Automated finite-element analysis for deformable registration of prostate images.

Two major factors preventing the routine clinical use of finite-element analysis for image registration are: 1) the substantial labor required to construct a finite-element model for an individual patient's anatomy and 2) the difficulty of determining an appropriate set of finite-element boundary conditions. This paper addresses these issues by presenting algorithms that automatically generate a high quality hexahedral finite-element mesh and automatically calculate boundary conditions for an imaged patient. Medial shape models called m-reps are used to facilitate these tasks and reduce the effort required to apply finite-element analysis to image registration. Encouraging results are presented for the registration of CT image pairs which exhibit deformation caused by pressure from an endorectal imaging probe and deformation due to swelling.

Comparison of human and automatic segmentations of kidneys from CT images.

PURPOSE: A controlled observer study was conducted to compare a method for automatic image segmentation with conventional user-guided segmentation of right and left kidneys from planning computerized tomographic (CT) images. METHODS AND MATERIALS: Deformable shape models called m-reps were used to automatically segment right and left kidneys from 12 target CT images, and the results were compared with careful manual segmentations performed by two human experts. M-rep models were trained based on manual segmentations from a collection of images that did not include the targets. Segmentation using m-reps began with interactive initialization to position the kidney model over the target kidney in the image data. Fully automatic segmentation proceeded through two stages at successively smaller spatial scales. At the first stage, a global similarity transformation of the kidney model was computed to position the model closer to the target kidney. The similarity transformation was followed by large-scale deformations based on principal geodesic analysis (PGA). During the second stage, the medial atoms comprising the m-rep model were deformed one by one. This procedure was iterated until no changes were observed. The transformations and deformations at both stages were driven by optimizing an objective function with two terms. One term penalized the currently deformed m-rep by an amount proportional to its deviation from the mean m-rep derived from PGA of the training segmentations. The second term computed a model-to-image match term based on the goodness of match of the trained intensity template for the currently deformed m-rep with the corresponding intensity data in the target image. Human and m-rep segmentations were compared using quantitative metrics provided in a toolset called Valmet. Metrics reported in this article include (1) percent volume overlap; (2) mean surface distance between two segmentations; and (3) maximum surface separation (Hausdorff distance). RESULTS: Averaged over all kidneys the mean surface separation was 0.12 cm, the mean Hausdorff distance was 0.99 cm, and the mean volume overlap for human segmentations was 88.8%. Between human and m-rep segmentations the mean surface separation was 0.18-0.19 cm, the mean Hausdorff distance was 1.14-1.25 cm, and the mean volume overlap was 82-83%. CONCLUSIONS: Overall in this study, the best m-rep kidney segmentations were at least as good as careful manual slice-by-slice segmentations performed by two experienced humans, and the worst performance was no worse than typical segmentations from our clinical setting. The mean surface separations for human-m-rep segmentations were slightly larger than for human-human segmentations but still in the subvoxel range, and volume overlap and maximum surface separation were slightly better for human-human comparisons. These results were expected because of experimental factors that favored comparison of the human-human segmentations. In particular, m-rep agreement with humans appears to have been limited largely by fundamental differences between manual slice-by-slice and true three-dimensional segmentation, imaging artifacts, image voxel dimensions, and the use of an m-rep model that produced a smooth surface across the renal pelvis.

A method and software for segmentation of anatomic object ensembles by deformable m-reps.

Deformable shape models (DSMs) comprise a general approach that shows great promise for automatic image segmentation. Published studies by others and our own research results strongly suggest that segmentation of a normal or near-normal object from 3D medical images will be most successful when the DSM approach uses (1) knowledge of the geometry of not only the target anatomic object but also the ensemble of objects providing context for the target object and (2) knowledge of the image intensities to be expected relative to the geometry of the target and contextual objects. The segmentation will be most efficient when the deformation operates at multiple object-related scales and uses deformations that include not just local translations but the biologically important transformations of bending and twisting, i.e., local rotation, and local magnification. In computer vision an important class of DSM methods uses explicit geometric models in a Bayesian statistical framework to provide a priori information used in posterior optimization to match the DSM against a target image. In this approach a DSM of the object to be segmented is placed in the target image data and undergoes a series of rigid and nonrigid transformations that deform the model to closely match the target object. The deformation process is driven by optimizing an objective function that has terms for the geometric typicality and model-to-image match for each instance of the deformed model. The success of this approach depends strongly on the object representation, i.e., the structural details and parameter set for the DSM, which in turn determines the analytic form of the objective function. This paper describes a form of DSM called m-reps that has or allows these properties, and a method of segmentation consisting of large to small scale posterior optimization of m-reps. Segmentation by deformable m-reps, together with the appropriate data representations, visualizations, and user interface, has been implemented in software that accomplishes 3D segmentations in a few minutes. Software for building and training models has also been developed. The methods underlying this software and its abilities are the subject of this paper.

