We propose a shape-based approach to curve evolution for the segmentation of medical images containing known object types. In particular, motivated by the work of Leventon, Grimson, and Faugeras, we derive a parametric model for an implicit representation of the segmenting curve by applying principal component analysis to a collection of signed distance representations of the training data. The parameters of this representation are then manipulated to minimize an objective function for segmentation. The resulting algorithm is able to handle multidimensional data, can deal with topological changes of the curve, is robust to noise and initial contour placements, and is computationally efficient. At the same time, it avoids the need for point correspondences during the training phase of the algorithm. We demonstrate this technique by applying it to two medical applications; two-dimensional segmentation of cardiac magnetic resonance imaging (MRI) and three-dimensional segmentation of prostate MRI.

%B IEEE Trans Med Imaging %V 22 %P 137-54 %8 2003 Feb %G eng %N 2 %1 http://www.ncbi.nlm.nih.gov/pubmed/12715991?dopt=Abstract %R 10.1109/TMI.2002.808355 %0 Journal Article %J Neuroimage %D 2003 %T Spatial normalization of diffusion tensor MRI using multiple channels %A Park, Hae-Jeong %A Kubicki, Marek %A Shenton, Martha E %A Guimond, Alexandre %A McCarley, Robert W %A Maier, Stephan E. %A Kikinis, Ron %A Jolesz, Ferenc A %A Westin, Carl-Fredrik %K Adult %K Algorithms %K Brain %K Brain Mapping %K Female %K Humans %K Magnetic Resonance Imaging %K Male %K Middle Aged %K Nerve Fibers %K Reference Standards %K Reference Values %X Diffusion Tensor MRI (DT-MRI) can provide important in vivo information for the detection of brain abnormalities in diseases characterized by compromised neural connectivity. To quantify diffusion tensor abnormalities based on voxel-based statistical analysis, spatial normalization is required to minimize the anatomical variability between studied brain structures. In this article, we used a multiple input channel registration algorithm based on a demons algorithm and evaluated the spatial normalization of diffusion tensor image in terms of the input information used for registration. Registration was performed on 16 DT-MRI data sets using different combinations of the channels, including a channel of T2-weighted intensity, a channel of the fractional anisotropy, a channel of the difference of the first and second eigenvalues, two channels of the fractional anisotropy and the trace of tensor, three channels of the eigenvalues of the tensor, and the six channel tensor components. To evaluate the registration of tensor data, we defined two similarity measures, i.e., the endpoint divergence and the mean square error, which we applied to the fiber bundles of target images and registered images at the same seed points in white matter segmentation. We also evaluated the tensor registration by examining the voxel-by-voxel alignment of tensors in a sample of 15 normalized DT-MRIs. In all evaluations, nonlinear warping using six independent tensor components as input channels showed the best performance in effectively normalizing the tract morphology and tensor orientation. We also present a nonlinear method for creating a group diffusion tensor atlas using the average tensor field and the average deformation field, which we believe is a better approach than a strict linear one for representing both tensor distribution and morphological distribution of the population. %B Neuroimage %V 20 %P 1995-2009 %8 2003 Dec %G eng %N 4 %1 http://www.ncbi.nlm.nih.gov/pubmed/14683705?dopt=Abstract %0 Journal Article %J Acad Radiol %D 2003 %T Statistical Validation Based on Parametric Receiver Operating Characteristic Analysis of Continuous Classification Data %A Zou, Kelly H %A Warfield, Simon K %A Fielding, Julia R %A Tempany, Clare M %A William, M Wells %A Kaus, Michael R %A Jolesz, Ferenc A %A Kikinis, Ron %K Algorithms %K Brain Neoplasms %K Humans %K Magnetic Resonance Imaging %K Male %K Prostate-Specific Antigen %K Prostatectomy %K Prostatic Neoplasms %K ROC Curve %K Tomography, X-Ray Computed %K Ureteral Calculi %XRATIONALE AND OBJECTIVES: The accuracy of diagnostic test and imaging segmentation is important in clinical practice because it has a direct impact on therapeutic planning. Statistical validations of classification accuracy was conducted based on parametric receiver operating characteristic analysis, illustrated on three radiologic examples, MATERIALS AND METHODS: Two parametric models were developed for diagnostic or imaging data. Example 1: A semi-automated fractional segmentation algorithm was applied to magnetic resonance imaging of nine cases of brain tumors. The tumor and background pixel data were assumed to have bi-beta distributions. Fractional segmentation was validated against an estimated composite pixel-wise gold standard based on multi-reader manual segmentations. Example 2: The predictive value of 100 cases of spiral computed tomography of ureteral stone sizes, distributed as bi-normal after a non-linear transformation, under two treatment options received. Example 3: One hundred eighty cases had prostate-specific antigen levels measured in a prospective clinical trial. Radical prostatectomy was performed in all to provide a binary gold standard of local and advanced cancer stages. Prostate-specific antigen level was transformed and modeled by bi-normal distributions. In all examples, areas under the receiver operating characteristic curves were computed. RESULTS. The areas under the receiver operating characteristic curves were: Example 1: Fractional segmentation of magnetic resonance imaging of brain tumors: meningiomas (0.924-0.984); astrocytomas (0.786-0.986); and other low-grade gliomas (0.896-0.983). Example 3: Ureteral stone size for treatment planning (0.813). Example 2: Prostate-specific antigen for staging prostate cancer (0.768). CONCLUSION: All clinical examples yielded fair to excellent accuracy. The validation metric area under the receiver operating characteristic curves may be generalized to evaluating the performances of several continuous classifiers related to imaging.

