This website is using a security service to protect itself from online attacks
The action you just performed triggered the security solution
There are several actions that could trigger this block including submitting a certain word or phrase
You can email the site owner to let them know you were blocked
Please include what you were doing when this page came up and the Cloudflare Ray ID found at the bottom of this page
ShareSaveInnovationScienceThe First Landslide Disaster Investigated By Geologists Happened In 1806ByDavid Bressan
Painting by Swiss artist Xaver Triner (1767-1824) of the landslide of Goldau
described the entire event in great detail:
'The morning of this sorrowful day started with strong rain
becoming less and lesser until at midday it stopped
there were fissures in the earth and cracks in the meadows visible.' Later that day
Some larger stones are already rolling down the mountain
smashing houses … Now suddenly the movement increases
whole rows of pine trees fall down and into the depths ..
A terrible roar is heard; whole sections of soil are pushed up
rock pieces as large or larger than houses
hiding the avalanche in darkness and running before it
as a dark cloud chased by the stormy wind ..
faster than a cannonball from a cannon."
Zay's description shows that the landslide started as a single mass. This mass, flowing down the slope, formed then a 'Trümmerstrom', also known as debris avalanche
Zay describes also warning signs before the landslide happened
Fissures in the earth and cracks in the meadows were visible
the sound of breaking trees could be heard in the forest and some minor rockfalls occurred days before the catastrophe
at the time nobody realized the danger until it was too late
The Rossberg consists of a succession of fine-grained clays
sandstone and thick layers of coarse-grained conglomerates
Tectonic movements tilted the layers and in the wet summer of 1806 water infiltrated in large quantities into the rocks
forming sliding planes between the conglomerate and clay beds
the upper layers of the mountain started slowly to creep down the valley
a 40 million cubic meters large stack of layers slipped down the mountain
The scar left by the landslide of Goldau on the Rossberg is still visible today
the Rigi attracts visitors every second Friday evening from June to October with small cultural and culinary highlights
mountain railway tickets are available at a special price and extra trains run until late
Write to us if you have any comments on this article or would like to report an error
The new «local mountain» offer on Mount Rigi makes it possible to enjoy the particularly charming evening hours in the mountains in a special way: Every second Friday from June to October
the «Queen of the Mountains» with its panoramic viewpoints as well as walking and hiking trails offers a special evening programme as well as mountain railway tickets at a special price of CHF 30 from 5 pm
Trains run down to Goldau and Vitznau until late in the evening
The cultural and culinary highlights on «Hausberg» Fridays:
Blue Hour at the Mineralbad & Spa Rigi Kaltbad: from 7 June to 11 October every second Friday until 10 p.m.
enjoy the tranquillity and magic of the blue hour - e.g
in the outdoor pool with a view of the mountains or at the cocktail bar with a selection of special drinks
Whisky appreciation evening at the Hotel Rigi Kaltbad: 7 June from 19:15
with Guido Stohler from the Hauptstross 100 whisky shop
The «Hausberg-Ticket» for CHF 30 (with or without half-fare travelcard) is valid from 5 pm on all Rigi Bahnen AG cable cars on the following dates: on 7 June
The descent with the cogwheel railway to Arth-Goldau and Vitznau is possible until late in the evening
an extra train runs from Rigi Kulm to Arth-Goldau at 22:30
while the last regular train to Vitznau runs as usual at 22:40
Information and booking, as well as the various events can be found on the website
Please enable JavaScript to see this content.Need help? Enable JavaScript in our browser
Volume 5 - 2011 | https://doi.org/10.3389/fninf.2011.00018
This article is part of the Research TopicMapping the connectome: Multi-level analysis of brain connectivityView all 21 articles
One of the most promising avenues for compiling connectivity data originates from the notion that individual brain regions maintain individual connectivity profiles; the functional repertoire of a cortical area (“the functional fingerprint”) is closely related to its anatomical connections (“the connectional fingerprint”) and
a segregated cortical area may be characterized by a highly coherent connectivity pattern
Diffusion tractography can be used to identify borders between such cortical areas
Each cortical area is defined based upon a unique probabilistic tractogram and such a tractogram is representative of a group of tractograms
