Deep Learning for Assessing the Corneal Endothelium from Specular Microscopy Images up to 1 Year after Ultrathin-DSAEK Surgery.

Purpose: To present a fully automatic method to estimate the corneal endothelium parameters from specular microscopy images and to use it to study a one-year follow-up after ultrathin Descemet stripping automated endothelial keratoplasty. Methods: We analyzed 383 post ultrathin Descemet stripping automated endothelial keratoplasty images from 41 eyes acquired with a Topcon SP-1P specular microscope at 1, 3, 6, and 12 months after surgery. The estimated parameters were endothelial cell density (ECD), coefficient of variation (CV), and hexagonality (HEX). Manual segmentation was performed in all images. Results: Our method provided an estimate for ECD, CV, and HEX in 98.4% of the images, whereas Topcon's software had a success rate of 71.5% for ECD/CV and 30.5% for HEX. For the images with estimates, the percentage error in our method was 2.5% for ECD, 5.7% for CV, and 5.7% for HEX, whereas Topcon's software provided an error of 7.5% for ECD, 17.5% for CV, and 18.3% for HEX. Our method was significantly better than Topcon's (P < 0.0001) and was not statistically significantly different from the manual assessments (P > 0.05). At month 12, the subjects presented an average ECD = 1377 ± 483 [cells/mm2], CV = 26.1 ± 5.7 [%], and HEX = 58.1 ± 7.1 [%]. Conclusions: The proposed method obtains reliable and accurate estimations even in challenging specular images of pathologic corneas. Translational Relevance: CV and HEX, not currently used in the clinic owing to a lack of reliability in automatic methods, are useful biomarkers to analyze the postoperative healing process. Our accurate estimations allow now for their clinical use.

A pharmacokinetic model including arrival time for two inputs and compensating for varying applied flip-angle in dynamic gadoxetic acid-enhanced MR imaging.

PURPOSE: Pharmacokinetic models facilitate assessment of properties of the micro-vascularization based on DCE-MRI data. However, accurate pharmacokinetic modeling in the liver is challenging since it has two vascular inputs and it is subject to large deformation and displacement due to respiration. METHODS: We propose an improved pharmacokinetic model for the liver that (1) analytically models the arrival-time of the contrast agent for both inputs separately; (2) implicitly compensates for signal fluctuations that can be modeled by varying applied flip-angle e.g. due to B1-inhomogeneity. Orton's AIF model is used to analytically represent the vascular input functions. The inputs are independently embedded into the Sourbron model. B1-inhomogeneity-driven variations of flip-angles are accounted for to justify the voxel's displacement with respect to a pre-contrast image. RESULTS: The new model was shown to yield lower root mean square error (RMSE) after fitting the model to all but a minority of voxels compared to Sourbron's approach. Furthermore, it outperformed this existing model in the majority of voxels according to three model-selection criteria. CONCLUSION: Our work primarily targeted to improve pharmacokinetic modeling for DCE-MRI of the liver. However, other types of pharmacokinetic models may also benefit from our approaches, since the techniques are generally applicable.

Semiautomatic Assessment of the Terminal Ileum and Colon in Patients with Crohn Disease Using MRI (the VIGOR++ Project).

RATIONALE AND OBJECTIVES: The objective of this study was to develop and validate a predictive magnetic resonance imaging (MRI) activity score for ileocolonic Crohn disease activity based on both subjective and semiautomatic MRI features. MATERIALS AND METHODS: An MRI activity score (the "virtual gastrointestinal tract [VIGOR]" score) was developed from 27 validated magnetic resonance enterography datasets, including subjective radiologist observation of mural T2 signal and semiautomatic measurements of bowel wall thickness, excess volume, and dynamic contrast enhancement (initial slope of increase). A second subjective score was developed based on only radiologist observations. For validation, two observers applied both scores and three existing scores to a prospective dataset of 106 patients (59 women, median age 33) with known Crohn disease, using the endoscopic Crohn's Disease Endoscopic Index of Severity (CDEIS) as a reference standard. RESULTS: The VIGOR score (17.1 × initial slope of increase + 0.2 × excess volume + 2.3 × mural T2) and other activity scores all had comparable correlation to the CDEIS scores (observer 1: r = 0.58 and 0.59, and observer 2: r = 0.34-0.40 and 0.43-0.51, respectively). The VIGOR score, however, improved interobserver agreement compared to the other activity scores (intraclass correlation coefficient = 0.81 vs 0.44-0.59). A diagnostic accuracy of 80%-81% was seen for the VIGOR score, similar to the other scores. CONCLUSIONS: The VIGOR score achieves comparable accuracy to conventional MRI activity scores, but with significantly improved reproducibility, favoring its use for disease monitoring and therapy evaluation.

