Standardization in Quantitative Imaging: A Multicenter Comparison of Radiomic Features from Different Software Packages on Digital Reference Objects and Patient Data Sets.

Radiomic features are being increasingly studied for clinical applications. We aimed to assess the agreement among radiomic features when computed by several groups by using different software packages under very tightly controlled conditions, which included standardized feature definitions and common image data sets. Ten sites (9 from the NCI's Quantitative Imaging Network] positron emission tomography-computed tomography working group plus one site from outside that group) participated in this project. Nine common quantitative imaging features were selected for comparison including features that describe morphology, intensity, shape, and texture. The common image data sets were: three 3D digital reference objects (DROs) and 10 patient image scans from the Lung Image Database Consortium data set using a specific lesion in each scan. Each object (DRO or lesion) was accompanied by an already-defined volume of interest, from which the features were calculated. Feature values for each object (DRO or lesion) were reported. The coefficient of variation (CV), expressed as a percentage, was calculated across software packages for each feature on each object. Thirteen sets of results were obtained for the DROs and patient data sets. Five of the 9 features showed excellent agreement with CV < 1%; 1 feature had moderate agreement (CV < 10%), and 3 features had larger variations (CV ≥ 10%) even after attempts at harmonization of feature calculations. This work highlights the value of feature definition standardization as well as the need to further clarify definitions for some features.

Radiomics in Brain Tumor: Image Assessment, Quantitative Feature Descriptors, and Machine-Learning Approaches.

Radiomics describes a broad set of computational methods that extract quantitative features from radiographic images. The resulting features can be used to inform imaging diagnosis, prognosis, and therapy response in oncology. However, major challenges remain for methodologic developments to optimize feature extraction and provide rapid information flow in clinical settings. Equally important, to be clinically useful, predictive radiomic properties must be clearly linked to meaningful biologic characteristics and qualitative imaging properties familiar to radiologists. Here we use a cross-disciplinary approach to highlight studies in radiomics. We review brain tumor radiologic studies (eg, imaging interpretation) through computational models (eg, computer vision and machine learning) that provide novel clinical insights. We outline current quantitative image feature extraction and prediction strategies with different levels of available clinical classes for supporting clinical decision-making. We further discuss machine-learning challenges and data opportunities to advance radiomic studies.

Imaging informatics: toward capturing and processing semantic information in radiology images.

OBJECTIVES: To identify challenges and opportunities in imaging informatics that can lead to the use of images for discovery, and that can potentially improve the diagnostic accuracy of imaging professionals. METHODS: Recent articles on imaging informatics and related articles from PubMed were reviewed and analyzed. Some new developments and challenges that recent research in imaging informatics will meet are identified and discussed. RESULTS: While much literature continues to be devoted to traditional imaging informatics topics of image processing, visualization, and computerized detection, three new trends are emerging: (1) development of ontologies to describe radiology reports and images, (2) structured reporting and image annotation methods to make image semantics explicit and machine-accessible, and (3) applications that use semantic image information for decision support to improve radiologist interpretation performance. The informatics methods being developed have similarities and synergies with recent work in the biomedical informatics community that leverage large high-throughput data sets, and future research in imaging informatics will build on these advances to enable discovery by mining large image databases. CONCLUSIONS: Imaging informatics is beginning to develop and apply knowledge representation and analysis methods to image datasets. This type of work, already commonplace in biomedical research with large scale molecular and clinical datasets, will lead to new ways for computers to work with image data. The new advances hold promise for integrating imaging with the rest of the patient record as well as molecular data, for new data-driven discoveries in imaging analogous to that in bioinformatics, and for improved quality of radiology practice.

Colon polyp detection using smoothed shape operators: preliminary results.

