Tag: statistics

Sci Rep. 2022 Apr 7;12(1):5900. doi: 10.1038/s41598-022-09938-8.

ABSTRACT

Recently, computational modeling has shifted towards the use of statistical inference, deep learning, and other data-driven modeling frameworks. Although this shift in modeling holds promise in many applications like design optimization and real-time control by lowering the computational burden, training deep learning models needs a huge amount of data. This big data is not always available for scientific problems and leads to poorly generalizable data-driven models. This gap can be furnished by leveraging information from physics-based models. Exploiting prior knowledge about the problem at hand, this study puts forth a physics-guided machine learning (PGML) approach to build more tailored, effective, and efficient surrogate models. For our analysis, without losing its generalizability and modularity, we focus on the development of predictive models for laminar and turbulent boundary layer flows. In particular, we combine the self-similarity solution and power-law velocity profile (low-fidelity models) with the noisy data obtained either from experiments or computational fluid dynamics simulations (high-fidelity models) through a concatenated neural network. We illustrate how the knowledge from these simplified models results in reducing uncertainties associated with deep learning models applied to boundary layer flow prediction problems. The proposed multi-fidelity information fusion framework produces physically consistent models that attempt to achieve better generalization than data-driven models obtained purely based on data. While we demonstrate our framework for a problem relevant to fluid mechanics, its workflow and principles can be adopted for many scientific problems where empirical, analytical, or simplified models are prevalent. In line with grand demands in novel PGML principles, this work builds a bridge between extensive physics-based theories and data-driven modeling paradigms and paves the way for using hybrid physics and machine learning modeling approaches for next-generation digital twin technologies.

PMID:35393511 | DOI:10.1038/s41598-022-09938-8

Sci Rep. 2022 Apr 7;12(1):5905. doi: 10.1038/s41598-022-09994-0.

ABSTRACT

Hearing is one of the earliest senses to develop and is quite mature by birth. Contemporary theories assume that regularities in sound are exploited by the brain to create internal models of the environment. Through statistical learning, internal models extrapolate from patterns to predictions about subsequent experience. In adults, altered brain responses to sound enable us to infer the existence and properties of these models. In this study, brain potentials were used to determine whether newborns exhibit context-dependent modulations of a brain response that can be used to infer the existence and properties of internal models. Results are indicative of significant context-dependence in the responsivity to sound in newborns. When common and rare sounds continue in stable probabilities over a very long period, neonates respond to all sounds equivalently (no differentiation). However, when the same common and rare sounds at the same probabilities alternate over time, the neonate responses show clear differentiations. The context-dependence is consistent with the possibility that the neonate brain produces more precise internal models that discriminate between contexts when there is an emergent structure to be discovered but appears to adopt broader models when discrimination delivers little or no additional information about the environment.

PMID:35393525 | DOI:10.1038/s41598-022-09994-0

J Phys Condens Matter. 2022 Apr 7. doi: 10.1088/1361-648X/ac655b. Online ahead of print.

ABSTRACT

Complex band structure emerges when translational symmetry is broken and material states with complex wavevectors become admissible. The resulting complex bands continuously connect conventional bands and their shapes are directly related to measurable physical quantities. To date, interpretations of complex bands usually assume they are semielliptical because this is the shape produced by the Su-Schrieffer-Heeger model. However, numerous studies have reported complex band structures with distinctly non-semielliptical shapes, including loops (essentially deformed, asymmetric semiellipses), spikes, and vertical lines. The primary goal of this work is to explore the phenomenology of these shapes such that deeper physical insight can be obtained from a qualitative inspection of a material’s complex band structure. By using several variations on the Su-Schrieffer-Heeger model, we find that (i) vertical lines are unphysical numerical artifacts, (ii) spikes indicate perfectly evanescent states in the material that couple adjacent layers but do not transfer amplitude, and (iii) asymmetric loops result from hybridization. Secondarily, we also develop a strategy for eliminating any unphysical vertical lines from calculations, thereby improving computational techniques for complex band structure.

PMID:35390781 | DOI:10.1088/1361-648X/ac655b

Epilepsy Res. 2022 Mar 31;182:106917. doi: 10.1016/j.eplepsyres.2022.106917. Online ahead of print.

ABSTRACT

OBJECTIVE: Epilepsies are severe chronic neurological diseases that impair several domains in life and are often accompanied by various somatic and psychiatric comorbidities. Associations between epilepsy and its comorbidities remain poorly understood. As epidemiological research mainly relies on cross-sectional designs and descriptive results, homogeneities regarding comorbidities in individuals suffering from epilepsy remain uncovered. Therefore, we aimed to identify clusters of individuals based on selected seizure-related variables and somatic comorbidities, and their respective risk of experiencing affective disorders, using a Latent Class Analysis (LCA).