A hybrid optimization strategy for registering images with large local deformations and intensity variations.

PURPOSE: To develop a method for intra-patient registration of pre- and post-contrast abdominal MR images with large local deformations and large intensity variations. METHOD: A hybrid method is proposed to deal with this problem. It consists of two coupled techniques: (1) descriptor matching (DM) at the original resolution using a discrete optimization strategy to avoid getting trapped in a local minimum; (2) continuous optimization to refine the registration outcome based on autocorrelation of local image structure (ALOST). Our method-called DM-ALOST-has become insensitive to the local uptake of contrast agent by exploiting the mean phase and the phase congruency extracted from the multi-scale monogenic signal. The method was extensively tested on abdominal MR data of 30 patients with Crohn's disease. RESULTS: DM-ALOST produced significantly larger mean Dice coefficients than two state-of-the-art methods [Formula: see text]. CONCLUSION: Both qualitative and quantitative tests demonstrated improved registration using the proposed method compared to the state-of-the-art. The DM-ALOST method facilitates measurement of corresponding features from different abdominal MR images, which can aid to assess certain diseases, particularly Crohn's disease.

Estimating the arterial input function from dynamic contrast-enhanced MRI data with compensation for flow enhancement (II): Applications in spine diagnostics and assessment of crohn's disease.

BACKGROUND: Pharmacokinetic (PK) models can describe microvascular density and integrity. An essential component of PK models is the arterial input function (AIF) representing the time-dependent concentration of contrast agent (CA) in the blood plasma supplied to a tissue. PURPOSE/HYPOTHESIS: To evaluate a novel method for subject-specific AIF estimation that takes inflow effects into account. STUDY TYPE: Retrospective study. SUBJECTS: Thirteen clinical patients referred for spine-related complaints; 21 patients from a study into luminal Crohn's disease with known Crohn's Disease Endoscopic Index of Severity (CDEIS). FIELD STRENGTH/SEQUENCE: Dynamic fast spoiled gradient echo (FSPGR) at 3T. ASSESSMENT: A population-averaged AIF, AIFs derived from distally placed regions of interest (ROIs), and the new AIF method were applied. Tofts' PK model parameters (including vp and Ktrans ) obtained with the three AIFs were compared. In the Crohn's patients Ktrans was correlated to CDEIS. STATISTICAL TESTS: The median values of the PK model parameters from the three methods were compared using a Mann-Whitney U-test. The associated variances were statistically assessed by the Brown-Forsythe test. Spearman's rank correlation coefficient was computed to test the correlation of Ktrans to CDEIS. RESULTS: The median vp was significantly larger when using the distal ROI approach, compared to the two other methods (P < 0.05 for both comparisons, in both applications). Also, the variances in vp were significantly larger with the ROI approach (P < 0.05 for all comparisons). In the Crohn's disease study, the estimated Ktrans parameter correlated better with the CDEIS (r = 0.733, P < 0.001) when the proposed AIF was used, compared to AIFs from the distal ROI method (r = 0.429, P = 0.067) or the population-averaged AIF (r = 0.567, P = 0.011). DATA CONCLUSION: The proposed method yielded realistic PK model parameters and improved the correlation of the Ktrans parameter with CDEIS, compared to existing approaches. LEVEL OF EVIDENCE: 3 Technical Efficacy Stage 1 J. Magn. Reson. Imaging 2018;47:1197-1204.