Computer-aided detection (CAD) algorithms identify locations in computed tomographic (CT) images of the colon that are most likely to contain polyps. Existing CAD methods treat the CT data as a voxelized, volume image. They estimate a curvature-based feature at the mucosal surface voxels. However, curvature is a smooth notion, while our data are discrete and noisy. As a second order differential quantity, curvature amplifies noise. In this paper, we present the smoothed shape operators method (SSO), which uses a geometry processing approach. We extract a triangle mesh representation of the colon surface, and estimate curvature on this surface using the shape operator. We then smooth the shape operators on the surface iteratively. Throughout, we use techniques explicitly designed for discrete geometry. All our computation occurs on the surface, rather than in the voxel grid. We evaluate our algorithm on patient data and provide free-response receiver-operating characteristic performance analysis over all size ranges of polyps. We also provide confidence intervals for our performance estimates. We compare our performance with the surface normal overlap (SNO) method for the same data. A preliminary evaluation of our method on 35 patients yielded the following results (polyp diameter range; sensitivity; false positives/case): (10mm; 100%; 17.5), (5-10 mm; 89.7%, 21.23), (<5 mm; 59.1%; 23.9) and (overall; 80.3%; 23.9). The evaluation of the SNO method yielded: (10 mm; 75%; 17.5), (5-10 mm; 43.1%; 21.23), (<5 mm; 15.9%; 23.9) and (overall; 38.5%; 23.9).

Prediction of aortoiliac stent-graft length: comparison of measurement methods.

PURPOSE: To determine the accuracy of helical computed tomography (CT), projectional angiography derived from CT angiography, and intravascular ultrasonographic withdrawal (IUW) length measurements for predicting appropriate aortoiliac stent-graft length. MATERIALS AND METHODS: Helical CT data from 33 patients were analyzed before and after endovascular repair of abdominal aortic aneurysm (Aneuryx graft, n = 31; Excluder graft, n = 2). The aortoiliac length of the median luminal centerline (MLC) and the shortest path (SP) that remained at least one common iliac arterial radius away from the vessel wall were calculated. Conventional angiographic measurements were simulated from CT data as the length of the three-dimensional MLC projected onto four standard viewing planes. These predeployment lengths and IUW length, available in 24 patients, were compared with the aortoiliac arterial length after stent-graft deployment. RESULTS: The mean error values of SP, MLC, the maximum projected MLC, and IUW were -2.1 mm +/- 4.6 (SD) (P =.013), 9.8 mm +/- 6.8 (P <.001), -5.2 mm +/- 7.8 (P <.001), and -14.1 mm +/- 9.3 (P <.001), respectively. The preprocedural prediction of the postprocedural aortoiliac length with the SP was significantly more accurate than that with the MLC (P <.001), maximum projected MLC (P <.001), and IUW (P <.001). CONCLUSION: The shortest aortoiliac path length maintaining at least one radius distance from the vessel wall most accurately enabled stent-graft length prediction for 31 AneuRx and two Excluder stent-grafts.

A statistical 3-D pattern processing method for computer-aided detection of polyps in CT colonography.

Adenomatous polyps in the colon are believed to be the precursor to colorectal carcinoma, the second leading cause of cancer deaths in United States. In this paper, we propose a new method for computer-aided detection of polyps in computed tomography (CT) colonography (virtual colonoscopy), a technique in which polyps are imaged along the wall of the air-inflated, cleansed colon with X-ray CT. Initial work with computer aided detection has shown high sensitivity, but at a cost of too many false positives. We present a statistical approach that uses support vector machines to distinguish the differentiating characteristics of polyps and healthy tissue, and uses this information for the classification of the new cases. One of the main contributions of the paper is the new three-dimensional pattern processing approach, called random orthogonal shape sections method, which combines the information from many random images to generate reliable signatures of shape. The input to the proposed system is a collection of volume data from candidate polyps obtained by a high-sensitivity, low-specificity system that we developed previously. The results of our ten-fold cross-validation experiments show that, on the average, the system increases the specificity from 0.19 (0.35) to 0.69 (0.74) at a sensitivity level of 1.0 (0.95).

