Category: Statistics

Chin Med J (Engl). 2021 Feb 9;134(5):564-572. doi: 10.1097/CM9.0000000000001400.

ABSTRACT

BACKGROUND: The pathogenesis of osteosarcoma (OS) is still unclear, and it is still necessary to find new targets and drugs for anti-OS. This study aimed to investigate the role and mechanism of the anti-OS effects of miR-296-5p.

METHODS: We measured the expression of miR-296-5p in human OS cell lines and tissues. The effect of miR-296-5p and its target gene staphylococcal nuclease and tudor domain containing 1 on proliferation, migration, and invasion of human OS lines was examined. The Student’s t test was used for statistical analysis.

RESULTS: We found that microRNA (miR)-296-5p was significantly downregulated in OS cell lines and tissues (control vs. OS, 1.802 ± 0.313 vs. 0.618 ± 0.235, t = 6.402, P < 0.01). Overexpression of miR-296-5p suppressed proliferation, migration, and invasion of OA cells. SND1 was identified as a target of miR-296-5p by bioinformatic analysis and dual-luciferase reporter assay. Overexpression of SND1 abrogated the effects induced by miR-296-5p upregulation (miRNA-296-5p vs. miRNA-296-5p + SND1, 0.294 ± 0.159 vs. 2.300 ± 0.277, t = 12.68, P = 0.003).

CONCLUSION: Our study indicates that miR-296-5p may function as a tumor suppressor by targeting SND1 in OS.

PMID:33652459 | DOI:10.1097/CM9.0000000000001400

J Stroke Cerebrovasc Dis. 2021 Feb 27;30(5):105680. doi: 10.1016/j.jstrokecerebrovasdis.2021.105680. Online ahead of print.

ABSTRACT

INTRODUCTION: Body lateropulsion (BLP) is seen in neurological lesions involving the pathways responsible for body position and verticality. We report a case of isolated body lateropulsion (iBLP) as the presentation of lateral medullary infarction and conducted a systematic literature review.

METHODS: MEDLINE and EMBASE databases were searched up to December 3, 2020.

INCLUSION CRITERIA: age ≥ 18, presence of BLP, confirmed stroke on imaging.

EXCLUSION CRITERIA: age < 18, qualitative reviews, studies with inadequate patient data. Statistical analysis was performed using IBM® SPSS® Statistics 20.

RESULTS: A 64-year-old man presented with acute-onset iBLP. Brain MRI demonstrated acute infarction in the right caudolateral medulla. His symptoms progressed with ipsilateral Horner syndrome over the next 24 hours and contralateral hemisensory loss 10 days later. Repeat MRI showed an increase in infarct size. BLP resolved partially at discharge. Systematic review: 418 abstracts were screened; 59 studies were selected reporting 103 patients. Thirty-three patients had iBLP (32%). BLP was ipsilateral to stroke in 70 (68%) and contralateral in 32 (32%). The most common stroke locations were medulla (n = 63, 59%), pons (n = 16, 15%), and cerebellum (n = 16, 15%). Four strokes were cortical, 3 frontal and 1 temporoparietal (3%). The most common etiology was large-artery atherosclerosis (LAA) in 20 patients (32%), followed by small-vessel occlusion in 12 (19%). Seventeen (27%) had large-vessel occlusion (LVO), 12 involving the vertebral artery. Sixty (98%) had some degree of resolution of BLP; complete in 41 (70%). Median time-to-resolution was 14 days (IQR 10-21). There was no relationship between time-to-resolution and age, sex, side of BLP or side of stroke.

CONCLUSION: BLP was commonly seen with medullary infarction and was the isolated finding in one-third. LAA and LVO were the most common etiologies. Recovery of BLP was early and complete in most cases.

PMID:33652344 | DOI:10.1016/j.jstrokecerebrovasdis.2021.105680

Sex Reprod Healthc. 2021 Feb 24;28:100608. doi: 10.1016/j.srhc.2021.100608. Online ahead of print.

ABSTRACT

OBJECTIVE: To examine the socio-demographic variations in overwhelming existence of C-section deliveries in south India, with a comparison to rest of India.

METHODS: This study analyses data collected from 51,136 mothers under National Family Health Survey (NFHS)-3 (2005-06) and 2,52,183 mothers under NFHS-4 (2015-16), those who have given births during last five years preceding the survey.

