Category: Statistics

PLoS One. 2021 Jun 18;16(6):e0248350. doi: 10.1371/journal.pone.0248350. eCollection 2021.

ABSTRACT

PURPOSE: To investigate the repeatability in corneal thickness (CT) and epithelial thickness (ET) measurements using spectral domain anterior segment optical coherence tomography (AS-OCT, REVO NX, Optopol) in keratoconus, and examine the effect of corneal crosslinking (CXL) on repeatability.

METHODS: A cross-sectional study of 259 eyes of 212 patients with keratoconus attending the corneal disease clinic at a university hospital tertiary referral center were enrolled. Two groups were analysed: eyes with no prior history of CXL (Group A) and eyes with prior CXL (Group B). Repeatability of measurements was assessed using the intraclass correlation coefficient (ICC) and coefficient of variation (CV).

RESULTS: In Group A, central corneal thickness (CCT) was 472.18 ± 45.41μm, and the ET was found to be the thinnest in the inferior-temporal aspect at 51.79 ± 5.97μm and thickest at the superior-nasal aspect at 56.07 ± 5.70μm. In Group B, CCT was 465.11± 42.28μm, and the ET was the thinnest at the inferior-temporal aspect at 50.63 ± 5.52μm and thickest at the superior aspect at 56.80 ± 6.39μm. When evaluating CT measurements, ICC was above 0.86 and 0.83 for Group A and Group B respectively. When evaluating ET measurements, ICC was above 0.82 for both groups. CXL had no statistically significant impact on the repeatability of measurements.

CONCLUSIONS: AS-OCT provides repeatable CT and ET measurements in the central and peripheral cornea in patients with keratoconus. Repeatability is not affected by a history of CXL.

PMID:34143790 | DOI:10.1371/journal.pone.0248350

PLoS Negl Trop Dis. 2021 Jun 18;15(6):e0009410. doi: 10.1371/journal.pntd.0009410. eCollection 2021 Jun.

ABSTRACT

BACKGROUND: Leptospirosis has been described as a biphasic disease consisting of hematogenous dissemination to major organs in the acute phase and asymptomatic renal colonization in the chronic phase. Several observational studies have suggested an association between leptospirosis and chronic kidney disease (CKD). We investigated the dynamics of leptospires and histopathological changes in the kidney to understand the relationship between them, and also investigated the extent of renal dysfunction in the acute and chronic phases of leptospirosis using a hamster model.

FINDINGS: Hamsters (n = 68) were subcutaneously infected with 1 × 104 cells of the Leptospira interrogans serovar Manilae strain UP-MMC-SM. A total of 53 infected hamsters developed fatal acute leptospirosis, and the remaining 15 hamsters recovered from the acute phase, 13 of which showed Leptospira colonization in the kidneys in the chronic phase. Five asymptomatic hamsters also had renal colonization in the chronic phase. Immunofluorescence staining showed that leptospires were locally distributed in the renal interstitium in the early acute phase and then spread continuously into the surrounding interstitium. The kidneys of the surviving hamsters in the chronic phase showed patchy lesions of atrophic tubules, a finding of chronic tubulointerstitial nephritis, which were substantially consistent with the distribution of leptospires in the renal interstitium. The degree of atrophic tubules in kidney sections correlated statistically with the serum creatinine level in the chronic phase (rs = 0.78, p = 0.01).

CONCLUSION: Subcutaneous infection with pathogenic leptospires could cause acute death or chronic leptospirosis in hamsters after surviving the acute phase. We suggest that the renal distribution of leptospires during the acute phase probably affected the extent of tubular atrophy, leading to CKD.

PMID:34143778 | DOI:10.1371/journal.pntd.0009410

PLoS One. 2021 Jun 18;16(6):e0250025. doi: 10.1371/journal.pone.0250025. eCollection 2021.

ABSTRACT

OBJECTIVE: Takayasu arteritis (TAK) is a rare immune-mediated vasculitis of the aorta and its branches. Aims were to calculate prevalence and incidence in Switzerland, to assess disease activity and performance of MR-Angiography (MRA).