Stair-step artifacts with single versus multiple detector-row helical CT.

PURPOSE: To compare the effects of acquisition parameters on the magnitude and appearance of artifacts between single and multiple detector-row helical computed tomography (CT). MATERIALS AND METHODS: A cylindric (12.7 x 305.0-mm) acrylic rod inclined 45 degrees relative to the z axis was scanned at the isocenter and 100 mm from the isocenter with single detector-row (single-channel) helical CT (beam width, 1-10 mm; pitch, 1.0, 2.0, or 3.0) and multiple detector-row (four-channel) helical CT (detector width, 1. 25, 2.5, 3.75, and 5 mm; pitch, 0.75 or 1.5). The SD of radius measurements along the rod (SD(r)) was used to quantify artifacts in all 72 data sets and to analyze their frequency patterns. Volume-rendered images of the data sets were ranked by six independent and blinded readers; findings were correlated with acquisition parameters and SD(r) measurements. RESULTS: SD(r) was smaller in four- than in single-channel helical CT for any given table increment (TI). In single-channel helical CT, SD(r) increased linearly with beam width and geometrically with pitch. In four-channel helical CT, SD(r) measurements were directly proportional to the TI, regardless of the detector width and pitch combination used. Off-center object position on average increased SD(r) by a factor of 1.6 for single-channel helical CT and by a factor of 2.0 for four-channel helical CT. Subjective rankings of image quality correlated excellently with SD(r) (Spearman r = 0.94, P <.001). CONCLUSION: Artifacts are quantitatively and subjectively smaller with four- compared with single-channel helical CT for any given TI.

Automated polyp detector for CT colonography: feasibility study.

An abdominal computed tomographic scan was modified by inserting 10 simulated colonic polyps with use of methods that closely mimic the attenuation, noise, and polyp-colon wall interface of naturally occurring polyps. A shape-based polyp detector successfully located six of the 10 polyps. When settings that enhanced the edge profile of polyps were chosen, eight of 10 polyps were detected. There were no false-positive detections. Shape analysis is technically feasible and is a promising approach to automated polyp detection.

Reconstruction algorithm for polychromatic CT imaging: application to beam hardening correction.

This paper presents a new reconstruction algorithm for both single- and dual-energy computed tomography (CT) imaging. By incorporating the polychromatic characteristics of the X-ray beam into the reconstruction process, the algorithm is capable of eliminating beam hardening artifacts. The single energy version of the algorithm assumes that each voxel in the scan field can be expressed as a mixture of two known substances, for example, a mixture of trabecular bone and marrow, or a mixture of fat and flesh. These assumptions are easily satisfied in a quantitative computed tomography (QCT) setting. We have compared our algorithm to three commonly used single-energy correction techniques. Experimental results show that our algorithm is much more robust and accurate. We have also shown that QCT measurements obtained using our algorithm are five times more accurate than that from current QCT systems (using calibration). The dual-energy mode does not require any prior knowledge of the object in the scan field, and can be used to estimate the attenuation coefficient function of unknown materials. We have tested the dual-energy setup to obtain an accurate estimate for the attenuation coefficient function of K2 HPO4 solution.

Medical image segmentation using analysis of isolable-contour maps.

A common challenge for automated segmentation techniques is differentiation between images of close objects that have similar intensities, whose boundaries are often blurred due to partial-volume effects. We propose a novel approach to segmentation of two-dimensional images, which addresses this challenge. Our method, which we call intrinsic shape for segmentation (ISeg), analyzes isolabel-contour maps to identify coherent regions that correspond to major objects. ISeg generates an isolabel-contour map for an image by multilevel thresholding with a fine partition of the intensity range. ISeg detects object boundaries by comparing the shape of neighboring isolabel contours from the map. ISeg requires only little effort from users; it does not require construction of shape models of target objects. In a formal validation with computed-tomography angiography data, we showed that ISeg was more robust than conventional thresholding, and that ISeg's results were comparable to results of manual tracing.

