Radiomics analysis for prediction and classification of submucosal tumors based on gastrointestinal endoscopic ultrasonography.

Objectives: To identify and classify submucosal tumors by building and validating a radiomics model with gastrointestinal endoscopic ultrasonography (EUS) images. Methods: A total of 144 patients diagnosed with submucosal tumors through gastrointestinal EUS were collected between January 2019 and October 2020. There are 1952 radiomic features extracted from each patient's EUS images. The statistical test and the customized least absolute shrinkage and selection operator regression were used for feature selection. Subsequently, an extremely randomized trees algorithm was utilized to construct a robust radiomics classification model specifically tailored for gastrointestinal EUS images. The performance of the model was measured by evaluating the area under the receiver operating characteristic curve. Results: The radiomics model comprised 30 selected features that showed good discrimination performance in the validation cohorts. During validation, the area under the receiver operating characteristic curve was calculated as 0.9203 and the mean value after 10-fold cross-validation was 0.9260, indicating excellent stability and calibration. These results confirm the clinical utility of the model. Conclusions: Utilizing the dataset provided curated from gastrointestinal EUS examinations at our collaborating hospital, we have developed a well-performing radiomics model. It can be used for personalized and non-invasive prediction of the type of submucosal tumors, providing physicians with aid for early treatment and management of tumor progression.

Differentiating Pressure Ulcer Risk Levels through Interpretable Classification Models Based on Readily Measurable Indicators.

Pressure ulcers carry a significant risk in clinical practice. This paper proposes a practical and interpretable approach to estimate the risk levels of pressure ulcers using decision tree models. In order to address the common problem of imbalanced learning in nursing classification datasets, various oversampling configurations are analyzed to improve the data quality prior to modeling. The decision trees built are based on three easily identifiable and clinically relevant pressure ulcer risk indicators: mobility, activity, and skin moisture. Additionally, this research introduces a novel tabular visualization method to enhance the usability of the decision trees in clinical practice. Thus, the primary aim of this approach is to provide nursing professionals with valuable insights for assessing the potential risk levels of pressure ulcers, which could support their decision-making and allow, for example, the application of suitable preventive measures tailored to each patient's requirements. The interpretability of the models proposed and their performance, evaluated through stratified cross-validation, make them a helpful tool for nursing care in estimating the pressure ulcer risk level.

Branching Out: A Retrospective Review of Tree Fall-Related Trauma.

Introduction Falls from trees (FFTs), although rare, represent a significant public health concern due to the severe consequences they can impose. Such incidents, while statistically uncommon across the wider population, have the potential to cause drastic, lasting alterations in patients' lives. The severity of these events is often substantial, highlighted by high Injury Severity Scores (ISSs) and prolonged hospital length of stay (LOS), which brings to light the urgent need for preventive strategies and heightened awareness. Our study aims to present a current epidemiological understanding of the patterns, nature, and severity of injuries caused by FFTs. Additionally, it provides an analysis and comparison of data obtained from a de-identified trauma database of patients presenting after FFTs. Methods This review presents data from a trauma registry system detailing trauma admissions from March 31, 2016, to December 27, 2021, at the Desert Regional Medical Center in Palm Springs, California, United States, a designated Level 1 trauma center. Throughout this period of nearly five years and eight months, a total of 3,148 patients were recorded to have visited the emergency department due to falls. Specifically, the study zeroes in on the subset of patients who were admitted after experiencing FFTs. From the comprehensive retrospective examination, it was noted that among the 3,148 fall incidents, there were 50 cases that involved FFTs. Results This retrospective analysis focused on 50 patients treated at the emergency department after FFTs, with a predominantly male demographic profile of 49 (98%) and an average age of 44 years. Hospitalization was required for the vast majority (44%), with approximately one-third necessitating ICU care. Surgical procedures were necessary for 35 (70%) of these cases. Upon discharge, 36 (72% of patients) were able to return home. Vertebral fractures were the most frequent injury, present in 24 (22% of admissions), followed closely by soft tissue injuries at 23 (21%). The mean ISS was 11, although those with extended hospital stays of over 10 days had higher ISS scores of 16, in contrast to an ISS of 10 for those with shorter stays. Conclusions FFTs constitute a lesser-known category of trauma-related injuries in the broader spectrum of fall-related incidents. Although relatively infrequent, these incidents result in significant injury burdens. The objective of this review is to compile and summarize the existing body of literature on FFTs. It involves an in-depth analysis of admission, discharge, and demographic data related to FFTs, highlighting the significant consequences associated with such accidents. Additionally, this review incorporates an analysis of a specialized dataset dedicated to injuries resulting from FFTs, facilitating a comparative assessment against current research in this field.

