Predicting survival of glioblastoma from automatic whole-brain and tumor segmentation of MR images

Survival prediction models can potentially be used to guide treatment of glioblastoma patients. However, currently available MR imaging biomarkers holding prognostic information are often challenging to interpret, have difficulties generalizing across data acquisitions, or are only applicable to pre-operative MR data. In this paper we aim to address these issues by introducing novel imaging features that can be automatically computed from MR images and fed into machine learning models to predict patient survival. The features we propose have a direct anatomical-functional interpretation: They measure the deformation caused by the tumor on the surrounding brain structures, comparing the shape of various structures in the patient's brain to their expected shape in healthy individuals. To obtain the required segmentations, we use an automatic method that is contrast-adaptive and robust to missing modalities, making the features generalizable across scanners and imaging protocols. Since the features we propose do not depend on characteristics of the tumor region itself, they are also applicable to post-operative images, which have been much less studied in the context of survival prediction. Using experiments involving both pre- and post-operative data, we show that the proposed features carry prognostic value in terms of overall- and progression-free survival, over and above that of conventional non-imaging features.

Which Symptoms do Best Predict Major Depressive Disorder (MDD) in Males? Results from an Icelandic Community Study

Background/objectives:

The concept of male depression is based on the hypothesis that typical symptoms of depression in men often seem to be masked by non-typical male distress symptoms not considered in common depression inventories. Despite clinical evidence, scientific evidence is still lacking. The study describes symptoms reported by males (on the Gotland Male Depression Scale (GMDS) in a community study and these were further validated against MDS criteria.

Methods:

In a community study in Reykjanes Iceland, 534 males were rated by the GMDS. In the second phase a sub-sample of healthy and depressive males were examined by a trained psychiatrist in a semi-structured interview (N=137). Diagnosis was according to the DSM-IV for MDD. The symptoms on GMDS are self-rated from 0-3.

Results:

In the community sample the prevalence (>1 GMDS) of single symptoms was following: More stressed out than usual (16.8%), aggressiveness (11.8%), feeling of being burned out (14.5%), tiredness (18.2%), more irritable (15.3%), difficulty making decisions (10.8%), sleep problems (21%), anxiety/uneasiness (10.5%), increased use of alcohol/substances, hyperactivity ( 10.3%), behavioural changes (4.7%), feeling sad/hopeless (10.4%), self-pity (4.8%). All symptoms on GMDS correlated strongly with MDD. In a stepwiselogistic regression model, controlling for medication and age, three items contributed independently to MDD diagnoses. These were aggressiveness, being more irritable and anxiety/uneasiness.

Conclusions:

Sleep problems and tiredness were most commonly reported symptoms. The best predicting symptoms for MDD were anxiety and acting out symptoms, aggression and irritability. This is of interest as these symptoms are not in the criteria for MDD.

The fall and rise of the Icelandic Arctic fox (Vulpes lagopus): a 50-year demographic study on a non-cyclic Arctic fox population

In territorial species, observed density dependence is often manifest in lowered reproductive output at high population density where individuals have fewer resources or are forced to inhabit low-quality territories. The Arctic fox (Vulpes lagopus) in Iceland is territorial throughout the year and feeds mostly on birds, since lemmings are absent from the country. Thus, the population does not exhibit short-term population cycles that are evident in most of the species' geographical range. The population has, however, gone through a major long-term fluctuation in population size. Because of the stability in hunting effort and reliable hunting records since 1958, the total number of adult foxes killed annually can be used as an index of population size (N t ). An index of carrying capacity (K) from population growth data for five separate time blocks during 1958-2007 revealed considerable variation in K and allowed a novel definition of population density in terms of K, or N t /K. Correlation analysis suggested that the reproductive rate was largely determined by the proportion of territorial foxes in the population. Variation in litter size and cub mortality was, on the other hand, related to climatic variation. Thus, Arctic foxes in Iceland engage in typical contest competition but can adapt their territory sizes in response to both temporal and spatial variation in carrying capacity, resulting in surprisingly little variation in litter size.