The underlying methodology is called connectivity-based cortex parcellation and requires clustering or grouping of similar diffusion tractograms
Despite the relative success of this technique in producing anatomically sensible results
existing clustering techniques in the context of connectivity-based parcellation typically depend on several non-trivial assumptions
we embody an unsupervised hierarchical information-based framework to clustering probabilistic tractograms that avoids many drawbacks offered by previous methods
Cortex parcellation of the inferior frontal gyrus together with the precentral gyrus demonstrates a proof of concept of the proposed method: The automatic parcellation reveals cortical subunits consistent with cytoarchitectonic maps and previous studies including connectivity-based parcellation
Further insight into the hierarchically modular architecture of cortical subunits is given by revealing coarser cortical structures that differentiate between primary as well as premotoric areas and those associated with pre-frontal areas
for the purpose of cortical area parcellation
probabilistic tractography does not necessarily have to accurately reflect the connectivity pattern of an individual area
The sensitivity of probabilistic tractography to differences in connectivity of cortical areas plays a much more important role
This motivates the application of tractography for connectivity-based parcellation: When each cortical area is characterized by unique cortico-cortical connections (“connectional fingerprint”)
any tractogram within an area should be similar
The aforementioned attempts at clustering probabilistic tractograms
impose several non-trivial assumptions about the underlying structure of the data
it is often difficult to justify the choice of a particular number of clusters a priori
the choice of the number of cortical subunits has been subject to forming representative
meaningful cortical regions while still maintaining relative consistency across subjects
two different types of clustering algorithms have been used to perform tractography-based parcellation:
(1) Similarity-based clustering methods, such as K-means clustering (Anwander et al., 2007; Klein et al., 2007; Nanetti et al., 2009) or spectral reordering (Johansen-Berg et al., 2004)
employ correlation as a predefined similarity measure and thus explicitly rely on the strength of linear dependency between tractograms in order to form clusters
whether similarity between tractograms should be defined by their linear dependency to one another
(2) Dirichlet process mixture models (Jbabdi et al., 2009) embody a Bayesian non-parametric model for clustering of probabilistic tractograms. Such stochastic processes typically assume data to be generated from a mixture of Gaussian distributions. In an application to multiple-subject parcellation of the thalamus, Jbabdi et al. (2009) represented tractograms as vectorial data and grouped them based upon a Gaussian likelihood function
Whether or not individual tractograms can be interpreted as vectors and subsequently clustered using Gaussian likelihood functions is undetermined
A broad transition region may reflect biologically genuine gradations
such that neurons within the transition region have anatomical and/or physiological characteristics intermediate between the neighboring subdivisions
an important issue concerning parcellation is to assess the spatial extend over which such architectonic transitions occur
The problem that an a priori determination of the number of clusters may not be possible clearly motivates an unsupervised clustering approach
The purpose of this study is therefore to formally adopt such an approach
we employ an information-theoretic framework to minimize the assumptions imposed on data
We assume for subsequent discussion that the connectivity pattern of each distinct cortical subunit retains a prototype property
referred to as exemplars in subsequent sections
such that a particular tractogram is approximately representative of the connectivity pattern of the entire cortical subunit
Further grouping of cortical subunits forms hierarchically modular structures that each contains multiple representative tractograms
we make the prior assumption that probabilistic tractography is capable of revealing information pertaining to nested structures
Our approach makes use of soft-constraint affinity propagation (SCAP; Leone et al., 2008) to seek exemplar tractograms that are each representative of cortical subunits
Global clusters of tractograms are formed by extracting disjoint sets of connected components each consisting of multiple exemplars
individual global clusters are allowed to share multiple centers (i.e.
