Multi-View Network for Colorectal Polyps Detection in CT Colonography

Standards for Gastroenterologists for Performing and Interpreting Diagnostic Computed Tomographic Colonography

Progress in Refining Virtual Colonoscopy for Colorectal Cancer Screening

Surface curvatures are important geometric features for the computer-aided analysis and detection of polyps in CT colonography (CTC). However, the general kernel approach for curvature computation can yield erroneous results for small polyps and for polyps that lie on haustral folds. Those erroneous curvatures will reduce the performance of polyp detection. This paper presents an analysis of interpolation's effect on curvature estimation for thin structures and its application on computer-aided detection of small polyps in CTC. The authors demonstrated that a simple technique, image interpolation, can improve the accuracy of curvature estimation for thin structures and thus significantly improve the sensitivity of small polyp detection in CTC. Our experiments showed that the merits of interpolating included more accurate curvature values for simulated data, and isolation of polyps near folds for clinical data. After testing on a large clinical data set, it was observed that sensitivities with linear, quadratic B-spline and cubic B-spline interpolations significantly improved the sensitivity for small polyp detection. The image interpolation can improve the accuracy of curvature estimation for thin structures and thus improve the computer-aided detection of small polyps in CTC.

Robust training of a deep convolutional neural network (DCNN) requires a very large number of annotated datasets that are currently not available in CT colonography (CTC). We previously demonstrated that deep transfer learning provides an effective approach for robust application of a DCNN in CTC. However, at high detection accuracy, the differentiation of small polyps from non-polyps was still challenging. In this study, we developed and evaluated a deep ensemble learning (DEL) scheme for reviewing of virtual endoluminal images to improve the performance of computer-aided detection (CADe) of polyps in CTC. Nine different types of image renderings were generated from virtual endoluminal images of polyp candidates detected by a conventional CADe system. Eleven DCNNs that represented three types of publically available pre-trained DCNN models were re-trained by transfer learning to identify polyps from the virtual endoluminal images. A DEL scheme that determines the final detected polyps by a review of the nine types of VE images was developed by combining the DCNNs using a random forest classifier as a meta-classifier. For evaluation, we sampled 154 CTC cases from a large CTC screening trial and divided the cases randomly into a training dataset and a test dataset. At 3.9 falsepositive (FP) detections per patient on average, the detection sensitivities of the conventional CADe system, the highestperforming single DCNN, and the DEL scheme were 81.3%, 90.7%, and 93.5%, respectively, for polyps ≥6 mm in size. For small polyps, the DEL scheme reduced the number of false positives by up to 83% over that of using a single DCNN alone. These preliminary results indicate that the DEL scheme provides an effective approach for improving the polyp detection performance of CADe in CTC, especially for small polyps.

CT colonography (CTC) is a feasible and minimally invasive method for the detection of colorectal polyps and cancer screening. Computer-aided detection (CAD) of polyps has improved consistency and sensitivity of virtual colonoscopy interpretation and reduced interpretation burden. A CAD system typically consists of four stages: (1) image preprocessing including colon segmentation; (2) initial detection generation; (3) feature selection; and (4) detection classification. In our experience, three existing problems limit the performance of our current CAD system. First, highdensity orally administered contrast agents in fecal-tagging CTC have scatter effects on neighboring tissues. The scattering manifests itself as an artificial elevation in the observed CT attenuation values of the neighboring tissues. This pseudo-enhancement phenomenon presents a problem for the application of computer-aided polyp detection, especially when polyps are submerged in the contrast agents. Second, general kernel approach for surface curvature computation in the second stage of our CAD system could yield erroneous results for thin structures such as small (6-9 mm) polyps and for touching structures such as polyps that lie on haustral folds. Those erroneous curvatures will reduce the sensitivity of polyp detection. The third problem is that more than 150 features are selected from each polyp candidate in the third stage of our CAD system. These high dimensional features make it difficult to learn a good decision boundary for detection classification and reduce the accuracy of predictions. Therefore, an improved CAD system for polyp detection in CTC data is proposed by introducing three new techniques. First, a scale-based scatter correction algorithm is applied to reduce pseudo-enhancement effects in the image pre-processing stage. Second, a cubic spline interpolation method is utilized to accurately estimate curvatures for initial detection generation. Third, a new dimensionality reduction classifier, diffusion map and local linear embedding (DMLLE), is developed for classification and false positives (FP) reduction. Performance of the improved CAD system is evaluated and compared with our existing CAD system (without applying those techniques) using CT scans of 1186 patients. These scans are divided into a training set and a test set. The sensitivity of the improved CAD system increased 18% on training data at a rate of 5 FPs per patient and 15% on test data at a rate of 5 FPs per patient. Our results indicated that the improved CAD system achieved significantly better performance on medium-sized colonic adenomas with higher sensitivity and lower FP rate in CTC.

