Jacob Schack Vestergaard

Assessment of algorithms for mitosis detection in breast cancer histopathology images.

The proliferative activity of breast tumors, which is routinely estimated by counting of mitotic figures in hematoxylin and eosin stained histology sections, is considered to be one of the most important prognostic markers. However, mitosis counting is laborious, subjective and may suffer from low inter-observer agreement. With the wider acceptance of whole slide images in pathology labs, automatic image analysis has been proposed as a potential solution for these issues. In this paper, the results from the Assessment of Mitosis Detection Algorithms 2013 (AMIDA13) challenge are described. The challenge was based on a data set consisting of 12 training and 11 testing subjects, with more than one thousand annotated mitotic figures by multiple observers. Short descriptions and results from the evaluation of eleven methods are presented. The top performing method has an error rate that is comparable to the inter-observer agreement among pathologists.

HEp-2 Cell Classification Using Shape Index Histograms With Donut-Shaped Spatial Pooling

We present a new method for automatic classification of indirect immunoflourescence images of HEp-2 cells into different staining pattern classes. Our method is based on a new texture measure called shape index histograms that captures second-order image structure at multiple scales. Moreover, we introduce a spatial decomposition scheme which is radially symmetric and suitable for cell images. The spatial decomposition is performed using donut-shaped pooling regions of varying sizes when gathering histogram contributions. We evaluate our method using both the ICIP 2013 and the ICPR 2012 competition datasets. Our results show that shape index histograms are superior to other popular texture descriptors for HEp-2 cell classification. Moreover, when comparing to other automated systems for HEp-2 cell classification we show that shape index histograms are very competitive; especially considering the relatively low complexity of the method.

Reproductive isolation related to mimetic divergence in the poison frog Ranitomeya imitator

In a mimetic radiation--when a single species evolves to resemble different model species--mimicry can drive within-species morphological diversification, and, potentially, speciation. While mimetic radiations have occurred in a variety of taxa, their role in speciation remains poorly understood. We study the Peruvian poison frog Ranitomeya imitator, a species that has undergone a mimetic radiation into four distinct morphs. Using a combination of colour-pattern analysis, landscape genetics and mate-choice experiments, we show that a mimetic shift in R. imitator is associated with a narrow phenotypic transition zone, neutral genetic divergence and assortative mating, suggesting that divergent selection to resemble different model species has led to a breakdown in gene flow between these two populations. These results extend the effects of mimicry on speciation into a vertebrate system and characterize an early stage of speciation where reproductive isolation between mimetic morphs is incomplete but evident.

Mimetic Divergence and the Speciation Continuum in the Mimic Poison Frog Ranitomeya imitator.

While divergent ecological adaptation can drive speciation, understanding the factors that facilitate or constrain this process remains a major goal in speciation research. Here, we study two mimetic transition zones in the poison frog Ranitomeya imitator, a species that has undergone a Müllerian mimetic radiation to establish four morphs in Peru. We find that mimetic morphs are strongly phenotypically differentiated, producing geographic clines with varying widths. However, distinct morphs show little neutral genetic divergence, and landscape genetic analyses implicate isolation by distance as the primary determinant of among-population genetic differentiation. Mate choice experiments suggest random mating at the transition zones, although certain allopatric populations show a preference for their own morph. We present evidence that this preference may be mediated by color pattern specifically. These results contrast with an earlier study of a third transition zone, in which a mimetic shift was associated with reproductive isolation. Overall, our results suggest that the three known mimetic transition zones in R. imitator reflect a speciation continuum, which we have characterized at the geographic, phenotypic, behavioral, and genetic levels. We discuss possible explanations for variable progress toward speciation, suggesting that multifarious selection on both mimetic color pattern and body size may be responsible for generating reproductive isolation.

Parameter optimization in the regularized kernel minimum noise fraction transformation

Based on the original, linear minimum noise fraction (MNF) transformation and kernel principal component analysis, a kernel version of the MNF transformation was recently introduced. Inspired by we here give a simple method for finding optimal parameters in a regularized version of kernel MNF analysis. We consider the model signal-to-noise ratio (SNR) as a function of the kernel parameters and the regularization parameter. In 2-4 steps of increasingly refined grid searches we find the parameters that maximize the model SNR. An example based on data from the DLR 3K camera system is given.