exemplars) thereby allowing for the formation of irregularly shaped clusters
The number of clusters of the global partition is determined based upon the robustness of the clustering solution against uncertainty in the data measured by clustering several bootstrap dataset samples (Fischer and Buhmann, 2003)
The rationale behind this approach is to allow the uncertainty in the data to vote for the choice of exemplars and therefore the finest granularity level that gives rise to the most stable partitioning at a higher hierarchical level
Rate distortion theory (Tishby et al., 1999) is used to stochastically map tractograms to exemplars thereby inducing a soft partition between cortical areas. A more informative nested architecture is obtained using information-theoretic agglomerative grouping (Slonim and Tishby, 1999) of cortical areas by preserving as much information as possible about the representative tractograms through the partitioning at each step of the merging sequence
Diffusion-weighted data and high-resolution 3-dimensional (3D) T1- and T2-weighted images were acquired on a Siemens 3T Trio scanner with an eight-channel array head coil and maximum gradient strength of 40 mT/m
The diffusion-weighted data were acquired using spin-echo echo planar imaging (EPI; TR = 12 s
resolution 1.72 mm × 1.72 mm × 1.7 mm
A GRAPPA technique (reduction factor 2.0) was chosen as parallel imaging scheme
Diffusion weighting was isotropically distributed along 60 directions (b-value = 1000 s/mm2)
seven data sets with no diffusion weighting were acquired initially and interleaved after each block of 10 diffusion-weighted images as anatomical reference for motion correction
The high angular resolution of the diffusion weighting directions improves the robustness of the tensor estimation by increasing the signal-to-noise ratio (SNR) and reducing directional bias
scanning was repeated three times for averaging
requiring a total scan time for the dMRI protocol of approximately 45 min
dMRI data were acquired after the T2-weighted images in the same scanner reference system
Let each tractogram xi be a list of connectivity scores y(i) for all random paths originating from a particular seed voxel to every other white matter target voxel (i.e.
target voxel) such that the i-th tractogram is given by a ∈ Ω
where Ω denotes the set of all target voxels and η denotes the number of target voxels
for the purpose of unsupervised cortex parcellation
the set of imaging voxels compromises the whole white matter volume
General methodology to obtain connectivity information of a cortical area using probabilistic tractograms
(A) Seed voxels in the white matter near the white–gray matter interface form the region of interest
(B) A probabilistic tractogram is computed for each seed voxel
are similar with respect to their connectivity
our intuition about similar tractograms arises from the notion that one tractogram xi reveals information about connectivity associated with another tractogram xj and vice versa
From an information-theoretic point of view
an overlap in uncertainty between tractograms xi and xj translates into a gain in mutual information
Mutual information is one such quantity that provides a unique measure of the interdependence between tractograms:
where H(xi) is the entropy and a measure of uncertainty in connectivity associated with tractogram xi
the conditional entropy H(xi|xj) measures the remaining uncertainty in connectivity of tractogram xi after xj is observed
Mutual information is thus intuitively defined as the amount of uncertainty removed in xi after observing xj or equivalently the amount of information tractogram xi provides about tractogram xj
Mutual information is computed as follows:
and in common target voxels between two tractograms on the left hand side serve as coordinates of the frequency table given on the right hand side
Elements of the frequency table are incremented for each occurrence of a combination of connectivity scores
denoted by “responsibility” and “availability,” which together reflect the accumulated affinity tractogram xi has for choosing tractogram xq as its exemplar:
where the penalty term ρ serves as a free parameter
Tractogram xq infers its suitability for serving as an exemplar for tractogram xi by comparing its similarity with tractogram xi and the maximum of similarities between tractogram xi
corrected by availability a(xi,xq) and all other tractograms
A positive responsibility reveals that tractogram xi prefers tractogram xq as its exemplar
The sum of accumulated positive incoming responsibility messages computed by availability gathers further evidence as to whether candidate exemplar xq is a favorable exemplar for a group of tractograms
The goal of the message-passing procedure is to converge upon a set of exemplars such that the maximum net similarity of the data is attained
the exemplar choice of xi is extracted by selecting the candidate exemplar with which tractogram xi has maximum affinity (i.e.
