Multi-scale colorectal tumour segmentation using a novel coarse to fine strategy

This paper addresses the problem of colorectal tumour segmentation in complex real world imagery. For efficient segmentation, a multi-scale strategy is developed for extracting the potentially cancerous region of interest (ROI) based on colour histograms while searching for the best texture resolution. To achieve better segmentation accuracy, we apply a novel bag-of-visual-words method based on rotation invariant raw statistical features and random projection based l2-norm sparse representation to classify tumour areas in histopathology images. Experimental results on 20 real world digital slides demonstrate that the proposed algorithm results in better recognition accuracy than several state of the art segmentation techniques.

Magnetic properties contribution to the identification and provenance of marine sediments: distal IRD in the Galicia Interior Basin (NW Spain)

This paper discusses the advantages of using a combined environmagnetic and geochemical approach to the provenance and characterization of distal IRDs occurring during the Last Glacial Period in core CI12PC3 from the Galicia Interior Basin (GIB). Six Heinrich layers (HL1-6) have been identified in the area in base to the detection of distinct populations of exotic magnetic mineral assemblages alien to the local/regional sedimentation environment. Their extension has been determined by Ca/Sr and Si/Sr ratios and their provenance by 143Nd/144Nd and 87Sr/86Sr isotopic ratios and FORCs. The sedimentary expression of HL is characterized by the presence of distal Ice Rafted Detritus (IRD). Distal IRD magnetic signatures in the GIB consist of (i) an increase of one order of magnitude in the peak amplitude of magnetic susceptibility from background values, (ii) a general coarsening of the magnetic grain size in a mineral assemblage dominated by titano-magnetites, (iii) FORC distributions pushing towards the coarse MD or PSD component, and (iv) thermomagnetic curves depicting the occurrence of several magnetite phases. These four features are very different from the fine-grained biogenic magnetic assemblages characterized by the combination of lower MS and higher coercivity values that dominate the predominant mixtures of the non-interacting SSD and PSD components in the non-IRD influenced background sedimentation. Our results show that the last 70.000 yr of sedimentation in the GIB were controlled by the relative contribution of local detrital material derived from the Iberian Variscan Chain and IRD alien material from the iceberg melting during the Heinrich Events. They also show two main IRD provenance fields: Europe and Canada. And that the later is more important for for HL1, HL2, HL4 and HL5. FORCs analysis complemented the isotopic information and provided a very unique information, indicating that glacial flour may not always have the same provenance as IRD and that ice-melted derived suspended sediment has its own dynamics and may reach further and/or persists longer than IRD.

Intellectual disability, coarse face, relative macrocephaly, and cerebellar hypotrophy in two sisters

We report on two Portuguese sisters with a very similar phenotype characterized by severe intellectual disability, absent speech, relative macrocephaly, coarse face, cerebellar hypotrophy, and severe ataxia. Additional common features include increased thickness of the cranial vault, delayed dental eruption, talipes equino-varus, clinodactyly, and camptodactyly of the fifth finger. The older sister has retinal dystrophy and the younger sister has short stature. Their parents are consanguineous. We suggest this condition constitutes a previously unreported autosomal recessive entity.

The influence of nano-scale second-phase particles on deformation of fine grained calcite mylonites

Grey and white carbonate mylonites were collected along thrust planes of the Helvetic Alps. They are characterised by very small grain sizes and non-random grain shape (SPO) and crystallographic preferred orientation (CPO). Presumably they deformed in the field of grain size sensitive flow by recrystallisation accommodated intracrystalline deformation in combination with granular flow. Both mylonites show a similar mean grain size, but in the grey mylonites the grain size range is larger, the grain shapes are more elongate and the dynamically recrystallised calcite grains are more often twinned. Grey mylonites have an oblique CPO, while the CPO in white mylonites is symmetric with respect to the shear plane. Combustion analysis and TEM investigations revealed that grey mylonites contain a higher amount of highly structured kerogens with particle sizes of a few tens of nanometers, which are finely dispersed at the grain boundaries. During deformation of the rock, nano-scale particles reduced the migration velocity of grain boundaries by Zener drag resulting in slower recrystallisation rates of the calcite aggregate. In the grey mylonites, more strain increments were accommodated by individual grains before they became refreshed by dynamic recrystallisation than in white mylonites, where grain boundary migration was less hindered and recrystallisation cycles were faster. Consequently, grey mylonites represent ‘deformation’ microfabrics while white mylonites are characterised by ‘recrystallisation’ microfabrics. Field geologists must utilise this different deformation behavior when applying the obliquity in CPO and SPO of the respective mylonites as reliable shear sense indicators.

A new technique to automatically quantify microstructures of fine grained carbonate mylonites: two-step etching combined with SEM imaging and image analysis

A two-step etching technique for fine-grained calcite mylonites using 0.37% hydrochloric and 0.1% acetic acid produces a topographic relief which reflects the grain boundary geometry. With this technique, calcite grain boundaries become more intensely dissolved than their grain interiors but second phase minerals like dolomite, quartz, feldspars, apatite, hematite and pyrite are not affected by the acid and therefore form topographic peaks. Based on digital backscatter electron images and element distribution maps acquired on a scanning electron microscope, the geometry of calcite and the second phase minerals can be automatically quantified using image analysis software. For research on fine-grained carbonate rocks (e.g. dolomite calcite mixtures), this low-cost approach is an attractive alternative to the generation of manual grain boundary maps based on photographs from ultra-thin sections or orientation contrast images.

Coarse-graining and renormalisation group methods for the elucidation of the kinetics of complex nucleation and growth processes

We review our work on generalisations of the Becker-Doring model of cluster-formation as applied to nucleation theory, polymer growth kinetics, and the formation of upramolecular structures in colloidal chemistry. One valuable tool in analysing mathematical models of these systems has been the coarse-graining approximation which enables macroscopic models for observable quantities to be derived from microscopic ones. This permits assumptions about the detailed molecular mechanisms to be tested, and their influence on the large-scale kinetics of surfactant self-assembly to be elucidated. We also summarise our more recent results on Becker-Doring systems, notably demonstrating that cross-inhibition and autocatalysis can destabilise a uniform solution and lead to a competitive environment in which some species flourish at the expense of others, phenomena relevant in models of the origins of life.

Path-integral and Coarse-graining Strategies for Complex Molecular Phenomena

Molecular simulation provides a powerful tool for connecting molecular-level processes to physical observables. However, the facility to make those connections relies upon the application and development of theoretical methods that permit appropriate descriptions of the systems or processes to be studied. In this thesis, we utilize molecular simulation to study and predict two phenomena with very different theoretical challenges, beginning with (1) lithium-ion transport behavior in polymers and following with (2) equilibrium isotope effects with relevance to position-specific and clumped isotope studies. In the case of ion transport in polymers, there is motivation to use molecular simulation to provide guidance in polymer electrolyte design, but the length and timescales relevant for ion diffusion in polymers preclude the use of direct molecular dynamics simulation to compute ion diffusivities in more than a handful of candidate systems. In the case of equilibrium isotope effects, the thermodynamic driving forces for isotopic fractionation are often fundamentally quantum mechanical in nature, and the high precision of experimental instruments demands correspondingly accurate theoretical approaches. Herein, we describe respectively coarse-graining and path-integral strategies to address outstanding questions in these two subject areas.