METHODS: 31 patients were recorded in a database, 27 were followed prospectively up to 3 years. Prevalence was calculated based on data of the national statistical bureau. Disease activity was defined using the revised EULAR criteria. MRA depicted stenotic changes and aortic wall enhancement.

RESULTS: A disease prevalence of 14.5/1.000.000 inhabitants and an incidence of 0.3/1.000.000 per year was calculated. Aortic wall enhancement was found in 10 patients while in clinical and serological remission. EULAR criteria missed 5 patients with disease activity with isolated elevations of ESR/CRP. Arterial stenosis did not change over time in 5 cases, it improved in 2 and increased in 7. At follow-up 16 patients were treated with tocilizumab, 11/16 in monotherapy, 5 patients were treatment-free, 25/27 stayed in remission.

CONCLUSION: In addition to prevalence and incidence, our data show that MRA qualifies to detect subclinical disease activity, but, on the other hand, that EULAR criteria may miss disease activity in case of isolated elevation of ESR/CRP.

PMID:34143786 | DOI:10.1371/journal.pone.0250025

IEEE Trans Neural Netw Learn Syst. 2021 Jun 18;PP. doi: 10.1109/TNNLS.2021.3086570. Online ahead of print.

ABSTRACT

The novel 2019 Coronavirus (COVID-19) infection has spread worldwide and is currently a major healthcare challenge around the world. Chest computed tomography (CT) and X-ray images have been well recognized to be two effective techniques for clinical COVID-19 disease diagnoses. Due to faster imaging time and considerably lower cost than CT, detecting COVID-19 in chest X-ray (CXR) images is preferred for efficient diagnosis, assessment, and treatment. However, considering the similarity between COVID-19 and pneumonia, CXR samples with deep features distributed near category boundaries are easily misclassified by the hyperplanes learned from limited training data. Moreover, most existing approaches for COVID-19 detection focus on the accuracy of prediction and overlook uncertainty estimation, which is particularly important when dealing with noisy datasets. To alleviate these concerns, we propose a novel deep network named RCoNetks for robust COVID-19 detection which employs Deformable Mutual Information Maximization (DeIM), Mixed High-order Moment Feature (MHMF), and Multiexpert Uncertainty-aware Learning (MUL). With DeIM, the mutual information (MI) between input data and the corresponding latent representations can be well estimated and maximized to capture compact and disentangled representational characteristics. Meanwhile, MHMF can fully explore the benefits of using high-order statistics and extract discriminative features of complex distributions in medical imaging. Finally, MUL creates multiple parallel dropout networks for each CXR image to evaluate uncertainty and thus prevent performance degradation caused by the noise in the data. The experimental results show that RCoNetks achieves the state-of-the-art performance on an open-source COVIDx dataset of 15,134 original CXR images across several metrics. Crucially, our method is shown to be more effective than existing methods with the presence of noise in the data.

PMID:34143745 | DOI:10.1109/TNNLS.2021.3086570

PLoS Comput Biol. 2021 Jun 18;17(6):e1009089. doi: 10.1371/journal.pcbi.1009089. Online ahead of print.

ABSTRACT

The advent of high-throughput metagenomic sequencing has prompted the development of efficient taxonomic profiling methods allowing to measure the presence, abundance and phylogeny of organisms in a wide range of environmental samples. Multivariate sequence-derived abundance data further has the potential to enable inference of ecological associations between microbial populations, but several technical issues need to be accounted for, like the compositional nature of the data, its extreme sparsity and overdispersion, as well as the frequent need to operate in under-determined regimes. The ecological network reconstruction problem is frequently cast into the paradigm of Gaussian Graphical Models (GGMs) for which efficient structure inference algorithms are available, like the graphical lasso and neighborhood selection. Unfortunately, GGMs or variants thereof can not properly account for the extremely sparse patterns occurring in real-world metagenomic taxonomic profiles. In particular, structural zeros (as opposed to sampling zeros) corresponding to true absences of biological signals fail to be properly handled by most statistical methods. We present here a zero-inflated log-normal graphical model (available at https://github.com/vincentprost/Zi-LN) specifically aimed at handling such “biological” zeros, and demonstrate significant performance gains over state-of-the-art statistical methods for the inference of microbial association networks, with most notable gains obtained when analyzing taxonomic profiles displaying sparsity levels on par with real-world metagenomic datasets.