similarity corrected by the availability):
Clustering of synthetic data to illustrate the capability of soft-constraint affinity propagation (SCAP) in capturing two levels of clustering
SCAP identifies 12 exemplars shown as circles and therefore 12 sub-clusters as well as their preferred grouping in three global clusters
Arrows indicate the affinity between exemplars at the top-level of the nested hierarchy
Note that the penalty term ρ influences the number of global clusters K and therefore the number of exemplars
The following section discusses a means to infer the number of global clusters and therefore the optimal ρ independent of the clustering algorithm
The method applied in this paper to assess an optimal clustering solution (i.e., to yield an automatic estimation of the number of clusters) was originally developed by Fischer and Buhmann (2003) and concerns the reliability of clustering tractograms: Uncertainty in the partitioning is quantified by clustering B bootstrap samples drawn from the original dataset
The empirical distribution of cluster assignments learned from clustering B multiset replications quantifies the uncertainty in mapping tractogram xi to cluster k for the same number of clusters across the bootstrap samples
A problem related to estimating the empirical assignment probability is to identify equivalent clusters across partitionings of different datasets
A greedy approach is to search the particular permutation πb + 1 of cluster labels cb + 1 of dataset Xb + 1 that maximizes the sum over all cluster assignment probabilities learned from the previous b mappings:
The Hungarian method (Kuhn, 1955) finds the permutation πb + 1 efficiently without having to search through K! possible permutations. More precisely, the problem is formulated in terms of a weighted bipartite matching that contains two sets of nodes with each set containing a permutation of cluster labels (Fischer and Buhmann, 2003)
Edges between nodes give the original assignment of label k to the assignment of a label π(k) from a permuted set
Maximizing the sum over all possible weights using the Hungarian method with a running time of X(K3) is equivalent to solving Eq
Finding the optimal cluster relabeling in each of the bootstrap sample allows one to quantify the reliability of clustering tractograms across different data replicates based upon their maximum likelihood given by:
where defines the maximum likelihood mapping. Fischer and Buhmann (2003) propose a stability criterion that compares the reliability of the maximum likelihood mapping with the reliability of making random cluster assignments relative to the risk of misclassification:
where and are the mean of maximum and random probability assignments
By validating the global partitioning obtained by SCAP one yields a set of exemplars that give rise to the most stable global partitioning
Such a set of exemplars proves useful in identifying the finest level of detail of the hierarchy within a bottom-up approach as introduced in a latter section
previous attempts at tractography-based parcellation have formed hard borders between cortical subunits
it is unclear whether borders between cortical structures should be distinct or have a transitional property
A soft partitioning of the data should therefore be made
in order to account for transitional regions
Such a soft partitioning is induced by means of a stochastic mapping
and in order to map tractograms to exemplars as opposed to making hard assignments
The most straightforward approach is to convert dissimilarity measures (i.e.
distortion measures) into stochastic mappings using the following equations:
where t denotes the iteration sequence and A serves as the normalization constant
10(top set) with a likelihood function given by Note that the likelihood function contains the distortion measure between tractograms and exemplars together with the computational temperature T that sets the scale for converting dissimilarity measures into probabilities
10(bottom) is computed by summing over all conditional probabilities
However we require that the conditional as well as marginal probabilities remain consistent (i.e., they do not change with respect to one another). Within an information-based framework the problem can be formulated in terms of rate distortion theory where the conditional entropy and the expected distortion determine the quality of the stochastic mapping (Tishby et al., 1999):
Variation of information serves as the distortion measure, between tractogram xi and the exemplar Note that conditional entropy characterizes the average information required, in bits per tractogram, to invoke a mapping of a tractogram to an exemplar without confusion (Tishby et al., 1999)
Rate distortion theory characterizes the tradeoff between information rate and expected distortion
where the objective is to allot membership probabilities to tractograms in order to maximize compression (i.e.
equivalent to minimizing information rate) under the expected distortion constraint
Finding the rate distortion function is solved by introducing the Lagrange multiplier or inverse temperature
and minimizing the corresponding functional:
Minimization of the functional yields the set of self-consistent equations (Eq. 10) that are each iterated over convex sets of normalized distributions given by Blahut (1978). More precisely, Blahut (1978) proves that
iterating over the conditional and marginal probabilities in Eq
10yields the global minimum of the functional F in Eq
Note that both conditional and marginal probabilities remain consistent at the global minimum of F
evidence suggests that clusters of anatomical connectivity patterns are organized into a hierarchical structure; whereby bottom-level clusters reveal finer structures and top-level clusters (i.e.
coarser clusters reveal a collection of finer structures) within the region of interest
The SCAP approach identified cortical subunits as well as inferred their preferred grouping into a global partitioning
from an anatomical point of view the nested structure (i.e.
preferred grouping of cortical subunits) might exist at multiple levels thereby constituting a more informative hierarchical structure
which preserves as much information about exemplars as possible [i.e.