Otolith shape: a population marker for Atlantic herring Clupea harengus

Otolith shape variation of seven Atlantic herring Clupea harengus populations from Canada, the Faroe Islands, Iceland, Ireland, Norway and Scotland, U.K., covering a large area of the species' distribution, was studied in order to see if otolith shape can be used to discriminate between populations. The otolith shape was obtained using quantitative shape analysis, transformed with Wavelet and analysed with multivariate methods. Significant differences were detected among the seven populations, which could be traced to three morphological structures in the otoliths. The differentiation in otolith shape between populations was not only correlated with their spawning time, indicating a strong environmental effect, but could also be due to differing life-history strategies. A model based on the shape differences discriminates with 94% accuracy between Icelandic summer spawners and Norwegian spring spawners, which are known to mix at feeding grounds. This study shows that otolith shape could become an accurate marker for C. harengus population discrimination.

Mosaic structure of native ant supercolonies

According to the inclusive fitness theory, some degree of positive relatedness is required for the evolution and maintenance of altruism. However, ant colonies are sometimes large interconnected networks of nests, which are genetically homogenous entities, causing a putative problem for the theory. We studied spatial structure and genetic relatedness in two supercolonies of the ant Formica exsecta, using nuclear and mitochondrial markers. We show that there may be multiple pathways to supercolonial social organization leading to different spatial genetic structures. One supercolony formed a genetically homogenous population dominated by a single mtDNA haplotype, as expected if founded by a small number of colonizers, followed by nest propagation by budding and domination of the habitat patch. The other supercolony had several haplotypes, and the spatial genetic structure was a mosaic of nuclear and mitochondrial clusters. Genetic diversity probably originated from long-range dispersal, and the mosaic population structure is likely a result of stochastic short-range dispersal of individuals. Such a mosaic spatial structure is apparently discordant with the current knowledge about the integrity of ant colonies. Relatedness was low in both populations when estimated among nestmates, but increased significantly when estimated among individuals sharing the same genetic cluster or haplogroup. The latter association indicates the important historical role of queen dispersal in the determination of the spatial genetic structure.

Parasites and parallel divergence of the number of individual MHC alleles between sympatric three-spined stickleback Gasterosteus aculeatus morphs in Iceland

Two pairs of sympatric three-spined stickleback Gasterosteus aculeatus morphs and two single morph populations inhabiting mud and lava or rocky benthic habitats in four Icelandic lakes were screened for parasites and genotyped for MHC class IIB diversity. Parasitic infection differed consistently between G. aculeatus from different benthic habitats. Gasterosteus aculeatus from the lava or rocky habitats were more heavily infected in all lakes. A parallel pattern was also found in individual MHC allelic variation with lava G. aculeatus morphs exhibiting lower levels of variation than the mud morphs. Evidence for selective divergence in MHC allele number is ambiguous but supported by two findings in addition to the parallel pattern observed. MHC allele diversity was not consistent with diversity reported at neutral markers (microsatellites) and in Þingvallavatn the most common number of alleles in each morph was associated with lower infection levels. In the Þingvallavatn lava morph, lower infection levels by the two most common parasites, Schistocephalus solidus and Diplostomum baeri, were associated with different MHC allele numbers.

Fourteen new microsatellite markers for Atlantic herring Clupea harengus

Fourteen new microsatellite loci were developed and tested on Atlantic herring Clupea harengus with 39 individuals from Iceland and 49 individuals from Norway. The microsatellites, which contain di, tri and tetranucleotide repeats, are polymorphic (7-30 alleles), with observed heterozygosity ranging between 0·69 and 1·00 and expected heterozygosity between 0·55 and 0·97.

Morphological variation among Betula nana (diploid), B. pubescens (tetraploid) and their triploid hybrids in Iceland

BACKGROUND AND AIMS

Introgressive hybridization between two co-existing Betula species in Iceland, diploid dwarf birch B. nana and tetraploid downy birch B. pubescens, has been well documented. The two species are highly variable morphologically, making taxonomic delineation difficult despite stable ploidy levels. Here an analysis is made of morphological variation within each ploidy group with an aim to establishing a reliable means to distinguish the species.

