Yangming Ou

Multiparametric Tissue Characterization of Brain Neoplasms and Their Recurrence Using Pattern Classification of MR Images

RATIONALE AND OBJECTIVES Treatment of brain neoplasms can greatly benefit from better delineation of bulk neoplasm boundary and the extent and degree of more subtle neoplastic infiltration. Magnetic resonance imaging (MRI) is the primary imaging modality for evaluation before and after therapy, typically combining conventional sequences with more advanced techniques such as perfusion-weighted imaging and diffusion tensor imaging (DTI). The purpose of this study is to quantify the multiparametric imaging profile of neoplasms by integrating structural MRI and DTI via statistical image analysis methods to potentially capture complex and subtle tissue characteristics that are not obvious from any individual image or parameter. MATERIALS AND METHODS Five structural MRI sequences, namely, B0, diffusion-weighted images, fluid-attenuated inversion recovery, T1-weighted, and gadolinium-enhanced T1-weighted, and two scalar maps computed from DTI (ie, fractional anisotropy and apparent diffusion coefficient) are used to create an intensity-based tissue profile. This is incorporated into a nonlinear pattern classification technique to create a multiparametric probabilistic tissue characterization, which is applied to data from 14 patients with newly diagnosed primary high-grade neoplasms who have not received any therapy before imaging. RESULTS Preliminary results demonstrate that this multiparametric tissue characterization helps to better differentiate among neoplasm, edema, and healthy tissue, and to identify tissue that is likely to progress to neoplasm in the future. This has been validated on expert assessed tissue. CONCLUSION This approach has potential applications in treatment, aiding computer-assisted surgery by determining the spatial distributions of healthy and neoplastic tissue, as well as in identifying tissue that is relatively more prone to tumor recurrence.

Evaluation of prostate segmentation algorithms for MRI: the PROMISE12 challenge

Prostate MRI image segmentation has been an area of intense research due to the increased use of MRI as a modality for the clinical workup of prostate cancer. Segmentation is useful for various tasks, e.g. to accurately localize prostate boundaries for radiotherapy or to initialize multi-modal registration algorithms. In the past, it has been difficult for research groups to evaluate prostate segmentation algorithms on multi-center, multi-vendor and multi-protocol data. Especially because we are dealing with MR images, image appearance, resolution and the presence of artifacts are affected by differences in scanners and/or protocols, which in turn can have a large influence on algorithm accuracy. The Prostate MR Image Segmentation (PROMISE12) challenge was setup to allow a fair and meaningful comparison of segmentation methods on the basis of performance and robustness. In this work we will discuss the initial results of the online PROMISE12 challenge, and the results obtained in the live challenge workshop hosted by the MICCAI2012 conference. In the challenge, 100 prostate MR cases from 4 different centers were included, with differences in scanner manufacturer, field strength and protocol. A total of 11 teams from academic research groups and industry participated. Algorithms showed a wide variety in methods and implementation, including active appearance models, atlas registration and level sets. Evaluation was performed using boundary and volume based metrics which were combined into a single score relating the metrics to human expert performance. The winners of the challenge where the algorithms by teams Imorphics and ScrAutoProstate, with scores of 85.72 and 84.29 overall. Both algorithms where significantly better than all other algorithms in the challenge (p<0.05) and had an efficient implementation with a run time of 8min and 3s per case respectively. Overall, active appearance model based approaches seemed to outperform other approaches like multi-atlas registration, both on accuracy and computation time. Although average algorithm performance was good to excellent and the Imorphics algorithm outperformed the second observer on average, we showed that algorithm combination might lead to further improvement, indicating that optimal performance for prostate segmentation is not yet obtained. All results are available online at http://promise12.grand-challenge.org/.

Multi-Atlas Skull-Stripping

RATIONALE AND OBJECTIVES We present a new method for automatic brain extraction on structural magnetic resonance images, based on a multi-atlas registration framework. MATERIALS AND METHODS Our method addresses fundamental challenges of multi-atlas approaches. To overcome the difficulties arising from the variability of imaging characteristics between studies, we propose a study-specific template selection strategy, by which we select a set of templates that best represent the anatomical variations within the data set. Against the difficulties of registering brain images with skull, we use a particularly adapted registration algorithm that is more robust to large variations between images, as it adaptively aligns different regions of the two images based not only on their similarity but also on the reliability of the matching between images. Finally, a spatially adaptive weighted voting strategy, which uses the ranking of Jacobian determinant values to measure the local similarity between the template and the target images, is applied for combining coregistered template masks. RESULTS The method is validated on three different public data sets and obtained a higher accuracy than recent state-of-the-art brain extraction methods. Also, the proposed method is successfully applied on several recent imaging studies, each containing thousands of magnetic resonance images, thus reducing the manual correction time significantly. CONCLUSIONS The new method, available as a stand-alone software package for public use, provides a robust and accurate brain extraction tool applicable for both clinical use and large population studies.