if the Jenson–Shannon distance between their conditional distributions corrected for marginal probabilities
the clusters that we have to merge is found by minimizing
the merge cost is a product of the weight sum of clusters
and the distance between them with respect to the exemplars measured by the Jenson–Shannon divergence
This optimization strategy was formerly introduced as agglomerative information bottleneck method (Slonim and Tishby, 1999) – Figure 4 illustrates our application
After merging clusters, the marginal and conditional probabilities for and are updated as follows (Slonim and Tishby, 1999):
Note that the sub-clusters obtained by SCAP are used as the initial hard partition between cortical subunits ZM
Assessment of clustering solutions in Figure 5 based upon the stability criterion (Eq
9) suggests four global clusters consisting of 15 exemplars and thus 15 cortical subunits as the most stable solution within the region of interest
Figure 5. Assignment probabilities plotted against the number of global clusters K. Dashed plot shows the mean assignment probability based upon the maximum likelihood mapping. Dotted plot shows the random cluster assignment probabilities. The stability index (Fischer and Buhmann, 2003) used to select the number of clusters K* is the relative difference between mean and random cluster assignment probabilities (solid plot)
The most stable partitioning K* is given by the preferred grouping of 15 cortical subunits
Figure 6A illustrates the preferred grouping of cortical subunits in the region of interest in four global cortical structures: The PCG is divided into two areas
a dorsal area (dPCG) and a ventral area (vPCG)
then a ventral transition into the posterior IFG resides at the ventral tip of the PCG and pars opercularis of the IFG
pars triangularis of the IFG and the deep frontal operculum together form the forth group
(A) Cortex parcellation of the IFG together with the PCG showing four global cortical structures on the gray matter surface
(B) Cortex parcellation of the same ROI at finest level of detail expressed by a hierarchy indicates 15 cortical subunits
Figure 7. Hierarchical tree of the nested hierarchical structure constructed by the agglomerative information bottleneck method (Slonim and Tishby, 1999)
The cardinality of the initial partition equals the number of exemplars
The length of each branch gives the normalized information loss due to merging operations
Bottom left: Internal organization of representative tractograms within four global cortical structures obtained by SCAP
Top left: Cortex parcellation showing the finest level of detail
Black dots indicate the location of exemplars while arrows show the preferred grouping of cortical subunits obtained by SCAP
For the latter there is further modular organization showing distinction between areas of ventral PCG and those of posterior IFG. The parcellation results given alongside the hierarchical tree in Figure 7 show the finest detail expressed by cortical structures. Arrows in Figure 7 illustrate the preferred grouping of cortical subunits into four global cortical structures shown in Figure 6A
Notice that the same four global cortical structures emerge from the agglomerative information bottleneck method
We propose an unsupervised information-based clustering technique for connectivity-based cortex parcellation suitable for automatic parcellation. The methodological framework used here to reveal complex properties of cortical subunits such as transitional borders as well as a modular hierarchical architecture is summarized in Figure 8
Overview of the unsupervised framework used in this study
SCAP together with clustering assessment using bootstrap sampling is used to obtain exemplars that each characterizes individual cortical subunits within a global and bottom-level partitioning
rate distortion theory is used to map tractograms to exemplars
The agglomerative information bottleneck method used information gained from rate distortion theory [i.e.
] to construct a more informative nested structure of cortical subunits
The sub-clusters obtained by SCAP define the bottom-level partitioning of the hierarchy
A proof of principle of the approach yields anatomically sensible results
The parcellation results advocate that dorsal PCG fields can be separated in agreement with the suggestion that this area consists of two premotor areas (Schubotz et al., 2010), as well as primary motor cortex (Geyer et al., 1996) and the frontal eye field at the rostral bank of precentral sulcus and the ventral branch of posterior superior frontal sulcus (Amiez and Petrides, 2009)
such as the ventral precentral transitional cortex 6r1
and areas in the frontal operculum op8 and op9
Our results therefore suggest a specific connectivity underlying IFJ
rendering this region as a distinct anatomical area
Figure 9. Parcellation of the sub-fields in the posterior ventral precentral cortex indicate a striking resemblance with cytoarchitectonic and multireceptor studies as those were recently reported by Amunts et al. (2010) – cf
The merging of the postcentral region (orange field in Figure 6B) with the ventral PCG at a rather high hierarchical level seems to be supported by findings in non-human primates, implying dense bidirectional connections between the rostral portion of the inferior parietal lobule and the adjacent opercular area, i.e., ventral premotor area 6 (cf., e.g., Schmahmann and Pandya, 2007)
whether this suggestion is indeed evident in tractography-based connectivity scores remains to be studied in detail
and specifically with respect to limitations potentially imposed by the choice of a particular tractography method and the underlying diffusion model
A well-known difficulty of most clustering algorithms is the choice of an appropriate similarity measure
since this ultimately determines the cluster structure that can be inferred from the data – i.e.