ASTRO's core physics curriculum for radiation oncology residents.

In 2002, the Radiation Physics Committee of the American Society of Therapeutic Radiology and Oncology (ASTRO) appointed an Ad-hoc Committee on Physics Teaching to Medical Residents. The main initiative of the committee was to develop a core curriculum for physics education. Prior publications that have analyzed physics teaching have pointed to wide discrepancies among teaching programs. The committee was composed of physicists or physicians from various residency program based institutions. Simultaneously, members had associations with the American Association of Physicists in Medicine (AAPM), ASTRO, Association of Residents in Radiation Oncology (ARRO), American Board of Radiology (ABR), and the American College of Radiology (ACR). The latter two organizations' representatives were on the physics examination committees, as one of the main agendas was to provide a feedback loop between the examining organizations and ASTRO. The document resulted in a recommended 54-h course. Some of the subjects were based on American College of Graduate Medical Education (ACGME) requirements (particles, hyperthermia), whereas the majority of the subjects along with the appropriated hours per subject were devised and agreed upon by the committee. For each subject there are learning objectives and for each hour there is a detailed outline of material to be covered. Some of the required subjects/h are being taught in most institutions (i.e., Radiation Measurement and Calibration for 4 h), whereas some may be new subjects (4 h of Imaging for Radiation Oncology). The curriculum was completed and approved by the ASTRO Board in late 2003 and is slated for dissemination to the community in 2004. It is our hope that teaching physicists will adopt the recommended curriculum for their classes, and simultaneously that the ABR for its written physics examination and the ACR for its training examination will use the recommended curriculum as the basis for subject matter and depth of understanding. To ensure that the subject matter and emphasis remain current and relevant, the curriculum will be updated every 2 years.

Thresholds for human detection of patient setup errors in digitally reconstructed portal images of prostate fields.

PURPOSE: Computer-assisted methods to analyze electronic portal images for the presence of treatment setup errors should be studied in controlled experiments before use in the clinical setting. Validation experiments using images that contain known errors usually report the smallest errors that can be detected by the image analysis algorithm. This paper offers human error-detection thresholds as one benchmark for evaluating the smallest errors detected by algorithms. Unfortunately, reliable data are lacking describing human performance. The most rigorous benchmarks for human performance are obtained under conditions that favor error detection. To establish such benchmarks, controlled observer studies were carried out to determine the thresholds of detectability for in-plane and out-of-plane translation and rotation setup errors introduced into digitally reconstructed portal radiographs (DRPRs) of prostate fields. METHODS AND MATERIALS: Seventeen observers comprising radiation oncologists, radiation oncology residents, physicists, and therapy students participated in a two-alternative forced choice experiment involving 378 DRPRs computed using the National Library of Medicine Visible Human data sets. An observer viewed three images at a time displayed on adjacent computer monitors. Each image triplet included a reference digitally reconstructed radiograph displayed on the central monitor and two DRPRs displayed on the flanking monitors. One DRPR was error free. The other DRPR contained a known in-plane or out-of-plane error in the placement of the treatment field over a target region in the pelvis. The range for each type of error was determined from pilot observer studies based on a Probit model for error detection. The smallest errors approached the limit of human visual capability. The observer was told what kind of error was introduced, and was asked to choose the DRPR that contained the error. Observer decisions were recorded and analyzed using repeated-measures analysis of variance. RESULTS: The thresholds of detectability averaged over all observers were approximately 2.5 mm for in-plane translations, 1.6 degrees for in-plane rotations, 1 degrees for out-of-plane rotations, and 8% change in magnification for out-of-plane translations along the central axis. When one inexperienced observer is excluded, the average threshold for change in magnification is 5%. Experienced observers tended to perform better, but differences between groups were not statistically significant. Thresholds were computed as averages over all observers. Because of the broad range of observer capabilities, some detection tasks were too difficult for some observers, leading to missing threshold values in our data analysis. The missing values were excluded from computation of the average thresholds reported above. The effect of the missing values is to bias the average values toward the best human performance. CONCLUSIONS: Under favorable conditions, humans can detect small errors in setup geometry. The thresholds for error detection reported in this study are believed to represent rigorous but reasonable benchmarks that can be incorporated into studies evaluating algorithms for computer-assisted detection of setup errors in electronic portal images.