Integration and relative value of biomarkers for prediction of MCI to AD progression: Spatial patterns of brain atrophy, cognitive scores, APOE genotype and CSF biomarkers

This study evaluates the individual, as well as relative and joint value of indices obtained from magnetic resonance imaging (MRI) patterns of brain atrophy (quantified by the SPARE-AD index), cerebrospinal fluid (CSF) biomarkers, APOE genotype, and cognitive performance (ADAS-Cog) in progression from mild cognitive impairment (MCI) to Alzheimers disease (AD) within a variable follow-up period up to 6 years, using data from the Alzheimers Disease Neuroimaging Initiative-1 (ADNI-1). SPARE-AD was first established as a highly sensitive and specific MRI-marker of AD vs. cognitively normal (CN) subjects (AUC = 0.98). Baseline predictive values of all aforementioned indices were then compared using survival analysis on 381 MCI subjects. SPARE-AD and ADAS-Cog were found to have similar predictive value, and their combination was significantly better than their individual performance. APOE genotype did not significantly improve prediction, although the combination of SPARE-AD, ADAS-Cog and APOE ε4 provided the highest hazard ratio estimates of 17.8 (last vs. first quartile). In a subset of 192 MCI patients who also had CSF biomarkers, the addition of Aβ1–42, t-tau, and p-tau181p to the previous model did not improve predictive value significantly over SPARE-AD and ADAS-Cog combined. Importantly, in amyloid-negative patients with MCI, SPARE-AD had high predictive power of clinical progression. Our findings suggest that SPARE-AD and ADAS-Cog in combination offer the highest predictive power of conversion from MCI to AD, which is improved, albeit not significantly, by APOE genotype. The finding that SPARE-AD in amyloid-negative MCI patients was predictive of clinical progression is not expected under the amyloid hypothesis and merits further investigation.

MUSE: MUlti-atlas region Segmentation utilizing Ensembles of registration algorithms and parameters, and locally optimal atlas selection

Atlas-based automated anatomical labeling is a fundamental tool in medical image segmentation, as it defines regions of interest for subsequent analysis of structural and functional image data. The extensive investigation of multi-atlas warping and fusion techniques over the past 5 or more years has clearly demonstrated the advantages of consensus-based segmentation. However, the common approach is to use multiple atlases with a single registration method and parameter set, which is not necessarily optimal for every individual scan, anatomical region, and problem/data-type. Different registration criteria and parameter sets yield different solutions, each providing complementary information. Herein, we present a consensus labeling framework that generates a broad ensemble of labeled atlases in target image space via the use of several warping algorithms, regularization parameters, and atlases. The label fusion integrates two complementary sources of information: a local similarity ranking to select locally optimal atlases and a boundary modulation term to refine the segmentation consistently with the target images intensity profile. The ensemble approach consistently outperforms segmentations using individual warping methods alone, achieving high accuracy on several benchmark datasets. The MUSE methodology has been used for processing thousands of scans from various datasets, producing robust and consistent results. MUSE is publicly available both as a downloadable software package, and as an application that can be run on the CBICA Image Processing Portal (https://ipp.cbica.upenn.edu), a web based platform for remote processing of medical images.

Non-rigid registration between histological and MR images of the prostate: A joint segmentation and registration framework

This paper presents a 3D non-rigid registration algorithm between histological and MR images of the prostate with cancer. To compensate for the loss of 3D integrity in the histology sectioning process, series of 2D histological slices are first reconstructed into a 3D histological volume. After that, the 3D histology-MRI registration is obtained by maximizing a) landmark similarity and b) cancer region overlap between the two images. The former aims to capture distortions at prostate boundary and internal blob-like structures; and the latter aims to capture distortions specifically at cancer regions. In particular, landmark similarities, the former, is maximized by an annealing process, where correspondences between the automatically-detected boundary and internal landmarks are iteratively established in a fuzzy-to-deterministic fashion. Cancer region overlap, the latter, is maximized in a joint cancer segmentation and registration framework, where the two interleaved problems - segmentation and registration - inform each other in an iterative fashion. Registration accuracy is established by comparing against human-rater-defined landmarks and by comparing with other methods. The ultimate goal of this registration is to warp the histologically-defined cancer ground truth into MRI, for more thoroughly understanding MRI signal characteristics of the prostate cancerous tissue, which will promote the MRI-based prostate cancer diagnosis in the future studies.