%B Acad Radiol %V 10 %P 1359-68 %8 2003 Dec %G eng %N 12 %1 http://www.ncbi.nlm.nih.gov/pubmed/14697004?dopt=Abstract %0 Journal Article %J J Magn Reson Imaging %D 2003 %T Three-dimensional analysis of the geometry of individual multiple sclerosis lesions: detection of shape changes over time using spherical harmonics %A Goldberg-Zimring, Daniel %A Achiron, Anat %A Guttmann, Charles R G %A Azhari, Haim %K Adult %K Follow-Up Studies %K Humans %K Imaging, Three-Dimensional %K Magnetic Resonance Imaging %K Male %K Multiple Sclerosis %X PURPOSE: To suggest a quantitative method for assessing the temporal changes in the geometry of individual multiple sclerosis (MS) lesions in follow-up studies of MS patients. MATERIALS AND METHODS: Computer simulated and in vivo magnetic resonance (MR) imaged MS lesions were studied. Ten in vivo MS lesions were identified from sets of axial MR images acquired from a patient scanned consecutively for 24 times during a one-year period. Each of the lesions was segmented and its three-dimensional surface approximated using spherical harmonics (SH). From the obtained SH polynomial coefficients, indices of shape were defined, and analysis of the temporal changes in each lesion's geometry throughout the year was performed by determining the mean discrete total variation of the shape indices. RESULTS: The results demonstrate that most of the studied lesions undergo notable geometrical changes with time. These changes are not necessarily associated with similar changes in size/volume. Furthermore, it was found that indices corresponding to changes in lesion shape could be 1.4 to 8.0 times higher than those corresponding to changes in the lesion size/volume. CONCLUSION: Quantitative three-dimensional shape analysis can serve as a new tool for monitoring MS lesion activity and study patterns of MS lesion evolution over time. %B J Magn Reson Imaging %V 18 %P 291-301 %8 2003 Sep %G eng %N 3 %1 http://www.ncbi.nlm.nih.gov/pubmed/12938123?dopt=Abstract %R 10.1002/jmri.10365 %0 Journal Article %J Neuroimage %D 2003 %T Time-series analysis of MRI intensity patterns in multiple sclerosis %A Meier, Dominik S %A Guttmann, Charles R G %K Algorithms %K Artifacts %K Brain Mapping %K Disease Progression %K Follow-Up Studies %K Humans %K Image Interpretation, Computer-Assisted %K Magnetic Resonance Imaging %K Multiple Sclerosis %K Reproducibility of Results %X In progressive neurological disorders, such as multiple sclerosis (MS), magnetic resonance imaging (MRI) follow-up is used to monitor disease activity and progression and to understand the underlying pathogenic mechanisms. This article presents image postprocessing methods and validation for integrating multiple serial MRI scans into a spatiotemporal volume for direct quantitative evaluation of the temporal intensity profiles. This temporal intensity signal and its dynamics have thus far not been exploited in the study of MS pathogenesis and the search for MRI surrogates of disease activity and progression. The integration into a four-dimensional data set comprises stages of tissue classification, followed by spatial and intensity normalization and partial volume filtering. Spatial normalization corrects for variations in head positioning and distortion artifacts via fully automated intensity-based registration algorithms, both rigid and nonrigid. Intensity normalization includes separate stages of correcting intra- and interscan variations based on the prior tissue class segmentation. Different approaches to image registration, partial volume correction, and intensity normalization were validated and compared. Validation included a scan-rescan experiment as well as a natural-history study on MS patients, imaged in weekly to monthly intervals over a 1-year follow-up. Significant error reduction was observed by applying tissue-specific intensity normalization and partial volume filtering. Example temporal profiles within evolving multiple sclerosis lesions are presented. An overall residual signal variance of 1.4% +/- 0.5% was observed across multiple subjects and time points, indicating an overall sensitivity of 3% (for axial dual echo images with 3-mm slice thickness) for longitudinal study of signal dynamics from serial brain MRI. %B Neuroimage %V 20 %P 1193-209 %8 2003 Oct %G eng %N 2 %1 http://www.ncbi.nlm.nih.gov/pubmed/14568488?dopt=Abstract %R 10.1016/S1053-8119(03)00354-9 %0 Journal Article %J IEEE Trans Biomed Eng %D 2003 %T Tumor detection in the bladder wall with a measurement of abnormal thickness in CT scans %A Jaume, Sylvain %A Ferrant, Matthieu %A Macq, Benoît %A Hoyte, Lennox %A Fielding, Julia R %A Schreyer, Andreas %A Kikinis, Ron %A Warfield, Simon K %K Anatomy, Cross-Sectional %K Echocardiography %K False Negative Reactions %K False Positive Reactions %K Humans %K Imaging, Three-Dimensional %K Radiographic Image