Semi-automatic bowel wall thickness measurements on MR enterography in patients with Crohn's disease.

OBJECTIVE: To evaluate a semi-automatic method for delineation of the bowel wall and measurement of the wall thickness in patients with Crohn's disease. METHODS: 53 patients with suspected or proven Crohn's disease were selected. Two radiologists independently supervised the delineation of regions with active Crohn's disease on MRI, yielding manual annotations (Ano1, Ano2). Three observers manually measured the maximal bowel wall thickness of each annotated segment. An active contour segmentation approach semi-automatically delineated the bowel wall. For each active region, two segmentations (Seg1, Seg2) were obtained by independent observers, in which the maximum wall thickness was automatically determined. The overlap between (Seg1, Seg2) was compared with the overlap of (Ano1, Ano2) using Wilcoxon's signed rank test. The corresponding variances were compared using the Brown-Forsythe test. The variance of the semi-automatic thickness measurements was compared with the overall variance of manual measurements through an F-test. Furthermore, the intraclass correlation coefficient (ICC) of semi-automatic thickness measurements was compared with the ICC of manual measurements through a likelihood-ratio test. RESULTS: Patient demographics: median age, 30 years; interquartile range, 25-38 years; 33 females. The median overlap of the semi-automatic segmentations (Seg1 vs Seg2: 0.89) was significantly larger than the median overlap of the manual annotations (Ano1 vs Ano2: 0.72); p = 1.4 × 10-5. The variance in overlap of the semi-automatic segmentations was significantly smaller than the variance in overlap of the manual annotations (p = 1.1 × 10-9). The variance of the semi-automated measurements (0.46 mm2) was significantly smaller than the variance of the manual measurements (2.90 mm2, p = 1.1 × 10-7). The ICC of semi-automatic measurement (0.88) was significantly higher than the ICC of manual measurement (0.45); p = 0.005. CONCLUSION: The semi-automatic technique facilitates reproducible delineation of regions with active Crohn's disease. The semi-automatic thickness measurement sustains significantly improved interobserver agreement. Advances in knowledge: Automation of bowel wall thickness measurements strongly increases reproducibility of these measurements, which are commonly used in MRI scoring systems of Crohn's disease activity.

Image Registration Based on Autocorrelation of Local Structure.

Registration of images in the presence of intra-image signal fluctuations is a challenging task. The definition of an appropriate objective function measuring the similarity between the images is crucial for accurate registration. This paper introduces an objective function that embeds local phase features derived from the monogenic signal in the modality independent neighborhood descriptor (MIND). The image similarity relies on the autocorrelation of local structure (ALOST) which has two important properties: 1) low sensitivity to space-variant intensity distortions (e.g., differences in contrast enhancement in MRI); 2) high distinctiveness for 'salient' image features such as edges. The ALOST method is quantitatively compared to the MIND approach based on three different datasets: thoracic CT images, synthetic and real abdominal MR images. The proposed method outperformed the NMI and MIND similarity measures on these three datasets. The registration of dynamic contrast enhanced and post-contrast MR images of patients with Crohn's disease led to relative contrast enhancement measures with the highest correlation (r=0.56) to the Crohn's disease endoscopic index of severity.

Measuring murine chromosome orientation in interphase nuclei.

The nuclear architecture of a cell may change as a result of various diseases, including cancer. A variety of nuclear features are, therefore, of interest to cell biologists. Recently, several studies have investigated the orientation of chromosomes in the interphase nucleus either visually or semi-automatically. In this article an automated method to measure this orientation is presented. The theoretical difference between performing these measurements in two and three dimensions is discussed and experimentally verified. The results computed from measurements of murine nuclei correspond with results from visual inspection. We found significant differences in the orientation of chromosome 11 between nuclei from a PreB cell line of BALB/c origin and primary B nuclei from congenic [T38HxBALB/c]N wild-type mice. Since our new automatic method concurs with both the visual and semi-automatic methods, we conclude that the automatic method can replace these methods in assessing chromosome orientation.