MAIN OUTCOME MEASURES: Descriptive statistics, bivariate analysis with Chi-squared tests and binary logistic regression models with 95% confidence intervals are used.

RESULTS: In south India at least one out of four women deliver through C-section and there was a notable rise in caesarean deliveries in public facilities as well as among tribal population. In aggregate, number of states exceeding 15% prevalence rate of C-section deliveries doubled to sixteen, while nineteen states registered more than 100% rise. Rural-urban difference is slim in south India, while likelihood for C-section deliveries for richest women as compared to poorest has gone down from 2.76 to 1.88 in south India and 7.75 to 4.58 in other regions during 2005-06 to 2015-16. The odds ratio for C-section is higher in private hospitals (3.26) of southern states with reference to public institutions, while the odds are 3.90 times higher for private facilities in other states. In south India, percentage of C-section deliveries were actually lower among those who reported about pregnancy complications.

CONCLUSIONS: Despite, several maternal and child health related programs being launched in India, their impact on improving the C-section scenario has remained microscopic, or they have continued to contribute towards a rising prevalence of C-section, especially in south India.

PMID:33652351 | DOI:10.1016/j.srhc.2021.100608

J Crit Care. 2021 Feb 20;63:98-103. doi: 10.1016/j.jcrc.2021.02.004. Online ahead of print.

ABSTRACT

PURPOSE: Vascular access patterns in the intensive care unit (ICU) have shifted from non-tunneled central venous catheters (CVCs) towards peripherally inserted central catheters (PICCs). We evaluated perceptions of critical care practitioners regarding these devices and variation in evidence-based practice.

MATERIALS: A 35-question survey on ICU vascular access was deployed in 13 Michigan hospitals. Descriptive statistics summarized responses. Differences in utilization, perceptions and evidence-based practices between PICCs and CVCs, by participant and site-level characteristics, were assessed.

RESULTS: 314 of 621 eligible providers responded to the survey (response rate 51%). 15% of providers reported not routinely using ultrasound when placing CVCs. Respondents whom were trainees, from larger hospitals, and from closed ICUs were more likely to use ultrasound (p < 0.001). Additionally, 21% of respondents stated they did not specify number of CVC lumens, while 46% did not specify number of PICC lumens (p < 0.001). The likelihood of specifying PICC lumens increased when vascular access protocols were in place (p = 0.001). 2/3 of respondents (n = 173, 66%) stated more research on ICU vascular access was needed.

CONCLUSION: Variation in guideline-based vascular access practices exists in the ICU. Defined local protocols may improve guideline adherence. Studies evaluating vascular access decisions and patient safety in the ICU appear necessary.

PMID:33652363 | DOI:10.1016/j.jcrc.2021.02.004

Neural Netw. 2021 Feb 17;138:140-149. doi: 10.1016/j.neunet.2021.02.007. Online ahead of print.

ABSTRACT

Few-shot learning aims to classify unseen classes with a few training examples. While recent works have shown that standard mini-batch training with carefully designed training strategies can improve generalization ability for unseen classes, well-known problems in deep networks such as memorizing training statistics have been less explored for few-shot learning. To tackle this issue, we propose self-augmentation that consolidates self-mix and self-distillation. Specifically, we propose a regional dropout technique called self-mix, in which a patch of an image is substituted into other values in the same image. With this dropout effect, we show that the generalization ability of deep networks can be improved as it prevents us from learning specific structures of a dataset. Then, we employ a backbone network that has auxiliary branches with its own classifier to enforce knowledge sharing. This sharing of knowledge forces each branch to learn diverse optimal points during training. Additionally, we present a local representation learner to further exploit a few training examples of unseen classes by generating fake queries and novel weights. Experimental results show that the proposed method outperforms the state-of-the-art methods for prevalent few-shot benchmarks and improves the generalization ability.

PMID:33652370 | DOI:10.1016/j.neunet.2021.02.007

Sci Total Environ. 2021 Feb 18;776:145935. doi: 10.1016/j.scitotenv.2021.145935. Online ahead of print.

