Tarik A. Chowdhury

A Fully Automatic CAD-CTC System Based on Curvature Analysis for Standard and Low-Dose CT Data

Computed tomography colonography (CTC) is a rapidly evolving noninvasive medical investigation that is viewed by radiologists as a potential screening technique for the detection of colorectal polyps. Due to the technical advances in CT system design, the volume of data required to be processed by radiologists has increased significantly, and as a consequence the manual analysis of this information has become an increasingly time consuming process whose results can be affected by inter- and intrauser variability. The aim of this paper is to detail the implementation of a fully integrated CAD-CTC system that is able to robustly identify the clinically significant polyps in the CT data. The CAD-CTC system described in this paper is a multistage implementation whose main system components are: 1) automatic colon segmentation; 2) candidate surface extraction; 3) feature extraction; and 4) classification. Our CAD-CTC system performs at 100% sensitivity for polyps larger than 10 mm, 92% sensitivity for polyps in the range 5 to 10 mm, and 57.14% sensitivity for polyps smaller than 5 mm with an average of 3.38 false positives per dataset. The developed system has been evaluated on synthetic and real patient CT data acquired with standard and low-dose radiation levels.

The use of 3D surface fitting for robust polyp detection and classification in CT colonography

In this paper we describe the development of a computationally efficient computer-aided detection (CAD) algorithm based on the evaluation of the surface morphology that is employed for the detection of colonic polyps in computed tomography (CT) colonography. Initial polyp candidate voxels were detected using the surface normal intersection values. These candidate voxels were clustered using the normal direction, convexity test, region growing and Gaussian distribution. The local colonic surface was classified as polyp or fold using a feature normalized nearest neighborhood classifier. The main merit of this paper is the methodology applied to select the robust features derived from the colon surface that have a high discriminative power for polyp/fold classification. The devised polyp detection scheme entails a low computational overhead (typically takes 2.20min per dataset) and shows 100% sensitivity for phantom polyps greater than 5mm. It also shows 100% sensitivity for real polyps larger than 10mm and 91.67% sensitivity for polyps between 5 to 10mm with an average of 4.5 false positives per dataset. The experimental data indicates that the proposed CAD polyp detection scheme outperforms other techniques that identify the polyps using features that sample the colon surface curvature especially when applied to low-dose datasets.

A statistical approach for robust polyp detection in CT colonography

In this paper we describe the development of a computationally efficient computer-aided detection (CAD) algorithm based on the statistical features derived from the local colonic surface that are used for the detection of colonic polyps in computed tomography (CT) colonography. The candidate surface voxels were detected and clustered using the surface normal intersection, convexity test, region growing and Hough transform. The main objective of this paper is the selection of the statistical features that optimally capture the convexity of the candidate surface and consequently provide a high discrimination between local surfaces defined by polyps and folds. The developed polyp detection scheme is computationally efficient (typically takes 3.9 minute per dataset) and shows 100% sensitivity for phantom polyps greater than 5 mm and 87.5% sensitivity for real polyps greater than 5 mm with an average of 4.05 false positives per dataset

Shape filtering for false positive reduction at computed tomography colonography

In this paper, we treat the problem of reducing the false positives (FP) in the automatic detection of colorectal polyps at Computer Aided Detection in Computed Tomography Colonography (CAD-CTC) as a shape-filtering task. From the extracted candidate surface, we obtain a reliable shape distribution function and analyse it in the Fourier domain and use the resulting spectral data to classify the candidate surface as belonging to a polyp or a non-polyp class. The developed shape filtering scheme is computationally efficient (takes approximately 2 seconds per dataset to detect the polyps from the colonic surface) and offers robust polyp detection with an overall false positive rate of 5.44 per dataset at a sensitivity of 100% for polyps greater than 10 mm when it was applied to standard and low dose CT data.

Note on Feature Selection for Polyp Detection in CT Colonography

In this paper we describe a computer aided detection (CAD) algorithm for robust detection of polyps in computed tomography (CT) colonography. The devised algorithm identifies suspicious polyp candidate surfaces using the surface normal intersection, Hough transform, 3D histogram analysis, region growing and a convexity test. From these detected surfaces we extract statistical and morphological features in order to evaluate if the surface in question is a polyp or fold. In order to devise the optimal classification scheme the performance of two different classifiers are evaluated when the algorithm is applied to synthetic and real patient data. The experimental results indicate that the overall polyp detection performance shows sensitivity higher than 92% for polyps larger than 5mm with an average of 4.7 to 6.0 false positives per dataset

A Fast and Accurate Method for Automatic Segmentation of Colons at CT Colonography Based on Colon Geometrical Features

In CT colonography, the first major step of colonic polyp detection is reliable segmentation of colon from CT data. In this paper, we propose a fast and accurate method for automatic colon segmentation from CT data using colon geometrical features. After removal of the lung and surrounding air voxels from CT data, labeling is performed to generate candidate regions for Colon segmentation. The centroid of the data, derived from the labeled objects is used to analyze the colon geometry. Other notable features that are used for colon segmentation are volume/length measure and end points. The proposed method was validated using a total of 99 patient datasets. Collapsed colon surface detection was 99.59% with an average of 1.59% extra colonic surface inclusion. The proposed technique takes 16.29 second to segment the colon from an abdomen CT dataset.

A Quantitative Assessment of 3D Facial Key Point Localization Fitting 2D Shape Models to Curvature Information

This work addresses the localization of 11 prominent facial landmarks in 3D by fitting state of the art shape models to 2D data. Quantitative results are provided for 34 scan sat high resolution (texture maps of 10 M-pixels) in terms of accuracy (with respect to manual measurements) and precision(repeatability on different images from the same individual). We obtain an average accuracy of approximately 3 mm, and median repeatability of inter-landmark distances typically below2 mm, which are values comparable to current algorithms on automatic localization of facial landmarks. We also show that, in our experiments, the replacement of texture information by curvature features produced little change in performance, which is an important finding as it suggests the applicability of the method to any type of 3D data.