Epicardial fat volume is related to atherosclerotic calcification in multiple vessel beds.

AIM: To investigate relationships between epicardial fat volume and atherosclerosis in multiple major vessel beds. METHODS AND RESULTS: From the population-based Rotterdam Study, 2298 participants underwent computed tomography examinations to quantify epicardial fat volume and atherosclerotic calcification volume in the coronary arteries, aortic arch, and extracranial and intracranial internal carotid arteries. Using linear regression modelling, we investigated relationships of epicardial fat volume with atherosclerotic calcification volume in each vessel bed, adjusting for conventional cardiovascular risk factors (waist circumference, systolic and diastolic blood pressure, total and high-density lipoprotein cholesterol, smoking, diabetes, and usage of blood pressure-lowering and lipid-lowering medication). To test whether associations of epicardial fat with calcification per vessel bed were independent of calcification elsewhere, we created a model in which all vessel beds were entered together. We found that a larger epicardial fat volume was associated with larger calcification volumes in the coronary arteries, aortic arch, and extracranial carotid arteries in both sexes. After adjustment for cardiovascular risk factors, larger epicardial fat volume was related to coronary and extracranial carotid artery calcification volume in males only [difference in calcification volume per SD increase in epicardial fat volume: 0.12 (95% confidence interval, CI: 0.04; 0.19) and 0.14 (95% CI: 0.06; 0.22)]. These associations remained unchanged after entering all vessel beds into one model. CONCLUSION: Larger volumes of epicardial fat are associated with larger amounts of coronary and extracranial carotid artery atherosclerosis in males, independent of cardiovascular risk factors. This could imply that epicardial fat also exerts a systemic effect on atherosclerosis development. Future longitudinal research is warranted to further disentangle these relationships with a specific focus on sex differences.

Expiration-phase template-based motion correction of free-breathing abdominal dynamic contrast enhanced MRI.

This paper studies a novel method to compensate for respiratory and peristaltic motions in abdominal dynamic contrast enhanced magnetic resonance imaging. The method consists of two steps: 1) expiration-phase "template" construction and retrospective gating of the data to the template; and 2) nonrigid registration of the gated volumes. Landmarks annotated by three experts were used to directly assess the registration performance. A tri-exponential function fit to time intensity curves from regions of interest was used to indirectly assess the performance. One of the parameters of the tri-exponential fit was used to quantify the contrast enhancement. Our method achieved a mean target registration error (MTRE) of 2.12, 2.27, and 2.33 mm with respect to annotations by expert, which was close to the average interobserver variability (2.07 mm). A state-of-the-art registration method achieved an MTRE of 2.83-3.10 mm. The correlation coefficient of the contrast enhancement parameter to the Crohn's disease endoscopic index of Severity (r = 0.60, p = 0.004) was higher than the correlation coefficient for the relative contrast enhancement measurements values of two observers ( r(Observer 1) = 0.29, p = 0.2; r(Observer 2) = 0.45, p = 0.04). Direct and indirect assessments show that the expiration-based gating and a nonrigid registration approach effectively corrects for respiratory motion and peristalsis. The method facilitates improved enhancement measurement in the bowel wall in patients with Crohn's disease.

Semi-automatic MRI segmentation and volume quantification of intra-plaque hemorrhage.