Enhancement %K Radiographic Image Interpretation, Computer-Assisted %K Reference Values %K Reproducibility of Results %K Sensitivity and Specificity %K Urinary Bladder Neoplasms %X Virtual cystoscopy is a developing technique for bladder cancer screening. In a conventional cystoscopy, an optical probe is inserted into the bladder and an expert reviews the appearance of the bladder wall. Physical limitations of the probe place restrictions on the examination of the bladder wall. In virtual cystoscopy, a computed tomography (CT) scan of the bladder is acquired and an expert reviews the appearance of the bladder wall as shown by the CT. The task of identifying tumors in the bladder wall has often been done without extensive computational aid to the expert. We have developed an image processing algorithm that aids the expert in the detection of bladder tumors. Compared with an expert observer reading the CT, our algorithm achieves 89% sensitivity, 88% specificity, 48% positive predictive value, and 98% negative predictive value. %B IEEE Trans Biomed Eng %V 50 %P 383-90 %8 2003 Mar %G eng %N 3 %1 http://www.ncbi.nlm.nih.gov/pubmed/12669995?dopt=Abstract %R 10.1109/TBME.2003.808828 %0 Journal Article %J Inf Process Med Imaging %D 2003 %T Coupled Multi-shape Model and Mutual Information for Medical Image Segmentation %A Tsai, A %A Wells III, William M %A Tempany, Clare M %A Grimson, W Eric L %A Willsky, Alan S %K Algorithms %K Artificial Intelligence %K Computer Simulation %K Humans %K Image Enhancement %K Image Interpretation, Computer-Assisted %K Imaging, Three-Dimensional %K Male %K Models, Biological %K Models, Statistical %K Pattern Recognition, Automated %K Pelvis %K Principal Component Analysis %K Prostate %K Subtraction Technique %XThis paper presents extensions which improve the performance of the shape-based deformable active contour model presented earlier in [9]. In contrast to that work, the segmentation framework that we present in this paper allows multiple shapes to be segmented simultaneously in a seamless fashion. To achieve this, multiple signed distance functions are employed as the implicit representations of the multiple shape classes within the image. A parametric model for this new representation is derived by applying principal component analysis to the collection of these multiple signed distance functions. By deriving a parametric model in this manner, we obtain a coupling between the multiple shapes within the image and hence effectively capture the co-variations among the different shapes. The parameters of the multi-shape model are then calculated to minimize a single mutual information-based cost functional for image segmentation. The use of a single cost criterion further enhances the coupling between the multiple shapes as the deformation of any given shape depends, at all times, upon every other shape, regardless of their proximity. We demonstrate the utility of this algorithm to the segmentation of the prostate gland, the rectum, and the internal obturator muscles for MR-guided prostate brachytherapy.

%B Inf Process Med Imaging %V 18 %P 185-97 %8 2003 Jul %G eng %1 http://www.ncbi.nlm.nih.gov/pubmed/15344457?dopt=Abstract %0 Journal Article %J Inf Process Med Imaging %D 2003 %T A Unified Variational Approach to Denoising and Bias Correction in MR %A Fan, Ayres %A Wells III, William M %A Fisher, John W %A Cetin, Müjdat %A Haker, Steven %A Mulkern, Robert %A Tempany, Clare M %A Willsky, Alan S %K Algorithms %K Artifacts %K Brain %K Computer Simulation %K Heart %K Humans %K Image Enhancement %K Image Interpretation, Computer-Assisted %K Imaging, Three-Dimensional %K Magnetic Resonance Imaging %K Male %K Prostate %K Quality Control %K Reproducibility of Results %K Sensitivity and Specificity %XWe propose a novel bias correction method for magnetic resonance (MR) imaging that uses complementary body coil and surface coil images. The former are spatially homogeneous but have low signal intensity; the latter provide excellent signal response but have large bias fields. We present a variational framework where we optimize an energy functional to estimate the bias field and the underlying image using both observed images. The energy functional contains smoothness-enforcing regularization for both the image and the bias field. We present extensions of our basic framework to a variety of imaging protocols. We solve the optimization problem using a computationally efficient numerical algorithm based on coordinate descent, preconditioned conjugate gradient, half-quadratic regularization, and multigrid techniques. We show qualitative and quantitative results demonstrating the effectiveness of the proposed method in producing debiased and denoised MR images.

%B Inf Process Med Imaging %V 18 %P 148-59 %8 2003 Jul %G eng %1 http://www.ncbi.nlm.nih.gov/pubmed/15344454?dopt=Abstract