Identifying the generalizable controls on insect associations of native and non-native trees.

Trees growing outside their native geographic ranges often exhibit exceptional growth and survival due in part to the lack of co-evolved natural enemies that may limit their spread and suppress population growth. While most non-native trees tend to accumulate natural enemies over time, it remains uncertain which host and insect characteristics affect these novel associations and whether novel associations follow patterns of assembly similar to those of native hosts. Here, we used a dataset of insect-host tree associations in Europe to model which native insect species are paired with which native tree species, and then tested the model on its ability to predict which native insects are paired with which non-native trees. We show that native and non-native tree species closely related to known hosts are more likely to be hosts themselves, but that native host geographic range size, insect feeding guild, and sampling effort similarly affect insect associations. Our model had a strong ability to predict which insect species utilize non-native trees as hosts, but evolutionarily isolated tree species posed the greatest challenge to the model. These results demonstrate that insect-host associations can be reliably predicted, regardless of whether insect and host trees have co-evolved, and provide a framework for predicting future pest threats using a select number of easily attainable tree and insect characteristics.

Development and Validation of a Decision Tree Analysis Model for Predicting Home Discharge in a Convalescent Ward: A Single Institution Study.

OBJECTIVES: Accurately predicting the likelihood of inpatients' home discharge in a convalescent ward is crucial for assisting patients and families in decision-making. While logistic regression analysis has been commonly used, its complexity limits practicality in clinical settings. We focused on decision tree analysis, which is visually straightforward. This study aimed to develop and validate the accuracy of a prediction model for home discharge for inpatients in a convalescent ward using a decision tree analysis. METHODS: The cohort consisted of 651 patients admitted to our convalescent ward from 2018 to 2020. We collected data from medical records, including disease classification, sex, age, duration of acute hospitalization, discharge destination (home or nonhome), and Functional Independence Measure (FIM) subitems at admission. We divided the cohort data into training and validation sets and developed a prediction model using decision tree analysis with discharge destination as the target and other variables as predictors. The model's accuracy was validated using the validation data set. RESULTS: The decision tree model identified FIM grooming as the first single discriminator of home discharge, diverging at four points and identifying subsequent branching for the duration of acute hospitalization. The model's accuracy was 86.7%, with a sensitivity of 0.96, specificity of 0.52, positive predictive accuracy of 0.88, and negative predictive accuracy of 0.80. The area under the receiver operating characteristic curve was 0.75. CONCLUSION: The predictive model demonstrated more than moderate predictive accuracy, suggesting its utility in clinical practice. Grooming emerged as a variable with the highest explanatory power for determining home discharge.

Chronic anthropogenic disturbance mediates the bottom-up influence of plant diversity on arthropods in tropical forests.

Understanding how primary productivity and diversity affect secondary productivity is an important debate in ecology with implications for biodiversity conservation. Particularly, how plant diversity influences arthropod diversity contributes to our understanding of trophic cascades and species coexistence. Previous studies show a positive correlation between plant and arthropod diversity. The theory of associational resistance suggests that plant herbivory rate will decrease with increasing plant diversity indicating feedbacks between primary diversity, productivity, and secondary productivity rates. However, our understanding of how these relations are mediated by anthropogenic disturbance is still limited. We surveyed 10 forest sites, half of which are disturbed by fire, logging, and tree pruning, distributed in two climatic zones in Benin, West Africa. We established 100 transects to record plant species and sampled arthropods using pitfall traps, ceramic plates with bait, and sweeping nets. We developed a structural equation model to test the mediating effect of chronic anthropogenic disturbance on plant diversity and how it influences arthropod diversity and abundance. Arthropod diversity increased but arthropod abundance decreased with increasing intensity of disturbance. We found no significant bottom-up influence of the plant diversity on arthropod diversity but a significant plant diversity-arthropod abundance relationship. Some arthropod guilds were significantly affected by plant diversity. Finally, herbivory rates were positively associated with arthropod diversity. Synthesis. Our results highlight how chronic anthropogenic disturbance can mediate the functional links between trophic levels in terms of diversity and productivity. Our study demonstrated a decoupled response of arthropod diversity and abundance to disturbance. The direct positive influence of plant diversity on herbivory rates we found in our study provides counter-support for the theory of associational resistance.