METHODS

Plant materials were collected from 14 woodlands in Iceland. The plants were identified based on 2n chromosome numbers. Morphological variation in species-specific characters within each ploidy group was analysed qualitatively and quantitatively. The morphological index was based on eight discrete characters, whereas the multivariate analysis was based on nine leaf variables.

KEY RESULTS

Of the 461 plants examined, 9.5 % were found to be triploid hybrids. The three ploidy groups were morphologically distinguishable but their variation overlapped. The diploid, triploid and tetraploid groups had average scores of 1.3, 4.1 and 8.3, respectively, in the morphology index scale from 0 (B. nana) to 13 (B. pubescens). A linear discriminant analysis also revealed significant separation among the three ploidy groups and the model assigned 96 % and 97 % of the B. nana and B. pubescens individuals correctly. The triploid hybrids were difficult to predict since only half of them could be assigned correctly. Leaf length was the most useful variable identifying triploid hybrids. Geographical patterns within the ploidy groups could partly be explained by differences in mean July temperature.

CONCLUSIONS

Hybridization between B. nana and B. pubescens is widespread in Iceland. The species can be distinguished from each other morphologically, and from the triploid hybrids. The overlapping morphological variation indicates bidirectional introgression between the two species via triploid hybrids. Iceland could be considered a birch hybrid zone, harbouring genetic variation which may be advantageous in subarctic regions.

Selection on a modifier of recombination rate due to linked deleterious mutations

Several models have been suggested to explain the origin and maintenance of recombination. Here I present the results from computer simulations of multilocus haploid and diploid genotypes in small populations. Each chromosome consisted of 1001 loci where deleterious mutations occurred. At "equilibrium" for mutation-selection-genetic drift balance a single recombination variant was introduced to the population in the middle of a chromosome. On average 75,000 replicates for each combination of parameters were followed to fixation or loss of the modifier allele. The results show that, in a small population, increased recombination can be selected, even in the absence of epistasis or beneficial mutations. The effect of the mutation rate for deleterious mutations depends on the ploidy level and the recessiveness of deleterious mutations. A higher deleterious mutation rate is required for an increase in recombination rate to be favored in haploid populations. Increased recombination could not evolve in the case of strong associative overdominance.

Dental anxiety in relation to mental health and personality factors. A longitudinal study of middle-aged and elderly women

Little is known about the longitudinal course of dental anxiety in relation to age, mental health and personality factors. In 1968 69 a representative sample of 778 women aged 38 to 54 yr took part in a psychiatric examination. Three hundred and ten were followed up in 1992-93. A phobia questionnaire, including assessment of dental fear, and the Eysenck Personality Inventory were distributed to the participants at both occasions. High dental fear was reported by 16.8% of the women at baseline and was associated with a higher number of other phobias, a higher level of neuroticism, more psychiatric impairment, more social disability due to phobic disorder, and a higher anxiety level. Among women who reported high dental fear in 1968 69 (n=36), 64% remitted and 36% remained fearful. Among women with low dental fear in 1968 69 (n = 274), 5% reported high dental fear in 1992-93. Chronicity was associated with higher neuroticism, lower extraversion, and more psychiatric impairment at base-line. Remission was associated with higher extraversion at baseline. Dental anxiety increased or decreased over time in concert with the number of other fears.

The effects of deleterious mutations in cyclically parthenogenetic organisms

Cyclically parthenogenetic organisms experience benefits of both sexual and asexual reproductive modes in a constant environment. Sexual reproduction generates new genotypes and may facilitate the purging of deleterious mutations whereas asexuality has a two-fold advantage and enables maintenance of well-fitted genotypes. Asexual reproduction can have a drawback as increased linkage may lead to the accumulation of deleterious mutations. This study presents the results of Monte Carlo simulations of small and infinite diploid populations, with deleterious mutations occurring at multiple loci. The recombination rate and the length of the asexual period, interrupted by sexual reproduction, are allowed to vary. Here I show that the fitness of cyclical parthenogenetic population is dependent on the length of the asexual period. Increased length of the asexual period can lead both to increased segregational load following sexual reproduction and to a stronger effect of deleterious mutations on variation at a linked neutral marker, either by reducing or increasing the variation.