ABSTRACT

Data-driven landslide susceptibility models formally integrate spatial landslide information with explanatory environmental variables that describe predisposing factors of slope instability. Well-performing models are commonly utilized to identify landslide-prone terrain or to understand the causes of slope instability. In most cases, however, the available landslide data is affected by spatial biases (e.g. underrepresentation of landslides far from infrastructure or in forests) and does therefore not perfectly represent the spatial distribution of past slope instabilities. Literature shows that implications of such data flaws are frequently ignored. This study was built upon landslide information that systematically relates to damage-causing and infrastructure-threatening events in South Tyrol, Italy (7400 km²). The created models represent three conceptually different strategies to deal with biased landslide information. The aims were to demonstrate why an inference of geomorphic causation from apparently well-performing models is invalid under common landslide data bias conditions (Model 1), to test a novel bias-adjustment approach (Model 2) and to exploit the underlying data bias to model areas likely affected by potentially damaging landslides (Model 3; intervention index), instead of landslide susceptibility. The study offers a novel perspective on how biases in landslide data can be considered within data-driven models by focusing not only on the process under investigation (landsliding), but also on the circumstances that led to the registration of landslide information (data collection effects). The results were evaluated in terms of statistical relationships, variable importance, predictive performance, and geomorphic plausibility. The results revealed that none of the models reflected landslide susceptibility. Despite partly high predictive performances, the models were unable to create geomorphically plausible spatial predictions. The impact-oriented intervention index, however, enabled to identify damage-causing landslides with high accuracy. We conclude that the frequent practice of inferring geomorphic causation from well-performing models without accounting for data limitations is invalid.

PMID:33652311 | DOI:10.1016/j.scitotenv.2021.145935

Clin Imaging. 2021 Feb 24;77:86-91. doi: 10.1016/j.clinimag.2021.02.031. Online ahead of print.

ABSTRACT

OBJECTIVE: To investigate whether textural analysis (TA) of MRI heterogeneity may play a role in the clinical assessment and classification of breast tumors.

MATERIALS AND METHODS: For this retrospective study, patients with breast masses ≥1 cm on contrast-enhanced MRI were obtained in 69 women (mean age: 51 years; range 21-78 years) with 77 masses (38 benign, 39 malignant) from 2006 to 2018. The selected single slice sagittal peak post-contrast T1-weighted image was analyzed with commercially available TA software [TexRAD Ltd., UK]. Eight histogram TA parameters were evaluated at various spatial scaling factors (SSF) including mean pixel intensity, standard deviation of the pixel histogram (SD), entropy, mean of the positive pixels (MPP), skewness, kurtosis, sigma, and Tx_sigma. Additional statistical tests were used to determine their predictiveness.

RESULTS: Entropy showed a significant difference between benign and malignant tumors at all textural scales (p < 0.0001) and kurtosis was significant at SSF = 0-5 (p = 0.0026-0.0241). The single best predictor was entropy at SSF = 4 with AUC = 0.80, giving a sensitivity of 95% and specificity of 53%. An AUC of 0.91 was found using a model combining entropy with sigma, which yielded better performance with a sensitivity of 92% and specificity of 79%.

CONCLUSION: TA of breast masses has the potential to assist radiologists in categorizing tumors as benign or malignant on MRI. Measurements of entropy, kurtosis, and entropy combined with sigma may provide the best predictability.

PMID:33652269 | DOI:10.1016/j.clinimag.2021.02.031

Arch Oral Biol. 2021 Feb 23;125:105091. doi: 10.1016/j.archoralbio.2021.105091. Online ahead of print.

ABSTRACT

OBJECTIVES: To evaluate in vitro the effects of adrenaline and noradrenaline on the biofilm formation on orthodontic brackets, acid production and expression of virulence genes of Streptococcus mutans UA159 (S. mutans).

DESIGN: S. mutans UA159 biofilm was formed on orthodontic brackets under exposure to adrenaline (100 μM), noradrenaline (50 μM) or PBS solution (control group) in triptone-yeast extract with 1 % sucrose. After 24 h, biofilm formation was quantified through Colony Forming Units / mL (CFU/mL) and RNA was extracted to perform gene expression analysis through real-time reverse transcriptase-PCR (RT-qPCR). Evaluation of acid production was carried out on planktonic cultures for 6 h. One-way ANOVA followed by Tukey’s test was carried to determine statistical difference. The level of significance was set at 5 %.