PMID:34143768 | DOI:10.1371/journal.pcbi.1009089

PLoS One. 2021 Jun 18;16(6):e0246913. doi: 10.1371/journal.pone.0246913. eCollection 2021.

ABSTRACT

Video games have become a ubiquitous part of demographically diverse cultures. Numerous studies have focused on analyzing the cognitive aspects involved in game playing that could help in providing an optimal gaming experience by improving video game design. To this end, we present a framework for classifying the game player’s expertise level using wearable electroencephalography (EEG) headset. We hypothesize that expert and novice players’ brain activity is different, which can be classified using frequency domain features extracted from EEG signals of the game player. A systematic channel reduction approach is presented using a correlation-based attribute evaluation method. This approach lead us in identifying two significant EEG channels, i.e., AF3 and P7, among fourteen channels available in Emotiv EPOC headset. In particular, features extracted from these two EEG channels contributed the most to the video game player’s expertise level classification. This finding is validated by performing statistical analysis (t-test) over the extracted features. Moreover, among multiple classifiers used, K-nearest neighbor is the best classifier in classifying game player’s expertise level with a classification accuracy of up to 98.04% (without data balancing) and 98.33% (with data balancing).

PMID:34143774 | DOI:10.1371/journal.pone.0246913

Sci Prog. 2021 Apr-Jun;104(2):368504211025873. doi: 10.1177/00368504211025873.

ABSTRACT

The coronavirus 2019 (COVID-19) pandemic has caused dramatic changes in the daily lives of Romanians, affecting their mental health. The COVID-19 pandemic has evolved at three significant peaks, which sequentially occurred on: April 29, 2020; September 18, 2020; and the third wave registered the highest severity on November 27, 2020. Little is known about the mental health changes during this phase of this pandemic. This study evaluated mental health levels in Romania at the end of the first wave of the pandemic and amidst the third and most severe wave. We administered a two-phase internet-based survey among 543 and 583 participants, respectively, recruited through snowball sampling at a 6-month interval. The IPAT Anxiety Scale measured anxiety, the Beck’s Depression Inventory measured depression, and the Dissociative Experiences Scale measured dissociation. We observed no statistically significant differences in the number of participants with clinically relevant scores at either time point. In the first survey, 23.8%, 19.2%, and 32.6% reported being clinically anxious, clinically depressed, and showed clinical dissociation, respectively. Binary logistic regressions indicated that age, education level, and previous traumatic events were significantly associated with clinical levels of anxiety and depression. Moreover, multiple linear regression analysis reported a collective significant effect of gender, age, psychological impact, traumatic events, and dissociation on predicting high levels of anxiety and depression. Romanian adults’ mental health status was affected during the COVID-19 pandemic, and it did not change 6 months after the first lockdown.

PMID:34143706 | DOI:10.1177/00368504211025873

Sci Prog. 2021 Apr-Jun;104(2):368504211026131. doi: 10.1177/00368504211026131.

ABSTRACT

Commonly used fastener positioning methods include pixel statistics (PS) method and template matching (TM) method. For the PS method, it is difficult to judge the image segmentation threshold due to the complex background of the track. For the TM method, the search in both directions of the global is easily affected by complex background, as a result, the locating accuracy of fasteners is low. To solve the above problems, this paper combines the PS method with the TM method and proposes a new fastener positioning method called local unidirectional template matching (LUTM). First, the rail positioning is achieved by the PS method based on the gray-scale vertical projection. Then, based on the prior knowledge, the image of the rail and the surrounding area of the rail is obtained which is referred to as the 1-shaped rail image; then, the 1-shaped rail image and the produced offline symmetrical fastener template is pre-processed. Finally, the symmetrical fastener template image is searched from top to bottom along the rail and the correlation is calculated to realize the fastener positioning. Experiments have proved that the method in this paper can effectively realize the accurate locating of the fastener for ballastless track and ballasted track at the same time.

