Lee A. D. Cooper

Feature-based registration of histopathology images with different stains: An application for computerized follicular lymphoma prognosis

Follicular lymphoma (FL) is the second most common type of non-Hodgkins lymphoma. Manual histological grading of FL is subject to remarkable inter- and intra-reader variations. A promising approach to grading is the development of a computer-assisted system that improves consistency and precision. Correlating information from adjacent slides with different stain types requires establishing spatial correspondences between the digitized section pair through a precise non-rigid image registration. However, the dissimilar appearances of the different stain types challenges existing registration methods. This study proposes a method for the automatic non-rigid registration of histological section images with different stain types. This method is based on matching high level features that are representative of small anatomical structures. This choice of feature provides a rich matching environment, but also results in a high mismatch probability. Matching confidence is increased by establishing local groups of coherent features through geometric reasoning. The proposed method is validated on a set of FL images representing different disease stages. Statistical analysis demonstrates that given a proper feature set the accuracy of automatic registration is comparable to manual registration.

Non-rigid Registration for Large Sets of Microscopic Images on Graphics Processors

Microscopic imaging is an important tool for characterizing tissue morphology and pathology. 3D reconstruction and visualization of large sample tissue structure requires registration of large sets of high-resolution images. However, the scale of this problem presents a challenge for automatic registration methods. In this paper we present a novel method for efficient automatic registration using graphics processing units (GPUs) and parallel programming. Comparing a C++ CPU implementation with Compute Unified Device Architecture (CUDA) libraries and pthreads running on GPU we achieve a speed-up factor of up to 4.11× with a single GPU and 6.68× with a GPU pair. We present execution times for a benchmark composed of two sets of large-scale images: mouse placenta (16K ×16K pixels) and breast cancer tumors (23K ×62K pixels). It takes more than 12xa0hours for the genetic case in C++ to register a typical sample composed of 500 consecutive slides, which was reduced to less than 2xa0hours using two GPUs, in addition to a very promising scalability for extending those gains easily on a large number of GPUs in a distributed system.

Spatial segmentation of temporal texture using mixture linear models

In this paper we propose a novel approach for the spatial segmentation of video sequences containing multiple temporal textures. This work is based on the notion that a single temporal texture can be represented by a low-dimensional linear model. For scenes containing multiple temporal textures, e.g. trees swaying adjacent a flowing river, we extend the single linear model to a mixture of linear models and segment the scene by identifying subspaces within the data using robust generalized principal component analysis (GPCA). Computation is reduced to minutes in Matlab by first identifying models from a sampling of the sequence and using the derived models to segment the remaining data. The effectiveness of our method has been demonstrated in several examples including an application in biomedical image analysis.

Fast Automatic Registration Algorithm for Large Microscopy Images

In this paper, a framework of fast registration algorithm for large microscopy images is presented. The rationale behind this approach is that the rigid transform gives the global mapping between the two images while the nonrigid components further refines the local matching of the pixels by taking care of local nonrigid distortion and variation. Therefore, to estimate rigid transform, the global features such as specific anatomical structures need to be used instead of point features which does not contain any global information. Then to estimate local nonrigid transform, the local features such as points are used. The algorithm is divided into two stages: the first stage is to find an accurate estimate of the rigid (Euclidean) transform between the two images. To achieve this goal, high level (global) features such as small regions with anatomical meanings such as clusters of cells or blood vessels are exploited for matching purposes. A voting scheme is used to confirm the matching and compute the rigid transformation between two consecutive images. This then transforms the foundation for the second stage of nonrigid registration. Using the accurate estimate of the rigid transform, a large number of point feature correspondence is established and used as control points for the nonrigid transform

An imaging workflow for characterizing phenotypical change in large histological mouse model datasets