RESULTS: Catecholamines stimulated biofilm formation of S. mutans in orthodontic brackets (p < 0,05) but did not interfere with acid production (pH reduction) or the expression of the tested genes related to biofilm formation (gtfB, gtfC, gbpA, gbpB, gbpC, gbpD and brpA), aciduric (relA) and acidogenic properties (ldh).

CONCLUSIONS: The present study was the first to demonstrate that catecholamines can stimulate S. mutans UA159 biofilm formation. These findings can contribute to clarify the role of stress on bacterial metabolism and contribute to the understanding of a possible role on caries development, mainly in orthodontic patients.

PMID:33652302 | DOI:10.1016/j.archoralbio.2021.105091

Sci Total Environ. 2021 Feb 23;776:146019. doi: 10.1016/j.scitotenv.2021.146019. Online ahead of print.

ABSTRACT

The inadequate management of soils and the absence of conservation practices favor the degradation of pastures and can trigger adverse environmental alterations and damage under the terms of Brazilian Federal Law no. 6.938/1981. Based on this premise, this study aimed to estimate soil losses caused by water erosion in pasture areas using the brightness index (BI) from the annual series of Landsat 8 images in different geological formations. A specifically prepared Google Earth Engine (GEE) script automatically extracted the BI from the images. The study occurred in the Environmental Protection Area (EPA) of Uberaba River basin (Minas Gerais, Brazil). To accomplish the goal, 180 digital 500-wide random buffers were selected from 3 geologic types (60 points per type), and then analyzed for zonal statistics of USLE (Universal Soil Loss Equation) soil loss and BI in a Geographic Information System. The regression models BI versus USLE soil loss allowed estimating BI soil losses over the pastures of EPA. The model fittings were remarkable. The validation of soil loss maps in the EPA occurred in pasture phytophysiognomies through the probing of penetration resistance in 37 randomly selected locations. The results were satisfactory, mostly those based on the BI. The BI losses increased for greater resistances. Amplified losses also occurred in regions exposed to environmental land use conflicts (actual uses that deviate from land capability or natural use). Overall, the BI approach proved efficient to accurately track soil losses and pasture degradation over large areas, with the advantage of standing on a single parameter easily accessed through remote sensed data. From an environmental standpoint, this is an important result, because the accurate diagnosis and prognosis of degraded pastures is paramount to implement mitigation measures following the “polluter pays principle”, even more in Brazil where the areas occupied by degraded pastures are enormous.

PMID:33652307 | DOI:10.1016/j.scitotenv.2021.146019

Sci Total Environ. 2021 Feb 18;776:145926. doi: 10.1016/j.scitotenv.2021.145926. Online ahead of print.

ABSTRACT

The estimation of geochemical background is complex in areas impacted by point sources of atmospheric emissions due to unknowns about pollutant dispersion, persistence of pollutants on the landscape, and natural concentrations of elements associated with parent material. This study combined mineralogical analysis with conventional statistical and geospatial methods to separate anthropogenically impacted soils from unimpacted soils in the Yellowknife area, Northwest Territories, Canada, a region that was exposed to 60 years of arsenic (As)-rich atmospheric mining emissions (1938-1999) and that hosts natural enrichments of As. High concentrations of As (up to 4700 mg kg^-1) were measured in publicly accessible soils near decommissioned roaster stacks in the region and strong relationships between As and distance from the main emission sources persisted in surface soils and soils at depth in the soil profile more than 60 years after the bulk of mining emissions were released. Mineralogical analysis provided unambiguous evidence regarding the source of As minerals and highlighted that most As in surface soils within 15 km of Yellowknife is hosted as anthropogenic arsenic trioxide (As₂O₃), produced by roaster stack emissions. Statistical protocols for the estimation of geochemical background were applied to an existing database of till geochemistry (N = 1490) after removing samples from mining impacted areas. Results suggested geochemical background for the region is 0.25-15 mg kg^-1 As, comparable to global averages, with upper thresholds elevated in volcanic units (30 mg kg^-1 As) that often host sulfide mineralization in greenstone belts in the region.

PMID:33652309 | DOI:10.1016/j.scitotenv.2021.145926