Although the screening of small, flat polyps is clinically important, the role of CT colonography (CTC) screening in their detection has not been thoroughly investigated. To evaluate the detection capability and usefulness of CTC in the screening of flat and polypoid lesions by comparing CTC with optic colonoscopy findings as the gold standard. We evaluated the CTC detection capability for flat colorectal polyps with a flat surface and a height not exceeding 3 mm (n = 42) by comparing to conventional polypoid lesions (n = 418) according to the polyp diameter. Four types of reconstruction images including multiplanar reconstruction, volume rendering, virtual gross pathology, and virtual endoscopic images were used for visual analysis. We compared the abilities of the four reconstructions for polyp visualization. Detection sensitivity for flat polyps was 31.3%, 44.4%, and 87.5% for lesions measuring 2-3 mm, 4-5 mm, and ≥6 mm, respectively; the corresponding sensitivity for polypoid lesions was 47.6%, 79.0%, and 91.7%. The overall sensitivity for flat lesions (47.6%) was significantly lower than polypoid lesions (64.1%). Virtual endoscopic imaging showed best visualization among the four reconstructions. Colon cancers were detected in eight patients by optic colonoscopy, and CTC detected colon cancers in all eight patients. CTC using 64-row multidetector CT is useful for colon cancer screening to detect colorectal polyps while the detection of small, flat lesions is still challenging.

CT colonography (CTC) is a non-invasive screening technique for the detection of colorectal polyps, as an alternative to optical colonoscopy in clinical practice. Computer-aided detection (CAD) for CTC refers to a scheme which automatically detects colorectal polyps and masses in CT images of the colon. It has the potential to increase radiologists’ detection performance and greatly shorten the detection time. Over the years, technical developments have advanced CAD for CTC substantially. In this paper, key techniques used in CAD for polyp detection are reviewed. Illustrations about the performance of existing CAD schemes show their relatively high sensitivity and low false positive rate. However, these CAD schemes are still suffering from technical or clinical problems. Some existing challenges faced by CAD are also pointed out at the end of this paper.

A fast and robust colorectal polyp detection framework in CT colonography was proposed. In order to speed the detection of polyp in CT colonography, a cascade-Adaboost framework was employed, and a lot of candidates were rejected quickly in the first stages of the cascade framework. To improve the performance of cascade-Adaboost, cascade indifference curve was explored to determine detection rate and false positive rate of cascade automatically. The experiments showed that the classifier could achieve an overall per-polyp sensitivity of 90% (for polyps' diameter 5 mm and greater), with false positives of 6 per volume on average.

ObjectiveTo examine use of CT colonography (CTC) in the English Bowel Cancer Screening Programme (BCSP) and investigate detection rates.DesignRetrospective analysis of routinely coded BCSP data. Guaiac faecal occult blood test...

CT colonography (CTC) is a feasible and minimally invasive method for the detection of colorectal polyps and cancer screening. Computer-aided detection (CAD) of polyps can improve consistency and sensitivity of virtual colonoscopy interpretation and reduce interpretation burden. However, high-density orally administered contrast agents have scatter effects on neighboring tissues. The scattering manifests itself as an artificial elevation in the observed CT attenuation values of the neighboring tissues. This pseudoenhancement phenomenon presents a problem for the application of computer-aided polyp detection, especially when polyps are submerged in the contrast agents. The authors have developed a scale-based correction method that minimizes scatter effects in CTC data by subtraction of the estimated scatter components from observed CT attenuations. By bringing a locally adaptive structure, object scale, into the correction framework, the region of neighboring tissues affected by contrast agents is automatically specified and adaptively changed in different parts of the image. The method was developed as one preprocessing step in the authors' CAD system and was tested by using leave-one-patient-out evaluation on 56 clinical CTC scans (supine or prone) from 28 patients. There were 50 colonoscopy-confirmed polyps measuring 6-9 mm. Visual evaluation indicated that the method reduced CT attenuation of pseudoenhanced polyps to the usual polyp Hounsfield unit range without affecting luminal air regions. For polyps submerged in contrast agents, the sensitivity of CAD with correction is increased 24% at a rate of ten false-positive detections per scan. For all polyps within 6-9 mm, the sensitivity of the authors' CAD with scatter correction is increased 8% at a rate of ten false-positive detections per scan. The authors' results indicated that CAD with this correction method as a preprocessing step can yield a high sensitivity and a relatively low FP rate in CTC.

