Rainer Stotzka

Model-based registration of X-ray mammograms and MR images of the female breast

We present a new approach for automatic registration of X-ray mammograms and MR images. Multimodal breast cancer diagnosis is supported by automatic localization of small lesions, which are only visible in the mammograms or in the MR image. To cope with the huge deformation of the breast during mammography, a finite element model of the deformable behavior of the breast is applied during the registration. An evaluation of the registration with six clinical data sets resulted in an accurate localization with a mean displacement of 4.3 mm (/spl plusmn/1 mm) and 3.9 mm (/spl plusmn/1.7 mm) for predicting the lesion position in mammograms and in the MR images, respectively.

Automated Processing of Zebrafish Imaging Data: A Survey

Due to the relative transparency of its embryos and larvae, the zebrafish is an ideal model organism for bioimaging approaches in vertebrates. Novel microscope technologies allow the imaging of developmental processes in unprecedented detail, and they enable the use of complex image-based read-outs for high-throughput/high-content screening. Such applications can easily generate Terabytes of image data, the handling and analysis of which becomes a major bottleneck in extracting the targeted information. Here, we describe the current state of the art in computational image analysis in the zebrafish system. We discuss the challenges encountered when handling high-content image data, especially with regard to data quality, annotation, and storage. We survey methods for preprocessing image data for further analysis, and describe selected examples of automated image analysis, including the tracking of cells during embryogenesis, heartbeat detection, identification of dead embryos, recognition of tissues and anatomical landmarks, and quantification of behavioral patterns of adult fish. We review recent examples for applications using such methods, such as the comprehensive analysis of cell lineages during early development, the generation of a three-dimensional brain atlas of zebrafish larvae, and high-throughput drug screens based on movement patterns. Finally, we identify future challenges for the zebrafish image analysis community, notably those concerning the compatibility of algorithms and data formats for the assembly of modular analysis pipelines.

2J-2 3D Ultrasound Computer Tomography: Results with a Clinical Breast Phantom

Ultrasound computer tomography (USCT) is an imaging method capable of producing volume images with sub-millimeter resolution and high image quality. The long term goal of the system under construction is 3D imaging for early breast cancer diagnosis. In this paper the first images of a clinical breast phantom are presented and discussed in respect to further modifications of our system

An Example-Based System to Support the Segmentation of Stellate Lesions

The detection and segmentation of stellate lesions in mammograms is a difficult task in image processing due to the high variances in their appearance. We present the application of an interactive generic system, that is trained to detect and segment stellate lesions based on their local features. The training is done by an expert presenting examples of stellate lesions to the system. With the data available good detection results are achieved, yet the performance of the system can be increased as more examples are presented.

AUTOMATIC IMAGE MATCHING FOR BREAST CANCER DIAGNOSTICS BY A 3D DEFORMATION MODEL OF THE MAMMA

X-ray mammograms and MR volumes provide complementary information for early breast cancer diagnosis. The breast is deformed during mammography, therefore the images can not be compared directly. A registration algorithm is investigated to fuse the images automatically. A finite element simulation was applied to a MR image of an underformed breast and compared to a compressed breast using different tissue models and boundary conditions. Based on the results a set of patient data was registered. To archive the requested accuracy distinguishing between the different tissue types of the breast was not necessary. A linear elastic model was sufficient. It was possible to simulate the deformation with an average deviation of approximately of the size of a voxel in the MRI data and retrieve the position of a lesion with an error of 3.8 mm in the patient data.

An ensemble-averaged, cell density-based digital model of zebrafish embryo development derived from light-sheet microscopy data with single-cell resolution

A new era in developmental biology has been ushered in by recent advances in the quantitative imaging of all-cell morphogenesis in living organisms. Here we have developed a light-sheet fluorescence microscopy-based framework with single-cell resolution for identification and characterization of subtle phenotypical changes of millimeter-sized organisms. Such a comparative study requires analyses of entire ensembles to be able to distinguish sample-to-sample variations from definitive phenotypical changes. We present a kinetic digital model of zebrafish embryos up to 16 h of development. The model is based on the precise overlay and averaging of data taken on multiple individuals and describes the cell density and its migration direction at every point in time. Quantitative metrics for multi-sample comparative studies have been introduced to analyze developmental variations within the ensemble. The digital model may serve as a canvas on which the behavior of cellular subpopulations can be studied. As an example, we have investigated cellular rearrangements during germ layer formation at the onset of gastrulation. A comparison of the one-eyed pinhead (oep) mutant with the digital model of the wild-type embryo reveals its abnormal development at the onset of gastrulation, many hours before changes are obvious to the eye.