PURPOSE: Intra-plaque hemorrhage (IPH) is associated with plaque instability. Therefore, the presence and volume of IPH in carotid arteries may be relevant in predicting the progression of atherosclerotic disease and the occurrence of clinical events. The aim of our work was to develop and evaluate a method for semi-automatic IPH segmentation in T1-weighted (T1w)-magnetic resonance imaging (MRI). MATERIAL AND METHODS: IPH segmentation is performed by a regional level set method that models the intensity of the IPH and the background in T1w-MRI to be smoothly varying. The method only requires minimal user interaction, i.e., one or more mouse clicks inside the hemorrhage serve as initialization. The parameters of the method are optimized using a leave-one-out strategy by maximizing the Dice similarity coefficient (DSC) between manual and semi-automatic segmentations. We evaluated the IPH segmentation method on 22 carotid arteries; 10 of which were annotated by two observers and 12 were scanned twice within a 2 week period. RESULTS: We obtained a DSC of 0.52 between the manual and level set segmentations on all 22 carotids. The inter-observer DSC on 10 arteries is 0.57, which is comparable to the DSC between the method and the manual segmentation (0.55). The correlation between the IPH volumes extracted from the level set segmentation and the manual segmentation is 0.88, which is close to the inter-observer volume correlation of 0.92. The reproducibility after rescanning 12 carotids yield an IPH volume correlation of 0.97. The robustness with respect to the initialization by manually clicking two sets of seed points in these 12 carotid artery pairs yields a volume correlation of 0.99. CONCLUSION: Semi-automatic segmentation and quantification of IPHs are feasible with an accuracy in the range of the inter-observer variability. The method has excellent reproducibility with respect to rescanning and manual initialization.

Three-dimensional structured illumination microscopy using Lukosz bound apodization reduces pixel negativity at no resolution cost.

The quality of the reconstructed image in structured illumination microscopy (SIM) depends on various aspects of the image filtering process. To optimize the trade-off between resolution and ringing artifacts, which lead to negative intensities, we extend Lukosz-bound filtering to 3D SIM and derive the parametrization of the 3D SIM cut-off. We compare the use of the Lukosz-bound as apodization filter to triangular apodization and find a tenfold reduction in the most negative pixel value with a minimal resolution loss. We test this algorithm on experimental SIM images of tubulin filaments and DAPI stained DNA structure in cancer cells and find a substantial reduction in the most negative pixel value and the percentage of pixels with a negative value. This means that there is no longer a need to clip the final image to avoid these negative pixel values.

Differences in nuclear DNA organization between lymphocytes, Hodgkin and Reed-Sternberg cells revealed by structured illumination microscopy.

Advances in light microscopy have enabled the visualization of DNA in the interphase nucleus with more detail than is visible with conventional light microscopy. The nuclear architecture is assumed to be different in cancer cells compared to normal cells. In this paper we have studied, for the first time, the organization of nuclear DNA and that of DNA-free space in control lymphocytes, Hodgkin cells and Reed-Sternberg cells using 3D structured illumination microscopy (SIM). We have observed detail in these SIM images that was not observed in conventional widefield images. We have measured the size distribution of the DNA structure using granulometry and noted a significant, progressive increase in the amount of sub-micron structures from control lymphocytes to Hodgkin cells to Reed-Sternberg cells. The DNA-free space changes as well; "holes" in the DNA distribution start to appear in the malignant cells. We have studied whether these "holes" are nucleoli by staining for upstream binding factor (UBF), a protein associated with the nucleolus. We have found that the relative UBF content progressively and significantly decreases-or is absent-in the DNA-free space when measured as either the Pearson correlation coefficient with the DNA-free space or as the number of "holes" that contain UBF. Similar differences exist within the population of Reed-Sternberg cells between binucleated and multinucleated cells with four or more subnuclei. To our knowledge, this is the first study that investigates the changes of the nuclear DNA structure in any disease with superresolution light microscopy.

Electronic cleansing for 24-h limited bowel preparation CT colonography using principal curvature flow.