elements within the same cluster share a common similarity quantified by the respective measure
clustering of tractograms should be based upon capturing the shape of probabilistic tractograms
probabilistic tractograms should be grouped together if they have similar shape
Such tractograms are represented as volumes containing connectivity scores for each target voxel
Defining their shape is therefore not straightforward
We define two tractograms as having similar shape if their connectivity scores in common (i.e.
corresponding) target voxels (not any target voxel) are similar
The similarity measure should therefore involve a pairwise comparison of connectivity scores with pairs of connectivity scores given by where a denotes the particular target voxel common to both tractograms i and j
In order to compute how probable it is that two tractograms have similar shape we consider the joint occurrence of connectivity scores p(y(i),y(j)) for any common target voxel; p(y(i),y(j)) is computed by constructing the frequency table using the frequency of occurrence of pairs of connectivity scores within all common target voxels. Constructing the frequency table as shown in Figure 2 already involves pairwise comparisons of connectivity scores in common target voxels
If p(y(i),y(j)) is high for all connectivity scores among common target voxels in tractograms i and j it follows that tractograms i and j are likely to have similar shape
Mutual information measures the dependency of one tractogram on another tractogram
Since we are interested in capturing the shape of tractograms
measures how dependent the shape of one tractogram is on the shape of another tractogram
mutual information will capture any type of dependency including linear and non-linear dependencies between the shapes of tractograms
An issue that draws less attention is the dependency of similarity measures on the representation of the data – i.e.
different transformations of the data will produce different similarity quantities
we have limited knowledge about the structure of clusters or about which type of relation should be considered
the similarity measure should be invariant to data representation
Mutual information has the useful property of being independent of representation of the data – i.e.
different invertible transformations on individual tractograms will yield the same mutual information quantity
uncertainty in the data gives rise to uncertainty in the clustering solution
A sensible global partitioning is one for which the uncertainty in the data has minimum influence on the clustering solution
12) is dependent upon the ordering of the bootstrap samples
particularly if the first sample leads to a poor clustering solution
we suggest computing the stability criterion for different permutations of bootstrap sample orderings in order to avoid the influence of initial poor solutions on the stability criterion
In contrast to other traditional hierarchical clustering methods we insist upon defining the finest level of detail of the partitioning other than simply associating it with the maximum number of clusters (i.e.
very fine detailed hierarchical organization of cortical subunits may exist
the level of detail that probabilistic tractography is capable of revealing is
limited by resolution offered by diffusion-weighted images
Information rate I() plotted against the inverse temperature 1/T
Solid and dashed plots illustrate the relationship between information rate and inverse temperature for 15 cortical subunits and 4 global cortical structures
1/T* gives the temperature for which the information rate has only slight changes for lower temperatures
Maximal amount of information is thus used to construct the hierarchy while taking into account some uncertainty in mapping tractograms to exemplars
a meaningful parcellation should be assumed to exist in all subjects with similar location
Note that it is unclear whether or not cortical subunits possess the aforementioned “prototype” (i.e., exemplar) characteristic. A clustering method that avoids defining cluster “prototypes” might therefore prove more suitable for parcelating regions without a priori knowledge (Slonim et al., 2005)
an obvious limitation of any model-based approach to reveal structure in the data is that it already assumes that there is structure in the data
In the context of hierarchically modular cortical subunits
hierarchical clustering models are already forced to find nested structures in the data
The question of whether connectivity patterns
prefer to be grouped in a nested structure or not should be addressed and is formulated in terms of model validation
Buhmann (2010) performs model validation based upon an indispensible requirement that the solution should be generalizable under the influence of noise
a nested structure of cortical subunits should be generalizable given noise in the diffusion measurements
Assessing the generalizability performance is given by a tradeoff between the informativeness as well as the robustness against noise of the nested clustering solution
a nested structure showing more branching is more informative
The question is to be answered is therefore: How informative can the nested structure of cortical subunits be (i.e.