Deformable M-Reps for 3D Medical Image Segmentation.

M-reps (formerly called DSLs) are a multiscale medial means for modeling and rendering 3D solid geometry. They are particularly well suited to model anatomic objects and in particular to capture prior geometric information effectively in deformable models segmentation approaches. The representation is based on figural models, which define objects at coarse scale by a hierarchy of figures - each figure generally a slab representing a solid region and its boundary simultaneously. This paper focuses on the use of single figure models to segment objects of relatively simple structure. A single figure is a sheet of medial atoms, which is interpolated from the model formed by a net, i.e., a mesh or chain, of medial atoms (hence the name m-reps), each atom modeling a solid region via not only a position and a width but also a local figural frame giving figural directions and an object angle between opposing, corresponding positions on the boundary implied by the m-rep. The special capability of an m-rep is to provide spatial and orientational correspondence between an object in two different states of deformation. This ability is central to effective measurement of both geometric typicality and geometry to image match, the two terms of the objective function optimized in segmentation by deformable models. The other ability of m-reps central to effective segmentation is their ability to support segmentation at multiple levels of scale, with successively finer precision. Objects modeled by single figures are segmented first by a similarity transform augmented by object elongation, then by adjustment of each medial atom, and finally by displacing a dense sampling of the m-rep implied boundary. While these models and approaches also exist in 2D, we focus on 3D objects. The segmentation of the kidney from CT and the hippocampus from MRI serve as the major examples in this paper. The accuracy of segmentation as compared to manual, slice-by-slice segmentation is reported.

Regional appearance in deformable model segmentation.

Automated medical image segmentation is a challenging task that benefits from the use of effective image appearance models. In this paper, we compare appearance models at three regional scales for statistically characterizing image intensity near object boundaries in the context of segmentation via deformable models. The three models capture appearance in the form of regional intensity quantile functions. These distribution-based regional image descriptors are amenable to Euclidean methods such as principal component analysis, which we use to build the statistical appearance models. The first model uses two regions, the interior and exterior of the organ of interest. The second model accounts for exterior inhomogeneity by clustering on object-relative local intensity quantile functions to determine tissue-consistent regions relative to the organ boundary. The third model analyzes these image descriptors per geometrically defined local region. To evaluate the three models, we present segmentation results on bladders and prostates in CT in the context of day-to-day adaptive radiotherapy for the treatment of prostate cancer. Results show improved segmentations with more local regions, probably because smaller regions better represent local inhomogeneity in the intensity distribution near the organ boundary.

Geometrically proper models in statistical training.

In deformable model segmentation, the geometric training process plays a crucial role in providing shape statistical priors and appearance statistics that are used as likelihoods. Also, the geometric training process plays a crucial role in providing shape probability distributions in methods finding significant differences between classes. The quality of the training seriously affects the final results of segmentation or of significant difference finding between classes. However, the lack of shape priors in the training stage itself makes it difficult to enforce shape legality, i.e., making the model free of local self-intersection or creases. Shape legality not only yields proper shape statistics but also increases the consistency of parameterization of the object volume and thus proper appearance statistics. In this paper we propose a method incorporating explicit legality constraints in training process. The method is mathematically sound and has proved in practice to lead to shape probability distributions over only proper objects and most importantly to better segmentation results.