Ju Han

Iterative Voting for Inference of Structural Saliency and Characterization of Subcellular Events

Saliency is an important perceptual cue that occurs at different levels of resolution. Important attributes of saliency are symmetry, continuity, and closure. Detection of these attributes is often hindered by noise, variation in scale, and incomplete information. This paper introduces the iterative voting method, which uses oriented kernels for inferring saliency as it relates to symmetry. A unique aspect of the technique is the kernel topography, which is refined and reoriented iteratively. The technique can cluster and group nonconvex perceptual circular symmetries along the radial line of an objects shape. It has an excellent noise immunity and is shown to be tolerant to perturbation in scale. The application of this technique to images obtained through various modes of microscopy is demonstrated. Furthermore, as a case example, the method has been applied to quantify kinetics of nuclear foci formation that are formed by phosphorylation of histone gammaH2AX following ionizing radiation. Iterative voting has been implemented in both 2-D and 3-D for multi image analysis

Molecular Predictors of 3D Morphogenesis by Breast Cancer Cell Lines in 3D Culture

Correlative analysis of molecular markers with phenotypic signatures is the simplest model for hypothesis generation. In this paper, a panel of 24 breast cell lines was grown in 3D culture, their morphology was imaged through phase contrast microscopy, and computational methods were developed to segment and represent each colony at multiple dimensions. Subsequently, subpopulations from these morphological responses were identified through consensus clustering to reveal three clusters of round, grape-like, and stellate phenotypes. In some cases, cell lines with particular pathobiological phenotypes clustered together (e.g., ERBB2 amplified cell lines sharing the same morphometric properties as the grape-like phenotype). Next, associations with molecular features were realized through (i) differential analysis within each morphological cluster, and (ii) regression analysis across the entire panel of cell lines. In both cases, the dominant genes that are predictive of the morphological signatures were identified. Specifically, PPARγ has been associated with the invasive stellate morphological phenotype, which corresponds to triple-negative pathobiology. PPARγ has been validated through two supporting biological assays.

Invariant Delineation of Nuclear Architecture in Glioblastoma Multiforme for Clinical and Molecular Association

Automated analysis of whole mount tissue sections can provide insights into tumor subtypes and the underlying molecular basis of neoplasm. However, since tumor sections are collected from different laboratories, inherent technical and biological variations impede analysis for very large datasets such as The Cancer Genome Atlas (TCGA). Our objective is to characterize tumor histopathology, through the delineation of the nuclear regions, from hematoxylin and eosin (H&E) stained tissue sections. Such a representation can then be mined for intrinsic subtypes across a large dataset for prediction and molecular association. Furthermore, nuclear segmentation is formulated within a multi-reference graph framework with geodesic constraints, which enables computation of multidimensional representations, on a cell-by-cell basis, for functional enrichment and bioinformatics analysis. Here, we present a novel method, multi-reference graph cut (MRGC), for nuclear segmentation that overcomes technical variations associated with sample preparation by incorporating prior knowledge from manually annotated reference images and local image features. The proposed approach has been validated on manually annotated samples and then applied to a dataset of 377 Glioblastoma Multiforme (GBM) whole slide images from 146 patients. For the GBM cohort, multidimensional representation of the nuclear features and their organization have identified 1) statistically significant subtypes based on several morphometric indexes, 2) whether each subtype can be predictive or not, and 3) that the molecular correlates of predictive subtypes are consistent with the literature. Data and intermediaries for a number of tumor types (GBM, low grade glial, and kidney renal clear carcinoma) are available at: http://tcga.lbl.gov for correlation with TCGA molecular data. The website also provides an interface for panning and zooming of whole mount tissue sections with/without overlaid segmentation results for quality control.