Automated Invariant Alignment to Improve Canonical Variates in Image Fusion of Satellite and Weather Radar Data

AbstractCanonical correlation analysis (CCA) maximizes the correlation between two sets of multivariate data. CCA is applied to multivariate satellite data and univariate radar data to produce a subspace descriptive of heavily precipitating clouds. A misalignment, inherent to the nature of the two datasets, was observed, corrupting the subspace. A method for aligning the two datasets is proposed to overcome this issue and render a useful subspace projection. The observed corruption of the subspace gives rise to the hypothesis that the optimal correspondence between a heavily precipitating cloud in the radar data and the associated cloud top registered in the satellite data is found by a scale, rotation, and translation invariant transformation together with a temporal displacement. The method starts by determining a conformal transformation of the radar data at the time of maximum precipitation for optimal correspondence with the satellite data at the same time. This optimization is repeated for an increa...

BIELSCHOWSKY’s STRABISMUS

Bielschowsky’s strabismus is a rather infrequent form of squinting. Still, if one is on the lookout for it, I think it is not quite so rare as generally assumed. In 1911 Bielschowsky presented before the Congress of Ophs thalmologists, meeting in Heidelberg, the first material coms prising several cases of this anomaly as a clinical entity with a characteristic complex of symptoms. The symptoms met with in these patients resemble very much those encountered in typical trochlear paresis, namely:

Number of genes controlling a quantitative trait in a hybrid zone of the aposematic frog Ranitomeya imitator

The number of genes controlling mimetic traits has been a topic of much research and discussion. In this paper, we examine a mimetic, dendrobatid frog Ranitomeya imitator, which harbours extensive phenotypic variation with multiple mimetic morphs, not unlike the celebrated Heliconius system. However, the genetic basis for this polymorphism is unknown, and not easy to determine using standard experimental approaches, for this hard-to-breed species. To circumvent this problem, we first develop a new protocol for automatic quantification of complex colour pattern phenotypes from images. Using this method, which has the potential to be applied in many other systems, we define a phenotype associated with differences in colour pattern between different mimetic morphs. We then proceed to develop a maximum-likelihood method for estimating the number of genes affecting a quantitative trait segregating in a hybrid zone. This method takes advantage of estimates of admixture proportions obtained using genetic data, such as microsatellite markers, and is applicable to any other system where a phenotype has been quantified in an admixture/introgression zone. We evaluate the method using extensive simulations and apply it to the R. imitator system. We show that probably one or two, or at most three genes, control the mimetic phenotype segregating in a R. imitator hybrid zone identified using image analyses.

Change detection in bi-temporal data by canonical information analysis

Canonical correlation analysis (CCA) is an established multivariate statistical method for finding similarities between linear combinations of (normally two) sets of multivariate observations. In this contribution we replace (linear) correlation as the measure of association between the linear combinations with the information theoretical measure mutual information (MI). We term this type of analysis canonical information analysis (CIA). MI allows for the actual joint distribution of the variables involved and not just second order statistics. Where CCA is ideal for Gaussian data, CIA facilitates analysis of variables with different genesis and therefore different statistical distributions. As a proof of concept we give a toy example. We also give an example with DLR 3K camera data from two time points covering a motor way.

Structure identification in high-resolution transmission electron microscopic images: an example on graphene.

A connection between microscopic structure and macroscopic properties is expected for almost all material systems. High-resolution transmission electron microscopy is a technique offering insight into the atomic structure, but the analysis of large image series can be time consuming. The present work describes a method to automatically estimate the atomic structure in two-dimensional materials. As an example graphene is chosen, in which the positions of the carbon atoms are reconstructed. Lattice parameters are extracted in the frequency domain and an initial atom positioning is estimated. Next, a plausible neighborhood structure is estimated. Finally, atom positions are adjusted by simulation of a Markov random field model, integrating image evidence and the strong geometric prior. A pristine sample with high regularity and a sample with an induced hole are analyzed. False discovery rate-controlled large-scale simultaneous hypothesis testing is used as a statistical framework for interpretation of results. The first sample yields, as expected, a homogeneous distribution of carbon-carbon (C-C) bond lengths. The second sample exhibits regions of shorter C-C bond lengths with a preferred orientation, suggesting either strain in the structure or a buckling of the graphene sheet. The precision of the method is demonstrated on simulated model structures and by its application to multiple exposures of the two graphene samples.