METHODS: Latent class analysis, is a model-driven statistical approach, which aims at latent, unobservable clusters on selected disease features. LCA has therefore the potential for uncovering previously unobservable groups or classes with similar comorbidity patterns. It allows for comparisons between those classes regarding risk or promotive factors – such as affective disorders. Our data derives from the Austrian cohort of the European Study on Burden and Care of Epilepsy (ESBACE; http://www.esbace.eu/). In ESBACE, multiple factors were collected to get a detailed picture on prevalence, epilepsy-related variables and comorbidities in a population-based cohort from the region of Salzburg, Austria. We used LCA to identify epilepsy-somatic-comorbidity-clusters and further, compared them to the observed the risk of suffering from affective disorders.

RESULTS: The prevalence of epilepsy in the study region was 9.14/1000 inhabitants. LCA unveiled a three-cluster solution, of which one cluster, mainly consisting of individuals with mixed seizure types, higher age, and discrete somatic comorbidities (stroke, cardiovascular – and respiratory/pulmonary diseases) had a higher risk of experiencing affective disorders.

SIGNIFICANCE: To our knowledge, this is the first large scale study that uses LCA to identify epilepsy-related comorbidity phenotypes, and therefore it might open a new way for epidemiological research.

PMID:35390702 | DOI:10.1016/j.eplepsyres.2022.106917

Sleep Med. 2022 Feb 19;92:96-102. doi: 10.1016/j.sleep.2022.02.011. Online ahead of print.

ABSTRACT

INTRODUCTION: Adenoid and tonsillar hypertrophy in children often leads to adverse respiratory symptoms and obstructive sleep apnea (OSA). Current clinical guidelines from the American Academy of Pediatrics and American Academy of Otolaryngology-Head and Neck Surgery recommend tonsillectomy as the first line of pediatric OSA treatment for children with tonsillar hypertrophy. Rapid palatal expansion (RPE) performed by orthodontists improves obstructive sleep apnea in children by reducing nasal airway resistance, increasing nasal volume, raising tongue posture, and enlarging pharyngeal airway. However, the role of RPE in alleviating adenoid and tonsillar hypertrophy remains elusive. In this study, we aim to evaluate the changes in adenoid and palatine tonsil sizes following RPE using 3D volumetric analysis of cone beam computational tomography (CBCT) imaging.

MATERIALS AND METHODS: In this retrospective cohort study, a total of 60 pediatric patients (mean age: 8.00, range: 5-15, 32 females and 28 males) who had tonsillar hypertrophy (size 3 and 4) were included and divided into the control group (n = 20) and expansion group (n = 40). The control group did not undergo any treatment. The expansion group underwent RPE using a conventional Hyrax expander, activated 0.25 mm per day for 4-6 weeks. Final CBCT scans (T2) were performed 13.8 ± 6.5 months after the initial scan (T1). Pediatric sleep questionnaire (PSQ) and BMI were obtained at each timepoint. Volumetric analysis of adenoid and palatine tonsils was performed using a combination of bony and soft tissue landmarks in CBCT scans through Anatomage Invivo 6 imaging software. Paired t-tests were used to evaluate the difference between the initial and final adenoid and tonsil volumes. p values less than 0.05 were considered statistically significant.

RESULTS: Compared to the control group, the expansion group experienced a statistically significant decrease in both adenoid and tonsil volume. There was non-statistically significant increase in volume from T1 to T2 for the control group. For the expansion group, 90.0% and 97.5% of patients experienced significant reduction in adenoid and tonsil volume, respectively. The average volume decrease of adenoids was 16.8% while that of tonsils was 38.5%. The patients had up to 51.6% and 75.4% reduction in adenoid and tonsil size, respectively, following RPE orthodontic treatment. Pearson correlation ranged from 0.88 to 0.99 for each measurement, representing excellent internal consistency. There was a significant reduction in the PSQ scores from 5.81 ± 3.31 to 3.75 ± 2.38 in expansion group (p < 0.001).

CONCLUSIONS: Our results demonstrated that RPE significantly reduced the size of both adenoid and palatine tonsils and revealed another long-term benefit of RPE treatment. To our knowledge, this is the first study to quantify the changes of adenoids and tonsils following RPE. RPE treatment can be considered as a valid and effective treatment option for pediatric OSA population with narrow high arch palate and adenotonsillar hypertrophy.

PMID:35390750 | DOI:10.1016/j.sleep.2022.02.011

J Environ Manage. 2022 Apr 4;313:114988. doi: 10.1016/j.jenvman.2022.114988. Online ahead of print.