ToTE: A global database on trees of the treeline ecotone.

Globally, treelines form a transition zone between tree-dominated forest downslope and treeless alpine vegetation upslope. Treelines represent the highest boundary of "tree" life form in high-elevation mountains and at high latitudes. Recently, treelines have been shifting upslope in response to climate warming, so it has become important to understand global tree diversity and treeline distributions. However, to the best of our knowledge, no global database on tree flora of treelines exists, which limits our capacity to undertake macroecological analyses. Here, for the first time, we present a global data set on the trees of the treeline ecotone, supported by an online ToTE database. We synthesized the database from 1202 studies published over the last 60 years (1962 to 2022) following the Preferred Reporting Items in Systematic Reviews and Meta-Analysis (PRISMA) protocol. We classified the tree species in the database into three categories: treeline tree (TL) species, near to treeline (NTL) tree species, and tree species with an upper montane range limit (TUMR). The ToTE Version-1 presents a total of 208 tree taxa, including 189 species, five subspecies, and 14 varieties, belonging to 54 genera and 26 families distributed across 34 mountain regions worldwide that either grow exactly at the treeline or have a range limit below the treeline. Of the total taxa, 155, 14, and 39 belong to TL, NTL, and TUMR, respectively. Genera such as Abies, Picea, Pinus, Larix, and Juniperus are more represented in the treeline tree category. On the other hand, Acer, Prunus, Populus, and Quercus have more representatives in the near to treeline category, whereas Erica, Nothofagus, and Polylepis contribute more tree species with an upper montane range limit. Furthermore, families such as Rosaceae and Pinaceae include trees that occur both at the treeline and with an upper montane range limit, whereas Sapindaceae includes trees that occur exclusively near to treeline. Our database also includes information on the global distribution patterns of treeline tree species richness across mountains and biomes. The mountains with the highest number of tree species are the Andes (39) followed by the Himalaya (37). Close to 67% of tree species show restricted distributions in different mountains, with the highest endemism in the Andes and the Himalaya. In terms of tree species distribution, Pinus sylvestris was widespread, with a distribution across nine mountain regions, followed by Picea glauca and Fagus sylvatica, both distributed across five mountain regions. In terms of species' distribution across biomes, the temperate biome harbors the highest treeline tree species richness (152 species), which may reflect the fact that the majority of studies are available from the temperate regions of the world. The remaining 56 species are distributed within five other biomes, with the least in dry tropical and subarctic (four species each). Furthermore, currently 40 treeline tree species fall under different International Union for Conservation of Nature threat categories. We anticipate that our database will help advance research on macroecological, biogeographic, evolutionary, climate-change, and conservation aspects of the treeline on a global scale. The data are released under a Creative Commons Attribution 4.0 international license. Please cite this data paper when the data are reused.

Using decision tree models and comprehensive statewide data to predict opioid overdoses following prison release.

PURPOSE: Identifying predictors of opioid overdose following release from prison is critical for opioid overdose prevention. METHODS: We leveraged an individually linked, state-wide database from 2015-2020 to predict the risk of opioid overdose within 90 days of release from Massachusetts state prisons. We developed two decision tree modeling schemes: a model fit on all individuals with a single weight for those that experienced an opioid overdose and models stratified by race/ethnicity. We compared the performance of each model using several performance measures and identified factors that were most predictive of opioid overdose within racial/ethnic groups and across models. RESULTS: We found that out of 44,246 prison releases in Massachusetts between 2015-2020, 2237 (5.1%) resulted in opioid overdose in the 90 days following release. The performance of the two predictive models varied. The single weight model had high sensitivity (79%) and low specificity (56%) for predicting opioid overdose and was more sensitive for White non-Hispanic individuals (sensitivity = 84%) than for racial/ethnic minority individuals. CONCLUSIONS: Stratified models had better balanced performance metrics for both White non-Hispanic and racial/ethnic minority groups and identified different predictors of overdose between racial/ethnic groups. Across racial/ethnic groups and models, involuntary commitment (involuntary treatment for alcohol/substance use disorder) was an important predictor of opioid overdose.