CT colonography (CTC) is one of the recommended methods for colorectal cancer screening. The subject's preparation is one of the most burdensome aspects of CTC with a cathartic bowel preparation. Tagging of the bowel content with an oral contrast medium facilitates CTC with limited bowel preparation. Unfortunately, such preparations adversely affect the 3-D image quality. Thus far, data acquired after very limited bowel preparation were evaluated with a 2-D reading strategy only. Existing cleansing algorithms do not work sufficiently well to allow a primary 3-D reading strategy. We developed an electronic cleansing algorithm, aimed to realize optimal 3-D image quality for low-dose CTC with 24-h limited bowel preparation. The method employs a principal curvature flow algorithm to remove heterogeneities within poorly tagged fecal residue. In addition, a pattern recognition-based approach is used to prevent polyp-like protrusions on the colon surface from being removed by the method. Two experts independently evaluated 40 CTC cases by means of a primary 2-D approach without involvement of electronic cleansing as well as by a primary 3-D method after electronic cleansing. The data contained four variations of 24-h limited bowel preparation and was based on a low radiation dose scanning protocol. The sensitivity for lesions ≥ 6 mm was significantly higher for the primary 3-D reading strategy (84%) than for the primary 2-D reading strategy (68%) (p = 0.031). The reading time was increased from 5:39 min (2-D) to 7:09 min (3-D) (p = 0.005); the readers' confidence was reduced from 2.3 (2-D) to 2.1 (3-D) ( p = 0.013) on a three-point Likert scale. Polyp conspicuity for cleansed submerged lesions was similar to not submerged lesions (p = 0.06). To our knowledge, this study is the first to describe and clinically validate an electronic cleansing algorithm that facilitates low-dose CTC with 24-h limited bowel preparation.

Vessel specific coronary artery calcium scoring: an automatic system.

RATIONALE AND OBJECTIVES: The aim of this study was to automatically detect and quantify calcium lesions for the whole heart as well as per coronary artery on non-contrast-enhanced cardiac computed tomographic images. MATERIALS AND METHODS: Imaging data from 366 patients were randomly selected from patients who underwent computed tomographic calcium scoring assessments between July 2004 and May 2009 at Erasmum MC, Rotterdam. These data included data sets with 1.5-mm and 3.0-mm slice spacing reconstructions and were acquired using four different scanners. The scores of manual observers, who annotated the data using commercially available software, served as ground truth. An automatic method for detecting and quantifying calcifications for each of the four main coronary arteries and the whole heart was trained on 209 data sets and tested on 157 data sets. Statistical testing included determining Pearson's correlation coefficients and Bland-Altman analysis to compare performance between the system and ground truth. Wilcoxon's signed-rank test was used to compare the interobserver variability to the system's performance. RESULTS: Automatic detection of calcified objects was achieved with sensitivity of 81.2% per calcified object in the 1.5-mm data set and sensitivity of 86.6% per calcified object in the 3.0-mm data set. The system made an average of 2.5 errors per patient in the 1.5-mm data set and 2.2 errors in the 3.0-mm data set. Pearson's correlation coefficients of 0.97 (P < .001) for both 1.5-mm and 3.0-mm scans with respect to the calcium volume score of the whole heart were found. The average R values over Agatston, mass, and volume scores for each of the arteries (left circumflex coronary artery, right coronary artery, and left main and left anterior descending coronary arteries) were 0.93, 0.96, and 0.99, respectively, for the 1.5-mm scans. Similarly, for 3.0-mm scans, R values were 0.94, 0.94, and 0.99, respectively. Risk category assignment was correct in 95% and 89% of the data sets in the 1.5-mm and 3-mm scans. CONCLUSIONS: An automatic vessel-specific coronary artery calcium scoring system was developed, and its feasibility for calcium scoring in individual vessels and risk category classification has been demonstrated.

A 4D statistical model of wrist bone motion patterns.

Direct imaging of ligament damage in the wrist remains a challenge. Still, such damage can be assessed indirectly through the analysis of changes in wrist pose and motion pattern. For this purpose we built a statistical reference model that describes healthy motion patterns. We show that such a model can also be used to detect and quantify pathologies. A model that only describes the global translations and rotations of the carpal bones is insufficiently accurate due to size and shape variations of the bones. We present a local statistical motion model that minimizes the influence of size and shape differences by analyzing the coordinate differences of pairs of points on adjacent bone surfaces. These differences are determined in a set of 14 healthy example wrists imaged in a range of poses by means of 4D-RX imaging. The distribution of the differences as a function of the pose form the local statistical motion model (LSMM). Translations of 2 mm and rotations of 20° with respect to the healthy example wrists are detected as outliers in the point pair distributions. An evaluation involving wrists with a damaged ligament between scaphoid and lunate shows that not only joint space widenings can be detected, but also shifts of congruent bone surfaces. The LSMM is also used to perform a virtual reconstruction of the most likely healthy wrist after a simulated perturbation of bones. The reconstruction precision is shown to be about 1 mm. Therefore, the presented 4D statistical model of wrist bone movement may become a valuable clinical tool for diagnosis and surgical planning.