ABSTRACT

Detailed information on spatial distribution of croplands and grain yields is crucial for agricultural management and food security, but is often limited by a lack of geospatial data. By integrating satellite observation and statistical data, this study produced new geospatial data of cropland areas and grain yields in China during 2000-2020. We found that the decrease of relatively high-yielding croplands in southern China mainly caused by the expansion of constructed land. Yet, the increase of croplands largely occurred in temperature/water-limited regions of Northern Arid and Semiarid Region (NASR) and Northeast China Plain (NCP). Croplands’ decrease in southern China and expansion in NCP and NASR jointly led to the continuous northward shift of the centre of gravity of croplands and grain yields. This spatial transfer of croplands resulted in relatively lower-than- average grain yield per unit area (AGYA) croplands decreasing from 38.96% (2000) to 35.75% (2020), but also relatively higher-than-AGYA croplands decreasing from 38.41% (2000) to 35.01% (2020), implying spatial challenges of grain production. Generally, every 1 km² of cropland loss in traditional high-yield zones required nearly 1-3 times expansion in area in NASR and NCP to balance grain yield losses. The new geospatial data and these findings from this study could provide valuable information for regional agriculture development and policy marking.

PMID:35390663 | DOI:10.1016/j.jenvman.2022.114988

Forensic Sci Int Genet. 2022 Mar 21;59:102691. doi: 10.1016/j.fsigen.2022.102691. Online ahead of print.

ABSTRACT

The interpretation of mixtures containing related individuals can be difficult due to allele sharing between the contributors. Challenges include the assignment of the number of contributors (NoC) to the mixture with the under assignment of NoC resulting in false exclusions of true donors. Non-donating relatives of the true contributors to mixtures of close relatives can result in likelihood ratios supporting their adventitious inclusion within the mixture. We examine the effect of non-donor likelihood ratios on mixtures of first order relatives. Mixtures of full siblings and parent-child were created by mixing the DNA from known family members in vitro, or by in silico simulation. Mixtures were interpreted using the probabilistic genotyping software STRmix™ and likelihood ratios were assigned for the true donors and non-donors who were either further relatives of the true donors or unrelated to the true donors. The two donor balanced mixtures deconvoluted straightforwardly when analysed as NoC = 2 giving approximately the experimental design 1:1 ratio. When analysed as NoC = 3 a very large number of non-donor genotypes produced LRs close to 1 including many instances of adventitious support. The in vitro three donor balanced mixtures proved difficult to assign as NoC = 3 by a blind examination of the profile. It is likely that many of these would be misassigned as NoC = 2. The analysis of the in vitro and in silico mixtures assuming NoC = 3 with no use of a conditioning profile or with the use of a conditioning profile but without informed priors on the mixture proportions (Mx priors) was ineffective. If the profile can be assigned as NoC = 3 then assignment of the Mx priors is straightforward. This analysis gave no false exclusions. Adventitious support did happen for relatives with high allele sharing. Adventitious support was not observed for any unrelated non-donors. The analysis of the three-person mixtures as NoC = 2 produced many false exclusions and fewer instances of adventitious support. The three donor unbalanced mixtures could all be assigned as NoC= 3. Analysis without Mx priors produced an alternate genotype explanation.

PMID:35390645 | DOI:10.1016/j.fsigen.2022.102691

J Hazard Mater. 2022 Mar 29;433:128809. doi: 10.1016/j.jhazmat.2022.128809. Online ahead of print.

ABSTRACT

Rapid detection methods are needed to investigate the environmental quality risk of soil and groundwater in contaminated lands. Currently there is lack of rapid detection methods to sensitively and accurately analyze contaminations of hexavalent chromium in soil due to the challenge of complex sample pretreatment or expensive instrumentation. Here we report a rapid accurate detection platform for quantifying hexavalent chromium in soil and groundwater with ultrasensitivity. The platform consists of a novel sensor of microProbing beads and a portable microscope. Each microProbing bead was a nanoliter reactor to selectively sequestrate Cr (VI) with the enrichment factor up to 150 ×. The microProbing beads presented the signal uniformity of ~97% for the statistical colorimetric imaging analysis. Combined with a miniaturized microscope, the microProbing beads allowed for detecting aqueous Cr (VI) and soluble Cr (VI) in soil within 45 min. The platform achieved high sensitivity with the detection limits of 0.003 ppb for aqueous Cr (VI) and 0.07 ppm for soil Cr (VI). It accurately detected soil and groundwater samples from a chromium contaminated land in Yangtze River Basin of China. The consistency to the laboratory standard methods was achieved with the low cost of ~0.20 US dollar per test. The microProbing imaging platform with the operational simplicity and device portability is highly promising for the field analysis of Cr (VI) in contaminated lands.