Cost identification of abdominal aortic aneurysm imaging by using time and motion analyses.

PURPOSE: To compare the costs of performing helical computed tomographic (CT) angiography with three-dimensional rendering versus intraarterial digital subtraction angiography (DSA) for preoperative imaging of abdominal aortic aneurysms (AAAs). MATERIALS AND METHODS: A single observer determined the variable direct costs of performing nine intraarterial DSA and 10 CT angiographic examinations in age- and general health-matched patients with AAA by using time and motion analyses. All personnel directly involved in the cases were tracked, and the involvement times were recorded to the nearest minute. All material items used during the procedures were recorded. The cost of labor was determined from personnel reimbursement data, and the cost of materials, from vendor pricing. The variable direct costs of laboratory tests and using the ambulatory treatment unit for postprocedural monitoring, as well as all fixed direct costs, were assessed from hospital accounting records. The total costs were determined for each procedure and compared by using the Student t test and calculating the CIs. RESULTS: The mean total direct cost of intraarterial DSA (+/- SD) was $1,052 +/- 71, and that of CT angiography was $300 +/- 30, which are significantly different (P < 4.1 x 10(-11)). With 95% confidence, intraarterial DSA cost 3.2-3.7 times more than CT angiography for the assessment of AAA. CONCLUSION: Assuming equal diagnostic utility and procedure-related morbidity, institutions may have substantial cost savings whenever CT angiography can replace intraarterial DSA for imaging AAAs.

Visualization modes for CT colonography using cylindrical and planar map projections.

PURPOSE: The purpose of this study was to demonstrate the limitations to the effectiveness of CT colonography, colloquially called virtual colonoscopy (VC), for detecting polyps in the colon and to describe a new technique, map projection CT colonography using Mercator projection and stereographic projection, that overcomes these limitations. METHOD: In one experiment, data sets from nine patients undergoing CT colonography were analyzed to determine the percentage of the mucosal surface visible in various visualization modes as a function of field of view (FOV). In another experiment, 40 digitally synthesized polyps of various sizes (10, 7, 5, and 3.5 mm) were randomly inserted into four copies of one patient data set. Both Mercator and stereographic projections were used to visualize the surface of the colon of each data set. The sensitivity and positive predictive value (PPV) were calculated and compared with the results of an earlier study of visualization modes using the same CT colonography data. RESULTS: The percentage of mucosal surface visualized by VC increases with greater FOV but only approaches that of map projection VC (98.8%) at a distorting, very high FOV. For both readers and polyp sizes of > or =7 mm, sensitivity for Mercator projection (87.5%) and stereographic projection (82.5%) was significantly greater (p < 0.05) than for viewing axial slices (62.5%), and Mercator projection was significantly more sensitive than VC (67.5%). Mercator and stereographic projection had PPVs of 75.4 and 78.9%, respectively. CONCLUSION: The sensitivity of conventional CT colonography is limited by the percentage of the mucosal surface seen. Map projection CT colonography overcomes this problem and provides a more sensitive method with a high PPV for detecting polyps than other methods currently being investigated.

Computed tomographic angiography: historical perspective and new state-of-the-art using multi detector-row helical computed tomography.

Since its clinical introduction in 1991, volumetric computed tomography scanning using spiral or helical scanners has resulted in a revolution for diagnostic imaging. In addition to new applications for computed tomography, such as computed tomographic angiography and the assessment of patients with renal colic, many routine applications such as the detection of lung and liver lesions have substantially improved. Helical computed tomographic technology has improved over the past eight years with faster gantry rotation, more powerful X-ray tubes, and improved interpolation algorithms, but the greatest advance has been the recent introduction of multi detector-row computed tomography scanners. These scanners provide similar scan quality at a speed gain of 3-6 times greater than single detector-row computed tomography scanners. This has a profound impact on the performance of computed tomography angiography, resulting in greater anatomic coverage, lower iodinated contrast doses, and higher spatial resolution scans than single detector-row systems.