Morphometic analysis of TCGA glioblastoma multiforme

BackgroundOur goals are to develop a computational histopathology pipeline for characterizing tumor types that are being generated by The Cancer Genome Atlas (TCGA) for genomic association. TCGA is a national collaborative program where different tumor types are being collected, and each tumor is being characterized using a variety of genome-wide platforms. Here, we have developed a tumor-centric analytical pipeline to process tissue sections stained with hematoxylin and eosin (H&E) for visualization and cell-by-cell quantitative analysis. Thus far, analysis is limited to Glioblastoma Multiforme (GBM) and kidney renal clear cell carcinoma tissue sections. The final results are being distributed for subtyping and linking the histology sections to the genomic data.ResultsA computational pipeline has been designed to continuously update a local image database, with limited clinical information, from an NIH repository. Each image is partitioned into blocks, where each cell in the block is characterized through a multidimensional representation (e.g., nuclear size, cellularity). A subset of morphometric indices, representing potential underlying biological processes, can then be selected for subtyping and genomic association. Simultaneously, these subtypes can also be predictive of the outcome as a result of clinical treatments. Using the cellularity index and nuclear size, the computational pipeline has revealed five subtypes, and one subtype, corresponding to the extreme high cellularity, has shown to be a predictor of survival as a result of a more aggressive therapeutic regime. Further association of this subtype with the corresponding gene expression data has identified enrichment of (i) the immune response and AP-1 signaling pathways, and (ii) IFNG, TGFB1, PKC, Cytokine, and MAPK14 hubs.ConclusionWhile subtyping is often performed with genome-wide molecular data, we have shown that it can also be applied to categorizing histology sections. Accordingly, we have identified a subtype that is a predictor of the outcome as a result of a therapeutic regime. Computed representation has become publicly available through our Web site.

Learning invariant features of tumor signatures

We present a novel method for automated learning of features from unlabeled image patches for classification of tumor architecture. In contrast to previous manually-designed feature detectors (e.g., Gabor basis function), the proposed method utilizes inexpensive un-labeled data to construct features. The algorithm, also known as reconstruction independent subspace analysis, can be described as a two-layer network with non-linear responses, where the second layer represents subspace structures. The technique is applied to tissue sections for characterizing necrosis, apoptotic, and viable regions of Glioblastoma Multifrome (GBM) from TCGA dataset. Experimental results show that this method outperforms more complex expert-designed approaches. The fact that our approach learns features automatically from unlabeled data promises a wider application of self-learning strategies for tissue characterization.

Multidimensional Profiling of Cell Surface Proteins and Nuclear Markers

Cell membrane proteins play an important role in tissue architecture and cell-cell communication. We hypothesize that segmentation and multidimensional characterization of the distribution of cell membrane proteins, on a cell-by-cell basis, enable improved classification of treatment groups and identify important characteristics that can otherwise be hidden. We have developed a series of computational steps to 1) delineate cell membrane protein signals and associate them with a specific nucleus; 2) compute a coupled representation of the multiplexed DNA content with membrane proteins; 3) rank computed features associated with such a multidimensional representation; 4) visualize selected features for comparative evaluation through heatmaps; and 5) discriminate between treatment groups in an optimal fashion. The novelty of our method is in the segmentation of the membrane signal and the multidimensional representation of phenotypic signature on a cell-by-cell basis. To test the utility of this method, the proposed computational steps were applied to images of cells that have been irradiated with different radiation qualities in the presence and absence of other small molecules. These samples are labeled for their DNA content and E-cadherin membrane proteins. We demonstrate that multidimensional representations of cell-by-cell phenotypes improve predictive and visualization capabilities among different treatment groups, and identify hidden variables.

Genetic Differences in Transcript Responses to Low-Dose Ionizing Radiation Identify Tissue Functions Associated with Breast Cancer Susceptibility