PMID:35390613 | DOI:10.1016/j.jhazmat.2022.128809

Mult Scler Relat Disord. 2022 Mar 27;61:103777. doi: 10.1016/j.msard.2022.103777. Online ahead of print.

ABSTRACT

BACKGROUND: Iran, as a middle income country, is one of the places with high and rising prevalence of multiple sclerosis (MS). Regarding the substantial economic burden, reviewing the trend in prescribed disease modifying treatments (DMTs) could be of help. Here we studied the DMT information of nearly 14000 MS cases and its trends change for 30 years to improve health services to patients.

METHODS: The population base of this descriptive-analytical (cross-sectional) study consisted of all MS patients in the nationwide MS registry of Iran (NMSRI), up to August 1, 2021. Registrars from 15 provinces, 24 cities, 13 hospitals,8 MS associations, 16 private offices, and 7 clinics had entered the data.

RESULTS: Overall, 14316 cases were enrolled. The majority (76.1%) were female. The youngest and eldest patients were 5 and 78 years old, respectively. Diagnosis delay was under one year in most cases (median: 0, IQR: 0 – 1). Most (61.4%) had RRMS. Generally, platform injectables (IFN beta, glatiramer acetate) were the most used DMTs until 2010. It seems that introduction of newer agents (antiCD20s and oral DMTs) resulted in a decrease in the use of former drugs since around 2015. Some unusual practices are prominent such as using not approved DMTs for PPMS over the years, or administering high efficacy drugs like natalizumab for CIS. The results indicate the remaining popularity of first line injectable DMTs in female and pediatric patients.

DISCUSSION: Mean age (SD) at onset in our study (29 ± 8.8) is near the statistics in Asia and Oceania (28 ± 0.7). Concerns about COVID-19 had a noticeable impact on administering high efficacy drugs like rituximab and fingolimod. However, in male patients this approach has not been the case. It may be related to more aggressive disease course in this group. The other possible explanation could be planning for pregnancy in female cases. The popularity of platform injectable drugs in pediatric MS may be related to its favorable safety profile over the years. Another point in this group, is the superiority of rituximab over other highly efficient medications.

PMID:35390594 | DOI:10.1016/j.msard.2022.103777

Cancer Epidemiol. 2022 Apr 4;78:102153. doi: 10.1016/j.canep.2022.102153. Online ahead of print.

ABSTRACT

BACKGROUND: We used, for the first time, data registered in the Nationwide Registry for Childhood Hematological Malignancies and Solid Tumors (NARECHEM-ST)-Greece to estimate incidence/time trends of the rare childhood (0-14 years) non-Wilms tumors (non-WT), and compared the results of malignant non-WT to those from the Surveillance, Epidemiology, and End Results Program (SEER)-USA.

METHODS: Fifty-five cases (n = 33 malignant-only) were extracted from NARECHEM-ST (2001-2020) and 332 malignant cases from SEER (1990-2017). To allow between-country comparisons, age-standardized incidence rates (AIR) of malignant-only non-WT were calculated, and temporal trends were evaluated using Poisson and joinpoint regressions.

RESULTS: In NARECHEM-ST, malignant and non-malignant non-WT accounted for 22.6% of all renal tumors. Among malignant tumors, the AIR was 1.0/10⁶ children in Greece, similar to that calculated for SEER, USA (AIR=0.9/10⁶). The proportion of infant malignant and non-malignant non-WT was 27% (20% before 6 months) in NARECHEM-ST. Most common non-WT in Greece were congenital mesoblastic nephromas (CMN) diagnosed mainly in infancy (CIR=7.2/10⁶). The proportion of infant malignant non-WT was 20% in SEER (AIR_infancy=2.5/10⁶), mainly attributed to rhabdoid tumors (CIR=1.6/10⁶). The male-to-female (M:F) ratio of malignant non-WT was 0.9 in NARECHEM-ST vs. 1.2 in SEER, whereas boys outnumbered girls with clear cell sarcoma in NARECHEM-ST (M:F=4.0). Lastly, significantly increasing trends in incidence rates were noted in NARECHEM-ST [+ 6.8%, 95% confidence intervals (CI): 0.5, 13.3] and in SEER (+7.3%, 95%CI: 5.6, 9.0).

CONCLUSIONS: Observed incidence, time trends and sociodemographic variations of non-WT may reflect differential registration practices and healthcare delivery patterns including differences regarding surveillance, coding and treatment practices.

PMID:35390585 | DOI:10.1016/j.canep.2022.102153