Display modes for CT colonography. Part I. Synthesis and insertion of polyps into patient CT data.

PURPOSE: To develop and validate a method for the insertion of digitally synthesized polyps into computed tomographic (CT) images of the human colon for use as ground truth for evaluation of virtual colonoscopy. MATERIALS AND METHODS: Spiral CT simulator software was used to generate 10 synthetic polyps in various configurations. Additional software was developed to insert these polyps into volume CT scans. Ten polyps in eight patients were selected for comparison. Three radiologists evaluated whether two-dimensional (2D) CT images and three-dimensional (3D) volume-rendered CT images showed synthetic or real polyps. RESULTS: Edge-response profiles and noise of simulated polyps matched those of native polyps. Frequency distributions of reviewers' responses were not significantly different for synthetic versus real polyps in either 3D or 2D images. Responses were clustered around the response of "unsure" if lesions were real or synthetic. Receiver operating characteristic curves had areas of 0.54 (95% CI = 0.39, 0.68) for 3D and 0.39 (95% CI = 0.25, 0.53) for 2D images, which were not significantly different from random guessing (P = .70 and .28 for 3D and 2D images, respectively). CONCLUSION: Synthetic polyps were indistinguishable from real polyps. This method can be used to generate ground truth experimental data for comparison of CT colonographic display and detection methods.

Display modes for CT colonography. Part II. Blinded comparison of axial CT and virtual endoscopic and panoramic endoscopic volume-rendered studies.

PURPOSE: To determine the sensitivity of radiologist observers for detecting colonic polyps by using three different data review (display) modes for computed tomographic (CT) colonography, or "virtual colonoscopy." MATERIALS AND METHODS: CT colonographic data in a patient with a normal colon were used as base data for insertion of digitally synthesized polyps. Forty such polyps (3.5, 5, 7, and 10 mm in diameter) were randomly inserted in four copies of the base data. Axial CT studies, volume-rendered virtual endoscopic movies, and studies from a three-dimensional mode termed "panoramic endoscopy" were reviewed blindly and independently by two radiologists. RESULTS: Detection improved with increasing polyp size. Trends in sensitivity were dependent on whether all inserted lesions or only visible lesions were considered, because modes differed in how completely the colonic surface was depicted. For both reviewers and all polyps 7 mm or larger, panoramic endoscopy resulted in significantly greater sensitivity (90%) than did virtual endoscopy (68%, P = .014). For visible lesions only, the sensitivities were 85%, 81%, and 60% for one reader and 65%, 62%, and 28% for the other for virtual endoscopy, panoramic endoscopy, and axial CT, respectively. Three-dimensional displays were more sensitive than two-dimensional displays (P < .05). CONCLUSION: The sensitivity of panoramic endoscopy is higher than that of virtual endoscopy, because the former displays more of the colonic surface. Higher sensitivities for three-dimensional displays may justify the additional computation and review time.

Modeling of polychromatic attenuation using computed tomography reconstructed images.

This paper presents a procedure for estimating an accurate model of the CT imaging process including spectral effects. As raw projection data are typically unavailable to the end-user, we adopt a post-processing approach that utilizes the reconstructed images themselves. This approach includes errors from x-ray scatter and the nonidealities of the built-in soft tissue correction into the beam characteristics, which is crucial to beam hardening correction algorithms that are designed to be applied directly to CT reconstructed images. We formulate this approach as a quadratic programming problem and propose two different methods, dimension reduction and regularization, to overcome ill conditioning in the model. For the regularization method we use a statistical procedure, Cross Validation, to select the regularization parameter. We have constructed step-wedge phantoms to estimate the effective beam spectrum of a GE CT-I scanner. Using the derived spectrum, we computed the attenuation ratios for the wedge phantoms and found that the worst case modeling error is less than 3% of the corresponding attenuation ratio. We have also built two test (hybrid) phantoms to evaluate the effective spectrum. Based on these test phantoms, we have shown that the effective beam spectrum provides an accurate model for the CT imaging process. Last, we used a simple beam hardening correction experiment to demonstrate the effectiveness of the estimated beam profile for removing beam hardening artifacts. We hope that this estimation procedure will encourage more independent research on beam hardening corrections and will lead to the development of application-specific beam hardening correction algorithms.