Neuroanatomical Classification in a Population-Based Sample of Psychotic Major Depression and Bipolar I Disorder with 1 Year of Diagnostic Stability

The presence of psychotic features in the course of a depressive disorder is known to increase the risk for bipolarity, but the early identification of such cases remains challenging in clinical practice. In the present study, we evaluated the diagnostic performance of a neuroanatomical pattern classification method in the discrimination between psychotic major depressive disorder (MDD), bipolar I disorder (BD-I), and healthy controls (HC) using a homogenous sample of patients at an early course of their illness. Twenty-three cases of first-episode psychotic mania (BD-I) and 19 individuals with a first episode of psychotic MDD whose diagnosis remained stable during 1 year of followup underwent 1.5 T MRI at baseline. A previously validated multivariate classifier based on support vector machine (SVM) was employed and measures of diagnostic performance were obtained for the discrimination between each diagnostic group and subsamples of age- and gender-matched controls recruited in the same neighborhood of the patients. Based on T1-weighted images only, the SVM-classifier afforded poor discrimination in all 3 pairwise comparisons: BD-I versus HC; MDD versus HC; and BD-I versus MDD. Thus, at the population level and using structural MRI only, we failed to achieve good discrimination between BD-I, psychotic MDD, and HC in this proof of concept study.

Simultaneous geometric - iconic registration

In this paper, we introduce a novel approach to bridge the gap between the landmark-based and the iconic-based voxel-wise registration methods. The registration problem is formulated with the use of Markov Random Field theory resulting in a discrete objective function consisting of thee parts. The first part of the energy accounts for the iconic-based volumetric registration problem while the second one for establishing geometrically meaningful correspondences by optimizing over a set of automatically generated mutually salient candidate pairs of points. The last part of the energy penalizes locally the difference between the dense deformation field due to the iconic-based registration and the implied displacements due to the obtained correspondences. Promising results in real MR brain data demonstrate the potentials of our approach.

Sampling the Spatial Patterns of Cancer: Optimized Biopsy Procedures for Estimating Prostate Cancer Volume and Gleason Score

Prostate biopsy is the current gold-standard procedure for prostate cancer diagnosis. Existing prostate biopsy procedures have been mostly focusing on detecting cancer presence. However, they often ignore the potential use of biopsy to estimate cancer volume (CV) and Gleason Score (GS, a cancer grade descriptor), the two surrogate markers for cancer aggressiveness and the two crucial factors for treatment planning. To fill up this vacancy, this paper assumes and demonstrates that, by optimally sampling the spatial patterns of cancer, biopsy procedures can be specifically designed for estimating CV and GS. Our approach combines image analysis and machine learning tools in an atlas-based population study that consists of three steps. First, the spatial distributions of cancer in a patient population are learned, by constructing statistical atlases from histological images of prostate specimens with known cancer ground truths. Then, the optimal biopsy locations are determined in a feature selection formulation, so that biopsy outcomes (either cancer presence or absence) at those locations could be used to differentiate, at the best rate, between the existing specimens having different (high vs. low) CV/GS values. Finally, the optimized biopsy locations are utilized to estimate whether a new-coming prostate cancer patient has high or low CV/GS values, based on a binary classification formulation. The estimation accuracy and the generalization ability are evaluated by the classification rates and the associated receiver-operating-characteristic (ROC) curves in cross validations. The optimized biopsy procedures are also designed to be robust to the almost inevitable needle displacement errors in clinical practice, and are found to be robust to variations in the optimization parameters as well as the training populations.

PROBABILISTIC SEGMENTATION OF BRAIN TUMORS BASED ON MULTI-MODALITY MAGNETIC RESONANCE IMAGES

In this paper, multi-modal magnetic resonance (MR) images are integrated into a tissue profile that aims at differentiating tumor components, edema and normal tissue. This is achieved by a tissue classification technique that learns the appearance models of different tissue types based on training samples identified by an expert and assigns tissue labels to each voxel. These tissue classifiers produce probabilistic tissue maps reflecting imaging characteristics of tumors and surrounding tissues that may be employed to aid in diagnosis, tumor boundary delineation, surgery and treatment planning. The main contributions of this work are: 1) conventional structural MR modalities are combined with diffusion tensor imaging data to create an integrated multimodality profile for brain tumors, and 2) in addition to the tumor components of enhancing and non-enhancing tumor types, edema is also characterized as a separate class in our framework. Classification performance is tested on 22 diverse tumor cases using cross-validation.