The prevalence of depression in relation to cerebral atrophy and cognitive performance in 70- and 74-year-old women in Gothenburg. The Women's Health Study

BACKGROUND

Hospital-based studies report that depression in the elderly is associated with brain atrophy. This notion could not be confirmed in a population study on 85-year-olds. We aimed to assess depression in relation to brain atrophy and cognition in 70- and 74-year-old women. METHODS; A representative sample of 70- and 74-year-old women (N = 501) was examined with a psychiatric examination including the Mini-Mental State Examination (MMSE), measuring global cognitive function, and computerized tomography (CT) of the brain (N = 268). Depression was diagnosed according to DSM-III-R. Previous depression was diagnosed by history and by information from previous examinations in this 24-year longitudinal study.

RESULTS

The prevalence of depression was 11.6%, including 8.4% with major depression (MDD). Among those who were currently mentally healthy, 43.0% had a history of previous depression. Women with current MDD had lower scores on the MMSE than the mentally healthy women. This association was only found in women with a lower level of education. Current depressives, previous depressives and mentally healthy women without a history of depression did not differ on CT with regard to brain atrophy or white matter lesions. The association between MDD and lower cognitive performance was independent of the association of cognitive performance with structural brain changes on CT.

CONCLUSIONS

Brain atrophy on CT is not associated with depression in the general population, despite the fact that individuals with depression have a worse cognitive performance. The finding that cognitive performance was not decreased in individuals with previous depression suggests that cognitive dysfunction is a state phenomenon in depression.

Microsatellite variation in Daphnia pulex from both sides of the Baltic Sea

Despite large genetic differentiation among neighbouring populations of many freshwater zooplankton species, a macrogeographical homogeneity of allozyme variation is generally observed. A study on breeding systems in Scandinavian populations of Daphnia pulex suggested a latitudinally related cline in breeding system with both diploid cyclic parthenogens and diploid obligate parthenogens at the latitude of 60-61 degrees N. Variation at neutral markers may be more affected by selection at linked loci in such species than in strictly sexual species. In this paper I present a study of variation at five microsatellite loci in a total of 34 populations from small ponds and rockpools on both sides of the Baltic Sea at 60-61 degrees N. Two major groups, which may represent different species of the D. pulex complex, are defined with the microsatellites. Neighbouring populations show both similar and well differentiated genetic composition. Populations separated by larger geographical distances show only a large differentiation and a macrogeographic pattern. The large differentiation observed at small distances can be explained with small effective population size: variation at the microsatellite loci has been shaped by population bottlenecks followed with expansion in size, and possibly by selection. No conclusive evidence is found for obligative parthenogenesis.

A population study on the influence of depression on neuropsychological functioning in 85-year-olds

OBJECTIVE

To examine cognitive function in very old depressed individuals.

METHOD

Individuals with major depression (MDS) or dysthymia according to the DSM-III-R were compared to mentally healthy regarding tests of verbal ability, inductive logical reasoning, spatial ability, perceptual speed, basic arithmetics, primary memory and secondary memory in a population-based sample of 85-year-olds.

RESULTS

Individuals with MDS performed worse than mentally healthy individuals in tests of verbal ability, inductive logical reasoning, spatial ability, perceptual speed and secondary memory. There were no differences between the groups regarding basic arithmetics and primary memory. The poor test performance was mainly associated with psychomotor retardation and decreased concentration in depressed individuals. Memory complaints were not correlated to poor test performance, neither in the mentally healthy nor in the depressed.

CONCLUSION

MDS in elderly individuals is associated with reduced cognitive test performance, especially regarding more complex and time-demanding tests and in tests of secondary memory.