Variation in the timing and duration of autumn leaf phenology among temperate deciduous trees, native shrubs and non-native shrubs.

The timing and duration of autumn leaf phenology marks important transitions in temperate deciduous forests, such as, start of senescence, declining productivity and changing nutrient cycling. Phenological research on temperate deciduous forests typically focuses on upper canopy trees, overlooking the contribution of other plant functional groups like shrubs. Yet shrubs tend to remain green longer than trees, while non-native shrubs, in particular, tend to exhibit an extended growing season that confers a competitive advantage over native shrubs. We monitored leaf senescence and leaf fall (2017-2020) of trees and shrubs (native and non-native) in an urban woodland fragment in Wisconsin, USA. Our findings revealed that, the start of leaf senescence did not differ significantly between vegetation groups, but leaf fall started (DOY 273) two weeks later in shrubs. Non-native shrubs exhibited a considerably delayed start (DOY 262) and end of leaf senescence (DOY 300), with leaf-fall ending (DOY 315) nearly four weeks later than native shrubs and trees. Overall, the duration of the autumn phenological season was longer for non-native shrubs than either native shrubs or trees. Comparison of the timing of spring phenophases with the start and end of leaf senescence revealed that when spring phenology in trees starts later in the season senescence also starts later and ends earlier. The opposite pattern was observed in native shrubs. In conclusion, understanding the contributions of plant functional groups to overall forest phenology requires future investigation to ensure accurate predictions of future ecosystem productivity and help address discrepancies with remote sensing phenometrics.

Optimizing lung cancer classification through hyperparameter tuning.

Artificial intelligence is steadily permeating various sectors, including healthcare. This research specifically addresses lung cancer, the world's deadliest disease with the highest mortality rate. Two primary factors contribute to its onset: genetic predisposition and environmental factors, such as smoking and exposure to pollutants. Recognizing the need for more effective diagnosis techniques, our study embarked on devising a machine learning strategy tailored to boost precision in lung cancer detection. Our aim was to devise a diagnostic method that is both less invasive and cost-effective. To this end, we proposed four methods, benchmarking them against prevalent techniques using a universally recognized dataset from Kaggle. Among our methods, one emerged as particularly promising, outperforming the competition in accuracy, precision and sensitivity. This method utilized hyperparameter tuning, focusing on the Gamma and C parameters, which were set at a value of 10. These parameters influence kernel width and regularization strength, respectively. As a result, we achieved an accuracy of 99.16%, a precision of 98% and a sensitivity rate of 100%. In conclusion, our enhanced prediction mechanism has proven to surpass traditional and contemporary strategies in lung cancer detection.

Variational Bayesian phylogenies through matrix representation of tree space.

In this article, we study the distance matrix as a representation of a phylogeny by way of hierarchical clustering. By defining a multivariate normal distribution on (a subset of) the entries in a matrix, this allows us to represent a distribution over rooted time trees. Here, we demonstrate tree distributions can be represented accurately this way for a number of published tree distributions. Though such a representation does not map to unique trees, restriction to a subspace, in particular one we call a "cube", makes the representation bijective at the cost of not being able to represent all possible trees. We introduce an algorithm "cubeVB" specifically for cubes and show through well calibrated simulation study that it is possible to recover parameters of interest like tree height and length. Although a cube cannot represent all of tree space, it is a great improvement over a single summary tree, and it opens up exciting new opportunities for scaling up Bayesian phylogenetic inference. We also demonstrate how to use a matrix representation of a tree distribution to get better summary trees than commonly used maximum clade credibility trees. An open source implementation of the cubeVB algorithm is available from https://github.com/rbouckaert/cubevb as the cubevb package for BEAST 2.

In-Electrospray source Hydrogen/Deuterium exchange coupled to multistage fragmentation for the investigation of the protonation and fragmentation pathways of gas phase ions.