PMID:34143708 | DOI:10.1177/00368504211026131

Ecol Lett. 2021 Jun 18. doi: 10.1111/ele.13820. Online ahead of print.

ABSTRACT

Exploring and accounting for the emergent properties of ecosystems as complex systems is a promising horizon in the search for general processes to explain common ecological patterns. For example the ubiquitous hollow-curve form of the species abundance distribution is frequently assumed to reflect ecological processes structuring communities, but can also emerge as a statistical phenomenon from the mathematical definition of an abundance distribution. Although the hollow curve may be a statistical artefact, ecological processes may induce subtle deviations between empirical species abundance distributions and their statistically most probable forms. These deviations may reflect biological processes operating on top of mathematical constraints and provide new avenues for advancing ecological theory. Examining ~22,000 communities, we found that empirical SADs are highly uneven and dominated by rare species compared to their statistical baselines. Efforts to detect deviations may be less informative in small communities-those with few species or individuals-because these communities have poorly resolved statistical baselines. The uneven nature of many empirical SADs demonstrates a path forward for leveraging complexity to understand ecological processes governing the distribution of abundance, while the issues posed by small communities illustrate the limitations of using this approach to study ecological patterns in small samples.

PMID:34142760 | DOI:10.1111/ele.13820

Health Rep. 2021 Jun 16;32(6):3-13. doi: 10.25318/82-003-x202100600001-eng.

ABSTRACT

BACKGROUND: Few quantitative studies have used national-level data to examine access to mental health consultation (MHC) by immigrants in Canada, and even fewer studies investigate MHCs using the following variables: immigrant admission category, duration in Canada since landing and world source regions. This study examines MHCs by immigrants and refugees-compared with those of Canadian-born respondents-while controlling for self-reported mental health (SRMH) and immigrant characteristics, using a population-based survey linked to immigrant landing information. This study, which is based on a linked database, allows for much richer insight into immigrant populations than most previous studies.

DATA AND METHODS: Based on data from four cycles (2011 to 2014) of the Canadian Community Health Survey linked to data from the Longitudinal Immigration Database, the odds ratios of having had MHCs are compared between the Canadian-born population and immigrants by immigration dimensions, while controlling for SRMH. Results are hierarchically adjusted for age, sex, socioeconomic factors and sense of belonging.

RESULTS: After the above-mentioned factors were controlled for, immigrants were much less likely than Canadian-born respondents to access MHCs. Specifically, compared with the Canadian-born population that had high levels of SRMH, immigrants with high levels of SRMH were statistically less likely to have had an MHC (odds ratio [OR]=0.5, 95% confidence interval [CI] from 0.4 to 0.5), while those with low SRMH levels were more likely to report an MHC (OR=4.8, 95% CI from 4.5 to 5.1, for the Canadian-born population but OR=1.8, 95% CI from 1.5 to 2.1, for immigrants). Most Asian immigrants with low SRMH levels were only as likely to report MHCs as Canadian-born respondents with high SRMH levels. Refugees with low SRMH levels also had only a slightly elevated MHC level (OR=1.6, 95% CI from 1.1 to 2.3) compared with Canadian-born individuals with high SRMH levels. Overall, refugees were not more likely than immigrants of other admission categories to report having had an MHC, even though previous findings have shown that refugees report low levels of SRMH.

DISCUSSION: This study provides new evidence on the differences in access to MHC between Canadian-born individuals and immigrants by various characteristics, while controlling for SRMH. Results probably reflect the structural or cultural barriers to MHC and point to a possible pathway to either maintain or improve mental health among immigrants.

PMID:34142786 | DOI:10.25318/82-003-x202100600001-eng