Analyzing spatio-temporal patterns of genuine cross-correlations

In multivariate time series analysis, the equal-time cross-correlation is a classic and computationally efficient measure for quantifying linear interrelations between data channels. When the cross-correlation coefficient is estimated using a finite amount of data points, its non-random part may be strongly contaminated by a sizable random contribution, such that no reliable conclusion can be drawn about genuine mutual interdependencies. The random correlations are determined by the signals' frequency content and the amount of data points used. Here, we introduce adjusted correlation matrices that can be employed to disentangle random from non-random contributions to each matrix element independently of the signal frequencies. Extending our previous work these matrices allow analyzing spatial patterns of genuine cross-correlation in multivariate data regardless of confounding influences. The performance is illustrated by example of model systems with known interdependence patterns. Finally, we apply the methods to electroencephalographic (EEG) data with epileptic seizure activity.

Analyzing Software Evolution through Feature Views

Features encapsulate the domain knowledge of a software system and thus are valuable sources of information for a reverse engineer. When analyzing the evolution of a system, we need to know how and which features were modified to recover both the change intention and its extent, namely which source artifacts are affected. Typically, the implementation of a feature crosscuts a number of source artifacts. To obtain a mapping between features to the source artifacts, we exercise the features and capture their execution traces. However this results in large traces that are difficult to interpret. To tackle this issue we compact the traces into simple sets of source artifacts that participate in a feature's runtime behavior. We refer to these compacted traces as feature views. Within a feature view, we partition the source artifacts into disjoint sets of characterized software entities. The characterization defines the level of participation of a source entity in the features. We then analyze the features over several versions of a system and we plot their evolution to reveal how and hich features were affected by changes in the code. We show the usefulness of our approach by applying it to a case study where we address the problem of merging parallel development tracks of the same system.

Analyzing, Capturing and Taming Software Change

Software systems need to continuously change to remain useful. Change appears in several forms and needs to be accommodated at different levels. We propose ChangeBoxes as a mechanism to encapsulate, manage, analyze and exploit changes to software systems. Our thesis is that only by making change explicit and manipulable can we enable the software developer to manage software change more effectively than is currently possible. Furthermore we argue that we need new insights into assessing the impact of changes and we need to provide new tools and techniques to manage them. We report on the results of some initial prototyping efforts, and we outline a series of research activities that we have started to explore the potential of ChangeBoxes.

A novel quantitative method for analyzing the distributions of nanoparticles between different tissue and intracellular compartments

The penetration, translocation, and distribution of ultrafine and nanoparticles in tissues and cells are challenging issues in aerosol research. This article describes a set of novel quantitative microscopic methods for evaluating particle distributions within sectional images of tissues and cells by addressing the following questions: (1) is the observed distribution of particles between spatial compartments random? (2) Which compartments are preferentially targeted by particles? and (3) Does the observed particle distribution shift between different experimental groups? Each of these questions can be addressed by testing an appropriate null hypothesis. The methods all require observed particle distributions to be estimated by counting the number of particles associated with each defined compartment. For studying preferential labeling of compartments, the size of each of the compartments must also be estimated by counting the number of points of a randomly superimposed test grid that hit the different compartments. The latter provides information about the particle distribution that would be expected if the particles were randomly distributed, that is, the expected number of particles. From these data, we can calculate a relative deposition index (RDI) by dividing the observed number of particles by the expected number of particles. The RDI indicates whether the observed number of particles corresponds to that predicted solely by compartment size (for which RDI = 1). Within one group, the observed and expected particle distributions are compared by chi-squared analysis. The total chi-squared value indicates whether an observed distribution is random. If not, the partial chi-squared values help to identify those compartments that are preferential targets of the particles (RDI > 1). Particle distributions between different groups can be compared in a similar way by contingency table analysis. We first describe the preconditions and the way to implement these methods, then provide three worked examples, and finally discuss the advantages, pitfalls, and limitations of this method.

The performance of computer-aided detection when analyzing prior mammograms of newly detected breast cancers with special focus on the time interval from initial imaging to detection

PURPOSE: The clinical role of CAD systems to detect breast cancer, which have not been on cancer containing mammograms not detected by the radiologist was proven retrospectively. METHODS: All patients from 1992 to 2005 with a histologically verified malignant breast lesion and a mammogram at our department, were analyzed in retrospect focussing on the time of detection of the malignant lesion. All prior mammograms were analyzed by CAD (CADx, USA). The resulting CAD printout was matched with the cancer containing images yielding to the radiological diagnosis of breast cancer. CAD performance, sensitivity as well as the association of CAD and radiological features were analyzed. RESULTS: 278 mammograms fulfilled the inclusion criteria. 111 cases showed a retrospectively visible lesion (71 masses, 23 single microcalcification clusters, 16 masses with microcalcifications, in one case two microcalcification clusters). 54/87 masses and 34/41 microcalcifications were detected by CAD. Detection rates varied from 9/20 (ACR 1) to 5/7 (ACR 4) (45% vs. 71%). The detection of microcalcifications was not influenced by breast tissue density. CONCLUSION: CAD might be useful in an earlier detection of subtle breast cancer cases, which might remain otherwise undetected.