how many branches if any should the dendrogram have) without fitting the sampling noise
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest
von Cramon for his continuing support and his endurance to teach us tedious anatomy
Further appreciation for insightful discussions is given to J
Swiss Federal Institute of Technology Zurich
The work was supported by the German Ministry of Education and Research (Grants No
Anatomical organization of the eye fields in the human and non-human primate frontal cortex
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Broca’s region: novel organizational principles and multiple receptor mapping
CrossRef Full Text
The anatomical segregation of the frontal cortex: what does it mean for function
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Connectivity-based parcellation of Broca’s area
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Statistical analysis of parieto-frontal cognitive-motor networks
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Deep and superficial amygdala nuclei projections revealed in vivo by probabilistic tractography
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Understanding complexity in the human brain
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Efficient physical embedding of topologically complex information processing networks in brains and computer circuits
CrossRef Full Text
Connectivity-based parcellation of human cingulate cortex and its relation to functional specialization
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Non-invasive mapping of connections between human thalamus and cortex using diffusion imaging
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Computation of channel capacity and rate distortion theory
CrossRef Full Text
The role of the inferior frontal junction area in cognitive control
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Vergleichende Lokalisationslehre der Groß hirnrinde in ihren Prinzipien dargestellt auf Grund des Zellaufbaues
“Information theoretic model validation for clustering,” in International Symposium on Information Theory (Austin
Heuristics for connectivity-based brain parcellation of SMA/pre-SMA through force-directed graph layout
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Neural activations at the junction of the inferior frontal sulcus and the inferior precentral sulcus: interindividual variability
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Reliable identification of the auditory thalamus using multi-modal structural analyses
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Evidence for segregated and integrative connectivity patterns in the human basal ganglia
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Über die variabilitat der grosshirnrindenstruktur
Regio occipitalis beim erwachsenen Menshchen
CrossRef Full Text
Structural connectivity of Broca’s area and medial frontal cortex
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Clustering by passing messages between data points
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Physiology and Neuropsychology of the Frontal Lobe
Baltimore: Lippincott/Williams & Wilkins
Two different areas within the primary motor cortex of man
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Functional neuroanatomy of the primate isocortical motor system
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
specificity and variability of corticocortical connectivity
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Multiple-subjects connectivity-based parcellation using hierarchical Dirichlet process mixture models
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Improved optimization for the robust and accurate linear registration and motion correction of brain images
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Functional-anatomical validation and individual variation of diffusion tractography-based segmentation of the human thalamus
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Changes in connectivity profiles define functionally distinct regions in human medial frontal cortex
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Using diffusion imaging to study human connectional anatomy
Challenges and limitations of quantifying brain connectivity in vivo with diffusion MRI
CrossRef Full Text
A tutorial in connectome analysis: topological and spatial features of brain networks
Development of multi-cluster cortical networks by time windows for spatial growth
CrossRef Full Text
Optimal hierarchical modular topologies for producing limited sustained activation of neural networks
CrossRef Full Text
Connectivity-based parcellation of human cortex using diffusion MRI: establishing reproducibility
validity and observer independence in BA 44/45 and SMA/pre-SMA
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Knösche
The role of long-range connectivity for the characterization of the functional-anatomical organization of the cortex
CrossRef Full Text
The Hungarian method for the assignment problem
CrossRef Full Text
Diffusion tensor fiber tracking shows distinct Corticostriatal circuits in humans
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Unsupervised and semi-supervised clustering by message passing: soft-constraint affinity propagation
CrossRef Full Text
and multimodal processing areas in the parietal lobe of the macaque monkey
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Weight consistency specifies regularities of macaque cortical networks
Topographic organization of the middle temporal visual area in the macaque monkey: representational biases and the relationship to callosal connections and myeloarchitectonic boundaries
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Connectivity-based segmentation of the substantia nigra in human and its implications in Parkinson’s disease
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Modular and hierarchically modular organization of brain networks
CrossRef Full Text