Virtual colonoscopy is a new method of imaging the colon in which thin-section, helical computed tomography (CT) is used to generate high-resolution, two-dimensional axial images. Three-dimensional images of the colon simulating those obtained with conventional colonoscopy are then reconstructed off-line. We compared the performance of virtual and conventional colonoscopy for the detection of colorectal polyps. We prospectively studied 100 patients at high risk for colorectal neoplasia (60 men and 40 women; mean age, 62 years). We performed virtual colonoscopy immediately before conventional colonoscopy. We inserted a rectal tube and insufflated the colon with air to the maximal level that the patient could tolerate. We administered 1 mg of glucagon intravenously immediately before CT scanning to minimize the degree of smooth-muscle spasm and peristalsis and to reduce the patient's discomfort. The entire colon was clearly seen by virtual colonoscopy in 87 patients and by conventional colonoscopy in 89. Fifty-one patients had normal findings on conventional colonoscopy. In the other 49, we identified a total of 115 polyps and 3 carcinomas. Virtual colonoscopy identified all 3 cancers, 20 of 22 polyps that were 10 mm or more in diameter (91 percent), 33 of 40 that were 6 to 9 mm (82 percent), and 29 of 53 that were 5 mm or smaller (55 percent). There were 19 false positive findings of polyps and no false positive findings of cancer. Of the 69 adenomatous polyps, 46 of the 51 that were 6 mm or more in diameter (90 percent) and 12 of the 18 that were 5 mm or smaller (67 percent) were correctly identified by virtual colonoscopy. Although discomfort was not specifically recorded, none of the patients requested that virtual colonoscopy be stopped because of discomfort or pain. In a group of patients at high risk for colorectal neoplasia, virtual and conventional colonoscopy had similar efficacy for the detection of polyps that were 6 mm or more in diameter.

Small and Diminutive Polyps: Implications for Colorectal Cancer Screening With Computed Tomography Colonography

We have developed a novel automated technique for segmenting colonic walls for the application of computer-aided polyp detection in CT colonography. In particular, the technique was designed to minimize the presence of extracolonic components, such as small bowel, in the segmented colon. The segmentation technique combines an improved version of our previously reported anatomy-oriented colon segmentation technique with a colon-based analysis step that performs self-adjusting volume-growing within the colonic lumen. Extracolonic components are eliminated by intersecting of the resulting two segmentations, so that the colonic walls remain in the intersection. The technique was evaluated on 88 CT colonography datasets. The colon segmentations were evaluated subjectively by four radiologists, as well as objectively by performance of an automated polyp detection on the segmentation. For comparison, the tests were also performed for the anatomy-oriented colon segmentation technique. On average, the technique covered 98% of the visible colonic walls. Approximately 50% of the extracolonic components remaining in the anatomy-oriented segmentation were removed, but 10-15% of the segmentation still contained extracolonic components. The dataset-based false-positive rate of the automated polyp detection was improved by 10% without compromising the 100% case-based sensitivity, and the case-based false-positive rate was improved by 15% over the previous false-positive rate. The technique segments practically all of the colonic walls in the region of diagnostic quality with a large reduction in the amount of extracolonic components over our previously used technique. The new segmentation improves the specificity of our computer-aided polyp detection scheme significantly without any degradation in detection sensitivity.

Serrated polyps were previously believed to be benign lesions with no cancer potential. However, recent studies have revealed a novel molecular pathway where also serrated polyps can develop into colorectal cancer. CT colonography (CTC) can detect serrated polyps using the radiomic biomarker of contrast coating, but this requires expertise from the reader and current computer-aided detection (CADe) systems have not been designed to detect the contrast coating. The purpose of this study was to develop a novel CADe method that makes use of deep learning to detect serrated polyps based on their contrast-coating biomarker in CTC. In the method, volumetric shape-based features are used to detect polyp sites over soft-tissue and fecal-tagging surfaces of the colon. The detected sites are imaged using multi-angular 2D image patches. A deep convolutional neural network (DCNN) is used to review the image patches for the presence of polyps. The DCNN-based polyp-likelihood estimates are merged into an aggregate likelihood index where highest values indicate the presence of a polyp. For pilot evaluation, the proposed DCNN-CADe method was evaluated with a 10-fold cross-validation scheme using 101 colonoscopy-confirmed cases with 144 biopsy-confirmed serrated polyps from a CTC screening program, where the patients had been prepared for CTC with saline laxative and fecal tagging by barium and iodine-based diatrizoate. The average per-polyp sensitivity for serrated polyps ≥6 mm in size was 93±7% at 0:8±1:8 false positives per patient on average. The detection accuracy was substantially higher that of a conventional CADe system. Our results indicate that serrated polyps can be detected automatically at high accuracy in CTC.

Colon cancer is one of the leading causes of cancer-related deaths, while the CT colonoscopy has become the primary means of early detection of colon cancer. However, the majority of automatic detector of colon polyps in CT colonoscopy was got through offline training, which cannot be updated, when new samples were coming; simultaneously, polyp detection suffers from imbalanced data sets where negative samples (non-polyp) are dominant. Therefore, an online learning asymmetric approach was employed, which not only can update detector, but also can solve the problem of imbalanced data sets. Finally, experimental results show that the proposed algorithm can achieve good classification performance, and a shorter running time.