Virtual endoscopy using perspective volume-rendered three-dimensional sonographic data: technique and clinical applications.

OBJECTIVE: We present a technique for obtaining three-dimensional external and virtual endoscopy views of organs using perspective volume-rendered gray-scale and Doppler sonographic data, and we explore potential clinical applications in the carotid artery, the female pelvis, and the bladder. CONCLUSION: Using the proposed methods, radiologists will find it possible to create virtual endoscopy and external perspective views using sonographic data. The technique works well for revealing the interior of fluid-filled structures and cavities. However, expected improvements in computer performance and integration with existing sonographic equipment will be necessary for the technique to become practical in the clinical environment.

Fast 3D cardiac cine MR imaging.

We describe a technique for three-dimensional cine MR imaging. By using short repetition times (TR) and interleaved slice encoding, volumetric cine data can be acquired throughout the cardiac cycle with a temporal resolution of approximately 80 msec. A T1-shortening agent is used to produce contrast between blood and myocardium. A comparison between the acquisition times of this and several other two-dimensional techniques is presented.

Longitudinal sampling and aliasing in spiral CT.

Although analyses of in-plane aliasing have been done for conventional computed tomography (CT) images, longitudinal aliasing in spiral CT has not been properly investigated. We propose a mathematical model of the three-dimensional (3-D) sampling scheme in spiral CT and analyze its effects on longitudinal aliasing. We investigated longitudinal aliasing as a function of the helical-interpolation algorithm, pitch, and reconstruction interval using CT simulations and actual phantom scans. Our model predicts, and we verified, that for a radially uniform object at the isocenter, the spiral sampling scheme results in spatially varying cancellation of the aliased spectral islands which, in turn, results in spatially varying longitudinal aliasing. The aliasing is minimal at the scanner isocenter, but worsens with distance from it and rapidly becomes significant. Our results agree with published results observed at the isocenter of the scanner and further provide new insight into the aliasing conditions at off-isocenter locations with respect to the pitch, interpolation algorithm, and reconstruction interval. We conclude that longitudinal aliasing at off-isocenter locations can be significant, and that its magnitude and effects cannot be predicted by measurements made only at the scanner isocenter.

Spatially varying longitudinal aliasing and resolution in spiral computed tomography.

Spiral computed tomography (CT) has revolutionized conventional CT as a truly three-dimensional imaging modality. A number of studies aimed at evaluating the longitudinal resolution in spiral CT have been presented, but the spatially varying nature of the longitudinal resolution in spiral CT has been largely left undiscussed. In this paper, we investigate the longitudinal resolution in spiral CT as affected by the spatially varying longitudinal aliasing. We propose the treatment of aliasing as a signal dependent, additive noise, and define a new image quality parameter, the contrast-to-aliased-noise ratio (CNaR), that relates to possible image degradation or loss of resolution caused by aliasing. We performed CT simulations and actual phantom scans using a resolution phantom consisting of sequences of spherical beads of different diameters, extending along the longitudinal axis. Our results show that the off-isocenter longitudinal resolution differs significantly from the longitudinal resolution at the isocenter and that the CNaR decreases with distance from the isocenter, and is a function of pitch and the helical interpolation algorithm used. The longitudinal resolution was observed to worsen with decreasing CNaR. We conclude that the longitudinal resolution in spiral CT is spatially varying, and can be characterized by the CNaR measured at the transaxial location of interest.