Identification of molecules in complex natural matrices relies on matching the fragmentation spectra of ions under investigation and the spectra acquired for the corresponding analytical standards. Currently, there are many databases of experimentally measured tandem mass spectrometry spectra (such as NIST, MzCloud, and Metlin), and considerable progress has been made in the development of software for predicting tandem mass spectrometry fragments in silico using combinatorial, machine learning, and quantum chemistry approaches (such as MetFrag, CFM-ID, and QCxMS). However, the electrospray ionization molecules can be ionized at different sites (protonated or deprotonated), and the fragmentation spectra of such ions are different. Here, we are using the combination of the in-ESI source hydrogen/deuterium exchange reaction and MSn fragmentation for the investigation of the fragmentation pathways for different protomers of organic molecules. It is shown that the distribution of the deuterium in the fragment ions reflects the presence of different protomers. For several molecules, the distribution of deuterium was traced up to the MS5 level of fragmentation revealing many unusual and unexpected effects. For example, we investigated the loss of HF from the ciprofloxacin and norfloxacin ions and observed that for ions protonated at -COOH group, the eliminating hydrogen always comes from -NH group. When ions are protonated at another site, the elimination of hydrogen with a probability of 30% occurs from the -NH group, and with a probability of 70%, it originates from other sites on the molecule. Such effects were not described previously. Quantum chemical simulation was used for the verification of the protonated structures and simulation of the corresponding fragmentation spectra.

Intercropping fruit trees in tea plantation improves soil properties and the formation of tea quality components.

Intercropping has been recommended as a beneficial cropping practice for improving soil characteristic and tea quality. However, there is limited research on the effects of intercropping fruit trees on soil chemical properties, soil aggregate structure, and tea quality components. In this study, intercropping fruit trees, specifically loquats and citrus, had a significant impact on the total available nutrients, AMN, and AP in soil. During spring and autumn seasons, the soil large-macroaggregates (>2 mm) proportion increased by 5.93% and 19.03%, as well as 29.23% and 19.14%, respectively, when intercropping loquats and citrus. Similarly, intercropping waxberry resulted in a highest small-macroaggregates (0.25 mm-2 mm) proportion at 54.89% and 77.32%. Soil aggregate stability parameters of the R0.25, MWD, and GMD were generally considered better soil aggregate stability indicators, and significantly improved in intercropping systems. Intercropping waxberry with higher values for those aggregate stability parameters and lower D values, showed a better soil aggregate distribution, while intercropping loquats and citrus at higher levels of AMN and AP in different soil aggregate sizes. As the soil aggregate sizes increased, the AMN and AP contents gradually decreased. Furthermore, the enhanced levels of amino acids were observed under loquat, waxberry, and citrus intercropping in spring, which increased by 27.98%, 27.35%, and 26.21%, respectively. The contents of tea polyphenol and caffeine were lower under loquat and citrus intercropping in spring. These findings indicated that intercropping fruit trees, specifically loquat and citrus, have immense potential in promoting the green and sustainable development of tea plantations.

Where the not-so-wild things are in cities? The influence of social-ecological factors in urban trees at multiple scales.

Green infrastructure plays an essential role in cities due to the ecosystem services it provides. However, these elements are shaped by social and ecological factors that influence their distribution and diversity, affecting ecological functions and human well-being. Here, we analyzed neighborhood tree distribution - trees in pocket parks, squares and along streets - in Lisbon (Portugal) and modelled tree abundance and taxonomic and functional diversity, at the parish and local scales, considering a comprehensive list of social and ecological factors. For the functional analyses, we included functional traits linked to dispersal, resilience to important perturbations in coastal Mediterranean cities, and ecosystem services delivery. Our results show not only that trees are unevenly distributed across the city, but that there is a strong influence of social factors on all biological indices considered. At the parish and local scales, abundance and diversity responded to different factors, with abundance being linked to both social and ecological variables. Although the influence of social factors on urban trees can be expected, by modelling their influence we can quantify how much humans modify urban landscapes at a structural and functional level. These associations can underlie potential biodiversity filters and should be analyzed over time to inform decisions that support long-term ecological resilience, maximize trait functional expression, and increase equity in ecosystem services delivery.