High dose ionizing radiation (IR) is a well-known risk factor for breast cancer but the health effects after low-dose (LD, <10 cGy) exposures remain highly uncertain. We explored a systems approach that compared LD-induced chromosome damage and transcriptional responses in strains of mice with genetic differences in their sensitivity to radiation-induced mammary cancer (BALB/c and C57BL/6) for the purpose of identifying mechanisms of mammary cancer susceptibility. Unirradiated mammary and blood tissues of these strains differed significantly in baseline expressions of DNA repair, tumor suppressor, and stress response genes. LD exposures of 7.5 cGy (weekly for 4 weeks) did not induce detectable genomic instability in either strain. However, the mammary glands of the sensitive strain but not the resistant strain showed early transcriptional responses involving: (a) diminished immune response, (b) increased cellular stress, (c) altered TGFβ-signaling, and (d) inappropriate expression of developmental genes. One month after LD exposure, the two strains showed opposing responses in transcriptional signatures linked to proliferation, senescence, and microenvironment functions. We also discovered a pre-exposure expression signature in both blood and mammary tissues that is predictive for poor survival among human cancer patients (p = 0.0001), and a post-LD-exposure signature also predictive for poor patient survival (p<0.0001). There is concordant direction of expression in the LD-exposed sensitive mouse strain, in biomarkers of human DCIS and in biomarkers of human breast tumors. Our findings support the hypothesis that genetic mechanisms that determine susceptibility to LD radiation induced mammary cancer in mice are similar to the tissue mechanisms that determine poor-survival in breast cancer patients. We observed non-linearity of the LD responses providing molecular evidence against the LNT risk model and obtained new evidence that LD responses are strongly influenced by genotype. Our findings suggest that the biological assumptions concerning the mechanisms by which LD radiation is translated into breast cancer risk should be reexamined and suggest a new strategy to identify genetic features that predispose or protect individuals from LD-induced breast cancer.

Multireference Level Set for the Characterization of Nuclear Morphology in Glioblastoma Multiforme

Histological tissue sections provide rich information and continue to be the gold standard for the assessment of tissue neoplasm. However, there are a significant amount of technical and biological variations that impede analysis of large histological datasets. In this paper, we have proposed a novel approach for nuclear segmentation in tumor histology sections, which addresses the problem of technical and biological variations by incorporating information from both manually annotated reference patches and the original image. Subsequently, the solution is formulated within a multireference level set framework. This approach has been validated on manually annotated samples and then applied to the TCGA glioblastoma multiforme (GBM) dataset consisting of 440 whole mount tissue sections scanned with either a 20× or 40 × objective, in which, each tissue section varies in size from 40k × 40k pixels to 100k × 100k pixels. Experimental results show a superior performance of the proposed method in comparison with present state of art techniques.

Comparison of sparse coding and kernel methods for histopathological classification of gliobastoma multiforme

This paper compares the performance of redundant representation and sparse coding against classical kernel methods for classifying histological sections. Sparse coding has been proven an effective technique for restoration, and has recently been extended to classification. The main issue with histology sections classification is inherent heterogeneity, which is a result of technical and biological variations. Technical variations originate from sample preparation, fixation, and staining from multiple laboratories, whereas biological variations originate from tissue content. Image patches are represented with invariant features at local and global scales, where local refers to responses measured with Laplacian of Gaussians, and global refers to measurements in the color space. Experiments are designed to learn dictionaries through sparse coding, and to train classifiers through kernel methods using normal, necrotic, apoptotic, and tumor regions with characteristics of high cellularity. Two different kernel methods, that of a support vector machine (SVM) and a kernel discriminant analysis (KDA), were used for comparative analysis. Preliminary investigation on the histological samples of Glioblastoma multiforme (GBM) indicates the kernel methods perform as good, if not better, than sparse coding with redundant representation.

Multiscale iterative voting for differential analysis of stress response for 2D and 3D cell culture models.

Three‐dimensional (2D) cell culture models have emerged as the basis for improved cell systems biology. However, there is a gap in robust computational techniques for segmentation of these model systems that are imaged through confocal or deconvolution microscopy. The main issues are the volume of data, overlapping subcellular compartments and variation in scale or size of subcompartments of interest, which lead to ambiguities for quantitative analysis on a cell‐by‐cell basis. We address these ambiguities through a series of geometric operations that constrain the problem through iterative voting and decomposition strategies. The main contributions of this paper are to (i) extend the previously developed 2D radial voting to an efficient 3D implementation, (ii) demonstrate application of iterative radial voting at multiple subcellular and molecular scales, and (iii) investigate application of the proposed technology to two endpoints between 2D and 3D cell culture models. These endpoints correspond to kinetics of DNA damage repair as measured by phosphorylation of γH2AX, and the loss of the membrane‐bound E‐cadherin protein as a result of ionizing radiation.