The Large Scale Data Facility: Data Intensive Computing for Scientific Experiments

The Large Scale Data Facility (LSDF) at the Karlsruhe Institute of Technology was started end of 2009 with the aim of supporting the growing requirements of data intensive experiments. In close cooperation with the involved scientific communities, the LSDF provides them not only with adequate storage space but with a directly attached analysis farm and -- more importantly -- with value added services for their big scientific data-sets. Analysis workflows are supported through the mixed Hadoop and Open Nebula Cloud environments directly attached to the storage, and enable the efficient processing of the experimental data. Metadata handling is a central part of the LSDF, where a metadata repository, community specific metadata schemes, graphical tools, and APIs were developed for accessing and efficiently organizing the stored data-sets.

Medical imaging by ultrasound computer tomography

In breast cancer diagnosis, ultrasound examination provides useful additional diagnostic information. Moreover ultrasound does not harm biological tissue and can be applied frequently. But conventional ultrasound imaging methods lack both high spatial and temporal resolution. Usually, the scanner is operated manually and the tissue is deformed while getting as close as possible to the regions of interest. Therefore, image contents and image quality depend strongly on the operator. Exact measurement of tissue structures, like tumor size, is not possible. Instead of a manually controlled linear transducer array, we use ultrasound computer tomography (USCT) to image a volume directly. A few thousand ultrasound transducers are arranged in a cylindrical array around a tank containing the object to be examined coupled by water. Every single transducer is small enough to emit an almost spherical sound wave. While one transducer is transmitting, all others receive simultaneously. Afterwards a different transducer emits the next pulse. For volume reconstruction every transmitted, scattered and reflected signal is used. This new method allows reproducible image sequences with enhanced spatial and temporal resolution. For the benefit of more reconstructed 3D images per second, spatial resolution may be reduced offline. First tests with our prototype in a ring-shaped geometry have even showed nylon threads (0.4 mm) and an image quality superior to clinical ultrasound scanners.

Perspective of the Large Scale Data Facility (LSDF) Supporting Nuclear Fusion Applications

To cope with the growing requirements of data intensive scientific experiments, models and simulations the Large Scale Data Facility(LSDF) at KIT aims to support many scientific disciplines. The LSDFis a distributed storage facility at Exabyte scale providing storage, archives, data bases and meta data repositories. Open interfaces and APIs support a variety of access methods to the highly available services for high throughput data applications. Tools for an easy and transparent access allow scientists to use the LSDF without bothering with the internal structures and technologies. In close cooperation with the scientific communities the LSDF provides assistance to efficiently organize data and metadata structures, and develops and deploys community specific software on the directly connected computing infrastructure.

TV-based conjugate gradient method and discrete L-curve for few-view CT reconstruction of X-ray in vivo data

High-resolution, three-dimensional (3D) imaging of soft tissues requires the solution of two inverse problems: phase retrieval and the reconstruction of the 3D image from a tomographic stack of two-dimensional (2D) projections. The number of projections per stack should be small to accommodate fast tomography of rapid processes and to constrain X-ray radiation dose to optimal levels to either increase the duration of in vivo time-lapse series at a given goal for spatial resolution and/or the conservation of structure under X-ray irradiation. In pursuing the 3D reconstruction problem in the sense of compressive sampling theory, we propose to reduce the number of projections by applying an advanced algebraic technique subject to the minimisation of the total variation (TV) in the reconstructed slice. This problem is formulated in a Lagrangian multiplier fashion with the parameter value determined by appealing to a discrete L-curve in conjunction with a conjugate gradient method. The usefulness of this reconstruction modality is demonstrated for simulated and in vivo data, the latter acquired in parallel-beam imaging experiments using synchrotron radiation.