Group analyses of connectivity-based cortical parcellation using repeated k-means clustering
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
O’Muircheartaigh
Clustering probabilistic tractograms using independent component analysis applied to the thalamus
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
The anatomical basis of functional localization in the cortex
Pubmed Abstract | Pubmed Full Text
Motor and cognitive functions of the ventral premotor cortex
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
“Inter-subject connectivity-based parcellation of a patch of cerebral cortex,” in International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI) (New York: Springer-Verlag)
Architectonics of the human frontal lobe of the brain
With a demonstration of the principles of its formation as a reflection of phylogenetic differentiation of the cerebral cortex
Pubmed Abstract | Pubmed Full Text
Anatomical and functional parcellation of the human lateral premotor cortex
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Connectivity of the human periventricular-periaqueductal gray region
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
“Agglomerative information bottleneck,” in Neural Information Processing Systems (MIT Press) 12
Computational analysis of functional connectivity between areas of primate cerebral cortex
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
“The information bottleneck method,” in Proceedings of the 37th Annual Allerton Conference on Communication
Diffusion-weighted imaging tractography-based parcellation of the human lateral premotor cortex identifies dorsal and ventral subregions with anatomical and functional specializations
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Consequences of large interindividual variability for human brain atlases: converging macroscopical imaging and microscopical neuroanatomy
Pubmed Abstract | Pubmed Full Text | CrossRef Full Text
Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations
Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher
94% of researchers rate our articles as excellent or goodLearn more about the work of our research integrity team to safeguard the quality of each article we publish
Notices are posted by 10 am Monday through Saturday
Adjust Text Size: A+ A-
As published in Winnipeg Free Press on Sep 21
Share your memories and/or express your condolences below
Unfortunately with the need to moderate tributes for inappropriate content
your comments may take up to 48 hours to appear
CalgaryNewsTeenage girl gifted rescue horse for Valentines Day By Tyson FedorPublished: February 14, 2024 at 8:08PM EST
Twitter feed ©2025 BellMedia All Rights Reserved
Get daily business news from the region delivered straight to your inbox
reveals the secret to the jewellery brand’s success and where it is heading in the future
How has the gold and jewellery market changed since Au Finja was founded
“Au Finja as a brand has been in existence for over 20 years now
Before our existence the bangle industry only produced handcrafted pieces of jewellery
and we became the first true brand in this category
Today our range of machine-made wrist-wear is leading the market not only in sheer volume worldwide
but also the quality and precision of our designs
while quality is the most discussed aspect of the jewellery business
Most brands put out sub-standard products with superficial finishing
If something – anything – prevents you from passing the quality of an item
Every employee or worker has to come to terms with the strict quality protocol of the company
we haves set various benchmarks for the industry and has ushered in a quality-conscious climate.”
Has demand been affected at all by tougher economic conditions in the region over the last two years
One needs to bear in mind that tougher economic conditions may dampen consumer demand in several lifestyle categories
but not so heavily in categories such as gold jewellery
This is primarily due to the emotional connection it has to traditions among countries in the Indian sub-continent as well as the Far East
Jewellery purchase demand fluctuate mores with the rise and fall in gold prices and not because of any major economic downturn.”
Do you have any new collections coming to market before the end of the year
10 new ranges have been launched within our different collections
The introduction of the AUSUM and YUME collections has been typical of the company’s philosophy – explore and innovate
the design ranges of these two collections have become a talking point among the top jewellery brands in the world
A major international luxury brand has approached Au Finja for an exclusive range of AUSUM to be featured as part of their jewellery collection
Discussions are in the finalisation stage.”
What is the company’s strategy moving forward and are there any areas of the market you will be focussing on
“The company has moved forward by introducing wedding bands in the market
design innovation and precision manufacturing is key here
Our initial results are more than encouraging and the company will launch its range of wedding bands in important outlets across the world
“The company is also at the advanced stage of launching the Au Finja gold coin in configurations ranging from 1g up to 20g
This uniquely-designed coin can also be worn as an ornament
“Europe and the Americas have given us a warm welcome and our market presence is steadily growing across major cities on these continents
Global expansion and the availability of our brand across more than 16,000 retail outlets across the world is a special feeling but humbling too
We need to work harder at our craft and continue to delight the end user with our product.”
What are your expectations for the gold and jewellery market next year
“With the introduction of VAT in the UAE and other parts of the GCC everyone is in the wait and watch mode
I can emphatically state that nothing affects the buyer of gold jewellery