Impact of greening trends on biogenic volatile organic compound emissions in China from 1985 to 2022: Contributions of afforestation projects.

The rapid expansion of green areas in China has enhanced carbon sinks, but it also presents challenges regarding increased biogenic volatile organic compound (BVOC) emissions. This study examines the impact of greening trends on BVOC emissions in China from 1985 to 2001 and from 2001 to 2022, focusing on evaluating long-term trends in BVOC emissions within eight afforestation project areas during these two periods. Emission factors for 62 dominant tree species and provincial Plant Functional Types were updated. The BVOC emission inventories were developed for China at a spatial resolution of 27 km × 27 km using the Model of Emissions of Gases and Aerosols from Nature. The national BVOC emissions in 2018 were estimated at 54.24 Tg, with isoprene, monoterpenes, sesquiterpenes, and other BVOC contributing 26.94 Tg, 2.29 Tg, 0.44 Tg, and 24.57 Tg, respectively. Over the past 37 years, BVOC emissions experienced a slow growth rate of 1.7 % (0.79 Tg) during 1985-2001, followed by a significant increase of 12 % (6 Tg) from 2001 to 2022. BVOC emissions in the eight afforestation project areas increased by 2 % and 20 % during the two periods. From 2001 to 2022, at the regional scale, the Shelterbelt program for the middle reaches of the Yellow River area exhibited the largest rate of increase (43 %) in BVOC emissions. The Shelterbelt program for the upper and middle reaches of the Yangtze River made the most largest contribution (45 %) to the national increase in BVOC emissions. Afforestation projects have shifted towards planting more broadleaf trees than needleleaf trees from 2001 to 2022, and there also showed a change from herbaceous plants to broadleaf trees. These trends have led to higher average emission factors for vegetation, resulting in increased BVOC emissions. It underscores the importance of considering BVOC emissions when evaluating afforestation initiatives, emphasizing the need to balancing ecological benefits with potential atmospheric consequences.

Ocean to Tree: Leveraging Single-Molecule RNA-Seq to Repair Genome Gene Models and Improve Phylogenomic Analysis of Gene and Species Evolution.

Understanding gene evolution across genomes and organisms, including ctenophores, can provide unexpected biological insights. It enables powerful integrative approaches that leverage sequence diversity to advance biomedicine. Sequencing and bioinformatic tools can be inexpensive and user-friendly, but numerous options and coding can intimidate new users. Distinct challenges exist in working with data from diverse species but may go unrecognized by researchers accustomed to gold-standard genomes. Here, we provide a high-level workflow and detailed pipeline to enable animal collection, single-molecule sequencing, and phylogenomic analysis of gene and species evolution. As a demonstration, we focus on (1) PacBio RNA-seq of the genome-sequenced ctenophore Mnemiopsis leidyi, (2) diversity and evolution of the mechanosensitive ion channel Piezo in genetic models and basal-branching animals, and (3) associated challenges and solutions to working with diverse species and genomes, including gene model updating and repair using single-molecule RNA-seq. We provide a Python Jupyter Notebook version of our pipeline (GitHub Repository: Ctenophore-Ocean-To-Tree-2023 https://github.com/000generic/Ctenophore-Ocean-To-Tree-2023 ) that can be run for free in the Google Colab cloud to replicate our findings or modified for specific or greater use. Our protocol enables users to design new sequencing projects in ctenophores, marine invertebrates, or other novel organisms. It provides a simple, comprehensive platform that can ease new user entry into running their evolutionary sequence analyses.

Low Diversity and High Genetic Structure for Platonia insignis Mart., an Endangered Fruit Tree Species.

Platonia insignis is a fruit tree native to Brazil of increasing economic importance, with its pulp trading among the highest market values. This study aimed to evaluate the structure and genomic diversity of P. insignis (bacurizeiro) accessions from six locations in the Brazilian States of Roraima, Amazonas, Pará (Amazon biome), and Maranhão (Cerrado biome). A total of 2031 SNP markers were obtained using genotyping-by-sequencing (GBS), from which 625 outlier SNPs were identified. High genetic structure was observed, with most of the genetic variability (59%) concentrated among locations, mainly between biomes (Amazon and Cerrado). A positive and significant correlation (r = 0.85; p < 0.005) was detected between genetic and geographic distances, indicating isolation by distance. The highest genetic diversity was observed for the location in the Cerrado biome (HE = 0.1746; HO = 0.2078). The locations in the Amazon biome showed low genetic diversity indexes with significant levels of inbreeding. The advance of urban areas, events of burning, and expansion of agricultural activities are most probably the main factors for the genetic diversity reduction of P. insignis. Approaches to functional analysis showed that most of the outlier loci found may be related to genes involved in cellular and metabolic processes.