The effects of deleterious mutations on linked, neutral variation in small populations

The effects of recessive, deleterious mutations on genetic variation at linked neutral loci can be heterozygosity-decreasing because of reduced effective population sizes or heterozygosity-increasing because of associative overdominance. Here we examine the balance between these effects by simulating individual diploid genotypes in small panmictic populations. The haploid genome consists of one linkage group with 1000 loci that can have deleterious mutations and a neutral marker. Combinations of the following parameters are studied: gametic mutation rate to harmful alleles (U), population size (N), recombination rate (r), selection coefficient (s), and dominance (h). Tight linkage (r </= 10(-4)) gives significant associative effects, leading either to strong reduction of heterozygosity when the product Nhs is large or to a clear increase when the product Nhs is small, the boundary between these effects being 1 < Nhs < 4 in our simulations. Associative overdominance can lead to heterozygosities that are larger than predicted by the background selection models and even larger than the neutral expectation.

Depression, cerebral atrophy, cognitive performance and incidence of dementia. Population study of 85-year-olds

BACKGROUND

Hospital-based studies suggest that depression in old age relates to organic brain changes.

AIMS

To determine whether these findings are confirmed in a population-based sample.

METHOD

A population sample of non-demented 85-year-olds (227 mentally healthy and 62 with DSM-III-R depression were given a neuropsychiatric examination and computerised tomographic scans of the brain, and followed for three years.

RESULTS

On the Mini-Mental State Examination, those with a low educational level with major depression performed worse than the mentally healthy; this distinction was not evident among those who had received higher education. Measures of brain atrophy were similar in depressed and mentally healthy individuals. The three-year incidence of dementia was increased in those with early-onset major depression.

CONCLUSIONS

Higher education may protect against cognitive symptoms in depressed individuals. The association between depression and cerebral atrophy in the elderly is not very strong. The higher incidence of dementia in those with early-onset major depression may be due to a longer lifetime duration of depression, emphasising the importance of detecting and treating depression in the community.

Mitochondrial cytochrome b DNA sequence variation of Atlantic cod, Gadus morhua, from the Baltic and the White Seas

We studied sequence variation of a fragment of the mitochondrial cytochrome b gene using polymerase chain reaction (PCR) and direct sequencing of 109 Atlantic cod (Gadus morhua) from the Baltic and 19 cod from the White Sea. Eight nucleotide substitutions define 12 haplotypes differing from each other by one to six substitutions. All mutations are at silent sites and thus neutral to selection arising from protein action. Frequencies of four common haplotypes are 6%-57% and haplotype diversity is 0.64 +/- 0.031. Testing for neutral theory predictions reveals no departure from neutrality. The Baltic shares its haplotypes with distant localities in the Atlantic but differs in frequency. Within the Baltic none of the variation is explained by geographic location with an AMOVA. Temporal changes of haplotype frequencies of up to 50% are observed between sampling years and yearclasses. A potential confounding of temporal and microspatial variation cannot be excluded due to limited sample size.

Associative overdominance, heterozygosity and fitness

Our aim in this study was to examine the power of associative overdominance in creating a correlation between individual heterozygosity and fitness and in maintaining genetic polymorphism at neutral loci. This was undertaken by simulating a diploid model with five chromosomes, each with 1000 loci that could have deleterious mutations linked to neutral markers located on the same chromosomes. The simulations were carried out with various combinations of the following parameters: population size (N), number of crossovers (c), selection coefficient (s) and dominance (h). All combinations of parameter values resulted in a positive regression of the fitness on the individual marker heterozygosity, although the fitness differences were not very large when c = 2. The association between individual heterozygosity and fitness was clearest for recessive mutations (small h) with intermediate selection coefficients in small populations. The level of marker heterozygosity in the population was higher than the neutral expectation for many parameter values, even though regression of fitness on heterozygosity was not always steep at the individual level. These results agreed with the conclusion that associative overdominance could help to maintain polymorphisms in small populations. Strong selection occasionally yielded gene diversities lower than the neutral expectation, reflecting the prediction that strong selection reduces the effective population size.