MOTIVATIONnThis paper presents a workflow designed to quantitatively characterize the 3D structural attributes of macroscopic tissue specimens acquired at a micron level resolution using light microscopy. The specific application is a study of the morphological change in a mouse placenta induced by knocking out the retinoblastoma gene.nnnRESULTnThis workflow includes four major components: (i) serial section image acquisition, (ii) image preprocessing, (iii) image analysis involving 2D pair-wise registration, 2D segmentation and 3D reconstruction, and (iv) visualization and quantification of phenotyping parameters. Several new algorithms have been developed within each workflow component. The results confirm the hypotheses that (i) the volume of labyrinth tissue decreases in mutant mice with the retinoblastoma (Rb) gene knockout and (ii) there is more interdigitation at the surface between the labyrinth and spongiotrophoblast tissues in mutant placenta. Additional confidence stem from agreement in the 3D visualization and the quantitative results generated.nnnAVAILABILITYnThe source code is available upon request.

Multi-faceted computational assessment of risk and progression in oligodendroglioma implicates NOTCH and PI3K pathways

Oligodendrogliomas are diffusely infiltrative gliomas defined by IDH-mutation and co-deletion of 1p/19q. They have highly variable clinical courses, with survivals ranging from 6 months to over 20 years, but little is known regarding the pathways involved with their progression or optimal markers for stratifying risk. We utilized machine-learning approaches with genomic data from The Cancer Genome Atlas to objectively identify molecular factors associated with clinical outcomes of oligodendroglioma and extended these findings to study signaling pathways implicated in oncogenesis and clinical endpoints associated with glioma progression. Our multi-faceted computational approach uncovered key genetic alterations associated with disease progression and shorter survival in oligodendroglioma and specifically identified Notch pathway inactivation and PI3K pathway activation as the most strongly associated with MRI and pathology findings of advanced disease and poor clinical outcome. Our findings that Notch pathway inactivation and PI3K pathway activation are associated with advanced disease and survival risk will pave the way for clinically relevant markers of disease progression and therapeutic targets to improve clinical outcomes. Furthermore, our approach demonstrates the strength of machine learning and computational methods for identifying genetic events critical to disease progression in the era of big data and precision medicine.Brain cancer: AI reveals pathways implicated in oligodendrogliomaUsing artificial intelligence, researchers identified two key signaling pathways involved in the progression of a brain cancer known as oligodendroglioma. A team led by Daniel Brat from Northwestern University Feinberg School of Medicine in Chicago, Illinois, and Lee Cooper from Emory University in Atlanta, Georgia, took a machine-learning approach to find links between genomic records, radiographic brain imaging data and digitized pathology slides from patients with oligodendroglioma, a type of tumor that develops from brain cells known as an oligodendrocytes. Their computational model identified an association between worse clinical outcomes and genetic alterations that either inactivate the Notch signaling pathway or activate the phosphoinositide 3-kinase pathway. The findings could yield potential therapeutic targets or prognostic biomarkers, although more work is needed to elucidate the exact proteins in these two pathways that drive disease progression.

Informatics Approaches to Address New Challenges in the Classification of Lymphoid Malignancies

PurposenLymphoid malignancies are remarkably heterogeneous, with variations in outcomes and clinical, biologic, and histologic presentation complicating classification according to the World Health Organization guidelines. Incorrect classification of lymphoid neoplasms can result in suboptimal therapeutic strategies for individual patients and confound the interpretation of clinical trials involving personalized, class-based treatments. This review discusses the potential role of pathology informatics in improving the classification accuracy and objectivity for lymphoid malignancies.nnnDesignnWe identified peer-reviewed publications examining pathology informatics approaches for the classification of lymphoid malignancies, reviewed developments in the lymphoma classification systems, and summarized computational methods for pathologic assessment that can impact practice.nnnResultsnComputer-assisted pathology image analysis algorithms in lymphoma most commonly have been applied to follicular lymphoma to address biologic heterogeneity and subjectivity in the process of classification.nnnConclusionnObjective methods are available to assist pathologists in lymphoma classification and grading, and have been demonstrated to provide measurable benefits in specific contexts. Future validation and extension of these approaches will require datasets that link high resolution pathology images available for image analysis algorithms with clinical variables and follow up outcomes.