Machine Learning and Deep Learning Models for Nocturnal High- and Low-Glucose Prediction in Adults with Type 1 Diabetes.

Glucose management at night is a major challenge for people with type 1 diabetes (T1D), especially for those managed with multiple daily injections (MDIs). In this study, we developed machine learning (ML) and deep learning (DL) models to predict nocturnal glucose within the target range (3.9-10 mmol/L), above the target range, and below the target range in subjects with T1D managed with MDIs. The models were trained and tested on continuous glucose monitoring data obtained from 380 subjects with T1D. Two DL algorithms-multi-layer perceptron (MLP) and a convolutional neural network (CNN)-as well as two classic ML algorithms, random forest (RF) and gradient boosting trees (GBTs), were applied. The resulting models based on the DL and ML algorithms demonstrated high and similar accuracy in predicting target glucose (F1 metric: 96-98%) and above-target glucose (F1: 93-97%) within a 30 min prediction horizon. Model performance was poorer when predicting low glucose (F1: 80-86%). MLP provided the highest accuracy in low-glucose prediction. The results indicate that both DL (MLP, CNN) and ML (RF, GBTs) algorithms operating CGM data can be used for the simultaneous prediction of nocturnal glucose values within the target, above-target, and below-target ranges in people with T1D managed with MDIs.

Evaluation of High-Frequency Measurement Errors from Turned Surface Topography Data Using Machine Learning Methods.

Manufacturing processes in industry applications are often controlled by the evaluation of surface topography. Topography, in its overall performance, includes form, waviness, and roughness. Methods of measurement of surface roughness can be roughly divided into tactile and contactless techniques. The latter ones are much faster but sensitive to external disturbances from the environment. One type of external source error, while the measurement of surface topography occurs, is a high-frequency noise. This noise originates from the vibration of the measuring system. In this study, the methods for reducing high-frequency errors from the results of contactless roughness measurements of turned surfaces were supported by machine learning methods. This research delves into optimizing filtration methods for surface topography measurements through the application of machine learning models, focusing on enhancing the accuracy of surface roughness assessments. By examining turned surfaces under specific machining conditions and employing a variety of digital filters, the study identifies the Gaussian regression filter and spline filter as the most effective methods at a 22.5 µm cut-off. Utilizing neural networks, support vector machines, and decision trees, the research demonstrates the superior performance of SVMs, achieving remarkable accuracy and sensitivity in predicting optimal filtration methods.

Urban greening with shrubs can supercharge invertebrate abundance and diversity.

In urban areas, diverse and complex habitats for biodiversity are often lacking. This lack of diversity not only compromises essential ecological processes, such as pollination and nutrient cycling, but also diminishes the resilience of urban ecosystems to pests and diseases. To enhance urban biodiversity, a possible solution is to integrate shrubs alongside trees, thereby increasing the overall amount of vegetation, structural complexity and the associated resource diversity. Here, using a common garden experiment involving a variety of trees and shrubs planted alone and in combination, we evaluate how canopy-associated invertebrate assemblages are influenced by vegetation type. In particular, we test whether the presence of shrubs, alone or with trees, results in increased abundance and taxonomic richness of invertebrates, compared to trees on their own. We found that the overall abundance of invertebrates, and that of specific functional groups (e.g., herbivores, pollinators, detritivores), was higher on shrubs, compared to trees, and when trees and shrubs were planted in combination (relative to trees on their own). Our results suggest that planting shrub and tree species with wide and dense crowns can increase the associated abundance and taxonomic and functional group richness of invertebrate communities. Overall, our findings indicate that urban planning would benefit from incorporating shrubs alongside urban trees to maximise invertebrate abundance, diversity and function in urban landscapes.