower gastrointestinal endoscopy, were analysed using two commercially available CAD programs (CAD1 and CAD2). Results were compared against endoscopy to ascertain sensitivity and positive predictive value (PPV) for colorectal polyps. Time taken for CAD analysis was also calculated.

RESULTS

CAD1 demonstrated a sensitivity of 89.8%, PPV of 17.6% and mean analysis time of 125.8 seconds. CAD2 demonstrated a sensitivity of 75.5%, PPV of 44.0% and mean analysis time of 84.6 seconds.

CONCLUSION

The sensitivity and PPV for colorectal polyps and CAD analysis times can vary widely between current commercially available CAD programs. There is still room for improvement. Generally, there is a trade-off between sensitivity and PPV, and so further developments should aim to optimise both. Information on these factors should be made routinely available, so that an informed choice on their use can be made. This information could also potentially influence the radiologist's use of CAD results.

CT colonography with computer-aided detection: recognizing the causes of false-positive reader results

Computed tomography (CT) colonography is a screening modality used to detect colonic polyps before they progress to colorectal cancer. Computer-aided detection (CAD) is designed to decrease errors of detection by finding and displaying polyp candidates for evaluation by the reader. CT colonography CAD false-positive results are common and have numerous causes. The relative frequency of CAD false-positive results and their effect on reader performance on the basis of a 19-reader, 100-case trial shows that the vast majority of CAD false-positive results were dismissed by readers. Many CAD false-positive results are easily disregarded, including those that result from coarse mucosa, reconstruction, peristalsis, motion, streak artifacts, diverticulum, rectal tubes, and lipomas. CAD false-positive results caused by haustral folds, extracolonic candidates, diminutive lesions (<6 mm), anal papillae, internal hemorrhoids, varices, extrinsic compression, and flexural pseudotumors are almost always recognized and disregarded. The ileocecal valve and tagged stool are common sources of CAD false-positive results associated with reader false-positive results. Nondismissable CAD soft-tissue polyp candidates larger than 6 mm are another common cause of reader false-positive results that may lead to further evaluation with follow-up CT colonography or optical colonoscopy. Strategies for correctly evaluating CAD polyp candidates are important to avoid pitfalls from common sources of CAD false-positive results.

Ultra-low peak voltage CT colonography: effect of iterative reconstruction algorithms on performance of radiologists who use anthropomorphic colonic phantoms

PURPOSE

To analyze the effect of a decrease in computed tomographic (CT) colonographic voltage, from 100 and 120 kVp to 80 kVp and reconstructed with filtered back projection ( FBP filtered back projection ), on radiation dose, image noise, and diagnostic performance in anthropomorphic phantoms and to assess the effect of iterative reconstruction ( IR iterative reconstruction ) algorithms on radiologists' performance for 80-kVp CT colonography.

MATERIALS AND METHODS

Seven colon phantoms with 68 simulated polyps (≥6 mm) were scanned at three peak voltage settings (80, 100, 120 kVp) and 10 mAs. Images were reconstructed by using FBP filtered back projection , hybrid statistic-based IR iterative reconstruction , and knowledge-based IR iterative reconstruction algorithms. Effective radiation dose, image noise, and per-polyp sensitivity were recorded and compared by two reviewers with Friedman test, repeated measures analysis of variance, and McNemar test.

RESULTS

Median size-specific dose estimate and effective radiation dose of 80-kVp CT colonography was 0.231 mGy and 0.167 mSv, respectively, which was lower than with 100- and 120-kVp CT colonography, with significant difference between 80 and 120 kVp (P = .0005). Image noise (202.0 HU) at 80-kVp FBP filtered back projection CT colonography was significantly higher than at 100-kVp FBP filtered back projection (139.1 HU) and 120-kVp FBP filtered back projection (120.4 HU) (P < .0001). Per-polyp sensitivity (reviewer 1, 14.7% [10 of 68]; reviewer 2, 7.4% [five of 68]) at 80-kVp FBP filtered back projection was significantly lower than at 100-kVp FBP filtered back projection (reviewer 1, 57.4% [39 of 68]; reviewer 2, 39.7% [27 of 68]) and 120-kVp FBP filtered back projection (reviewer 1, 85.3% [58 of 68]; reviewer 2, 83.8% [57 of 68]) (P < .0001). With statistic-based IR iterative reconstruction , image noise at 80 kVp decreased significantly (52.8% [106.7 HU of 202.0 HU]) compared with that at 80-kVp FBP filtered back projection (P < .0001), but per-polyp sensitivity (reviewer 1, 79.4% [54 of 68]; reviewer 2, 66.2% [45 of 68]) at 80-kVp statistic-based IR iterative reconstruction remained significantly lower than at 100-kVp statistic-based IR iterative reconstruction (reviewer 1, 95.6% [65 of 68]; reviewer 2, 86.8% [59 of 68]) (P = .001) and 120-kVp statistic-based IR iterative reconstruction (reviewer 1, 98.5% [67 of 68]; reviewer 2, 89.7% [61 of 68]) (P < .001). For knowledge-based IR iterative reconstruction , per-polyp sensitivity at 80 kVp was improved to 98.5% (67 of 68) and 94.1% (64 of 68), not significantly different from that at 100 kVp (reviewer 1, 100% [68 of 68]; reviewer 2, 95.6% [65 of 68]) and 120 kVp (reviewer 1, 100% [68 of 68]; reviewer 2, 95.6% [65 of 68]) (P > .999).

CONCLUSION

A decrease in tube voltage to 80 kVp caused reduction in radiation dose (0.166 mSv) with deterioration in image noise and per-polyp sensitivity. By using a knowledge-based IR iterative reconstruction algorithm, radiologists' performance of 80-kVp CT colonography was acceptable and on par with that at 100- or 120-kVp CT colonography.

Electronic cleansing for CT colonography: does it help CAD software performance in a high-risk population for colorectal cancer?

OBJECTIVE

To compare the performance of computer-aided detection (CAD) for CT colonography (CTC) with and without electronic cleansing (EC) in a high-risk population tagged with a faecal tagging (FT) protocol.

METHODS

Thirty-two patients underwent CTC followed by same-day colonoscopy. All patients underwent bowel preparation and FT with barium and gastrografin. Each CTC dataset was processed with colon CAD with and without EC. Per-polyp sensitivity was calculated. The average number of false-positive (FP) results and their causes were also analysed and compared.

RESULTS

Eighty-six polyps were detected in 29 patients. Per-polyp sensitivities of CAD with EC (93.8% and 100%) were higher than those without EC (84.4% and 87.5%) for polyps >or=6 mm and >or=10 mm, respectively. However, the differences were not significant. The average number (6.3) of FPs of CAD with EC was significantly larger than that (3.1) without EC. The distribution of FPs in both CAD settings was also significantly different. The most common cause of FPs was the ileocaecal valve in both datasets. However, untagged faeces was a significantly less common cause of FPs with EC, EC-related artefacts being more common.

CONCLUSION

Electronic cleansing has the potential to improve per-polyp sensitivity of CTC CAD, although the significantly larger number of FPs with EC remains to be improved.