Computational modeling for assessment of IBD: to be or not to be?

The grading of inflammatory bowel disease (IBD) severity is important to determine the proper treatment strategy and to quantify the response to treatment. Traditionally, ileocolonoscopy is considered the reference standard for assessment of IBD. However, the procedure is invasive and requires extensive bowel preparation. Magnetic resonance imaging (MRI) has become an important tool for determining the presence of disease activity. Unfortunately, only moderate interobserver agreement is reported for most of the radiological severity measures. There is a clear demand for automated evaluation of MR images in Crohn's disease (CD). This paper aims to introduce a preliminary suite of fundamental tools for assessment of CD severity. It involves procedures for image analysis, classification and visualization to predict the colonoscopy disease scores.

Detection and segmentation of colonic polyps on implicit isosurfaces by second principal curvature flow.

Today's computer aided detection systems for computed tomography colonography (CTC) enable automated detection and segmentation of colorectal polyps. We present a paradigm shift by proposing a method that measures the amount of protrudedness of a candidate object in a scale adaptive fashion. One of the main results is that the performance of the candidate detection depends only on one parameter, the amount of protrusion. Additionally the method yields correct polyp segmentation without the need of an additional segmentation step. The supervised pattern recognition involves a clear distinction between size related features and features related to shape or intensity. A Mahalanobis transformation of the latter facilitates ranking of the objects using a logistic classifier. We evaluate two implementations of the method on 84 patients with a total of 57 polyps larger than or equal to 6 mm. We obtained a performance of 95% sensitivity at four false positives per scan for polyps larger than or equal to 6 mm.

Electronic cleansing for computed tomography (CT) colonography using a scale-invariant three-material model.

A well-known reading pitfall in computed tomography (CT) colonography is posed by artifacts at T-junctions, i.e., locations where air-fluid levels interface with the colon wall. This paper presents a scale-invariant method to determine material fractions in voxels near such T-junctions. The proposed electronic cleansing method particularly improves the segmentation at those locations. The algorithm takes a vector of Gaussian derivatives as input features. The measured features are made invariant to the orientation-dependent apparent scale of the data and normalized in a way to obtain equal noise variance. A so-called parachute model is introduced that maps Gaussian derivatives onto material fractions near T-junctions. Projection of the noisy derivatives onto the model yields improved estimates of the true, underlying feature values. The method is shown to render an accurate representation of the object boundary without artifacts near junctions. Therefore, it enhances the reading of CT colonography in a 3-D display mode.

Computer-aided detection of polyps in CT colonography using logistic regression.

We present a computer-aided detection (CAD) system for computed tomography colonography that orders the polyps according to clinical relevance. The CAD system consists of two steps: candidate detection and supervised classification. The characteristics of the detection step lead to specific choices for the classification system. The candidates are ordered by a linear logistic classifier (logistic regression) based on only three features: the protrusion of the colon wall, the mean internal intensity, and a feature to discard detections on the rectal enema tube. This classifier can cope with a small number of polyps available for training, a large imbalance between polyps and non-polyp candidates, a truncated feature space, unbalanced and unknown misclassification costs, and an exponential distribution with respect to candidate size in feature space. Our CAD system was evaluated with data sets from four different medical centers. For polyps larger than or equal to 6 mm we achieved sensitivities of respectively 95%, 85%, 85%, and 100% with 5, 4, 5, and 6 false positives per scan over 86, 48, 141, and 32 patients. A cross-center evaluation in which the system is trained and tested with data from different sources showed that the trained CAD system generalizes to data from different medical centers and with different patient preparations. This is essential to application in large-scale screening for colorectal polyps.