Conference Paper

Massive-training artificial neural networks for CAD for detection of polyps in CT colonography: False-negative cases in a large multicenter clinical trial

Dept. of Radiol., Univ. of Chicago, Chicago, IL
DOI: 10.1109/ISBI.2008.4541088 Conference: Proceedings of the 2008 IEEE International Symposium on Biomedical Imaging: From Nano to Macro, Paris, France, May 14-17, 2008
Source: DBLP


A major challenge in computer-aided detection (CAD) of polyps in CT colonography (CTC) is the detection of "difficult" polyps which radiologists are likely to miss. Our purpose was to develop massive-training artificial neural networks (MTANNs) for improving the performance of a CAD scheme on false-negative cases in a large multicenter clinical trial. We developed 3D MTANNs designed to differentiate between polyps and several types of non- polyps and tested on 14 polyps/masses that were actually "missed" by radiologists in the trial. Our initial CAD scheme detected 71.4% of "missed" polyps with 18.9 false positives (FPs) per case. The MTANNs removed 75% of the FPs without loss of any true positives; thus, the performance of our CAD scheme was improved to 4.8 FPs per case at the sensitivity of 71.4% of the polyps "missed" by radiologists.

Download full-text


Available from: Kenji Suzuki
  • Source
    • "A hierarchical probabilistic supervised classification is employed in Lu et al. (2008) where polar coordinate alignment of the polyp is performed and boosting is used for enhanced polyp surface boundary detection. By using various schemes of massive training artificial neural networks, Suzuki et al. (2008) improved the detection rate of polyps otherwise missed by standard methods by as much as 71.4% and removed 75% of otherwise original false positives at a new improved 4.8 false positives per dataset. A complete automatic polyp detection procedure is presented in Hong et al. (2006). "
    [Show abstract] [Hide abstract]
    ABSTRACT: This chapter presents a comprehensive scheme for automated detection of colorectal polyps in computed tomography colonography (CTC) with particular emphasis on robust learning that algorithms differ - entiate polyps from non-polyp shapes. The authors’ automated CTC scheme introduces two orientation independent features which encode the shape characteristics that aid in classification of polyps and non-polyps with high accuracy, low false positive rate, and low computations making the scheme suit-able for colorectal cancer screening initiatives. Experiments using state-of-the-art machine learning algorithms viz., lazy learning, support vector machines, and naïve Bayes classifiers reveal the robustness of the two features in detecting polyps at 100% sensitivity for polyps with diameter greater than 10 mm while attaining total low false positive rates, respectively, of 3.05, 3.47 and 0.71 per CTC dataset at specificities above 99% when tested on 58 CTC datasets. The results were validated using colonoscopy reports provided by expert radiologists.
    Full-text · Chapter · Jan 2009
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: A major challenge in the current computer-aided detection (CAD) of polyps in CT colonography (CTC) is to reduce the number of false-positive (FP) detections while maintaining a high sensitivity level. A pattern-recognition technique based on the use of an artificial neural network (ANN) as a filter, which is called a massive-training ANN (MTANN), has been developed recently for this purpose. The MTANN is trained with a massive number of subvolumes extracted from input volumes together with the teaching volumes containing the distribution for the "likelihood of being a polyp;" hence the term "massive training." Because of the large number of subvolumes and the high dimensionality of voxels in each input subvolume, the training of an MTANN is time-consuming. In order to solve this time issue and make an MTANN work more efficiently, we propose here a dimension reduction method for an MTANN by using Laplacian eigenfunctions (LAPs), denoted as LAP-MTANN. Instead of input voxels, the LAP-MTANN uses the dependence structures of input voxels to compute the selected LAPs of the input voxels from each input subvolume and thus reduces the dimensions of the input vector to the MTANN. Our database consisted of 246 CTC datasets obtained from 123 patients, each of whom was scanned in both supine and prone positions. Seventeen patients had 29 polyps, 15 of which were 5-9 mm and 14 were 10-25 mm in size. We divided our database into a training set and a test set. The training set included 10 polyps in 10 patients and 20 negative patients. The test set had 93 patients including 19 polyps in seven patients and 86 negative patients. To investigate the basic properties of a LAP-MTANN, we trained the LAP-MTANN with actual polyps and a single source of FPs, which were rectal tubes. We applied the trained LAP-MTANN to simulated polyps and rectal tubes. The results showed that the performance of LAP-MTANNs with 20 LAPs was advantageous over that of the original MTANN with 171 inputs. To test the feasibility of the LAP-MTANN, we compared the LAP-MTANN with the original MTANN in the distinction between actual polyps and various types of FPs. The original MTANN yielded a 95% (18/19) by-polyp sensitivity at an FP rate of 3.6 (338/93) per patient, whereas the LAP-MTANN achieved a comparable performance, i.e., an FP rate of 3.9 (367/93) per patient at the same sensitivity level. With the use of the dimension reduction architecture, the time required for training was reduced from 38 h to 4 h. The classification performance in terms of the area under the receiver-operating-characteristic curve of the LAP-MTANN (0.84) was slightly higher than that of the original MTANN (0.82) with no statistically significant difference (p-value =0.48).
    Full-text · Article · Nov 2010
  • [Show abstract] [Hide abstract]
    ABSTRACT: We present an experimental comparison of 3D feature descriptors with application to threat detection in Computed Tomography (CT) airport baggage imagery. The detectors range in complexity from a basic local density descriptor, through local region histograms and three-dimensional (3D) extensions to both to the RIFT descriptor and the seminal SIFT feature descriptor. We show that, in the complex CT imagery domain containing a high degree of noise and imaging artefacts, a specific instance object recognition system using simpler descriptors appears to outperform a more complex RIFT/SIFT solution. Recognition rates in excess of 95% are demonstrated with minimal false-positive rates for a set of exemplar 3D objects.
    No preview · Article · Sep 2013 · Pattern Recognition