Vegetation diversity has been found to influence ecosystem function and provide essential ecosystem services, intimately linked to societal wellbeing. However, the relationship between vegetation diversity and function is very complex and still not fully comprehended at different spatial-temporal scales. Indeed, in recent years, remote sensing has shown great promise to better monitor plant diversity at different scales, most commonly through the Spectral Variability Hypothesis (SVH), which links spectral diversity to plant diversity. However, there is still some debate over the generality of the SVH, especially in semi-arid grasslands, which tend to be less studied even though they dominate the trend and inter-annual variability of global water and carbon fluxes. This study focused on examining the relationship between functional diversity (FD) and optical traits of the herbaceous understory of a Mediterranean tree-grass ecosystem (TGE) using field spectroscopy and high resolution imagery from unmanned aerial vehicles (UAVs). Multiple field campaigns were performed from 2021 to 2023 in the Majadas de Tiétar experimental station located in Western Spain to collect in-situ measurements of plant traits (e.g. specific leaf area (SLA), chlorophyll content (Cab)), diversity metrics (functional dispersion (Fdis), Rao’s entropy (Qrao)), hyperspectral field spectroscopy (ASD Fieldspec® 3 portable spectroradiometer) and high-resolution visible-near-infrared (VNIR) and thermal infrared (TIR) imagery onboard UAVs. By applying partial-least-square regression (PLSR) models, high correlations were observed between field spectroscopy and plant traits (r2 > 0.7) with SWIR bands having the most weight in the predictive power of these empirical models, perhaps related to water being the principal limiting factor for herbaceous plants in these semi-arid conditions. By contrast, in-situ PLSR models showed little/no relation to plant diversity metrics (r2 < 0.1). However, preliminary results from the UAV images showed that the spatial heterogeneity of NDVI and land surface temperature (LST), quantified through Qrao using a 5 x 5 pixel window, were positively related to in-situ diversity metrics such as Fdis. Indeed, Qrao based on LST was found to have a more significant relationship to Fdis (p-value < 0.05) compared to Qrao based on NDVI (p-value > 0.05). While the remote sensing of plant functional diversity has concentrated on shortwave reflectances, the use of TIR imagery has large potential as it is more directly related to ecosystem function with its capabilities to act as a proxy for plant transpiration and inform on water use efficiency (WUE). This work is step forward to better understand the optical-diversity relationship in a semi-arid grassland using data acquired at different scales but also from different sources ranging from in-situ hyperspectral measurements to high-resolution TIR imagery.

Background: Accurate planning of the duration of stays at intensive care units is of utmost importance for resource planning. Currently, the discharge date used for resource management is calculated only at admission time and is called length of stay. However, the evolution of the treatment may be different from one patient to another, so a recalculation of the date of discharge should be performed, called days to discharge. The prediction of days to discharge during the stay at the ICU with statistical and data analysis methods has been poorly studied with low-quality results. This study aims to improve the prediction of the discharge date for any patient in intensive care units using artificial intelligence techniques. Methods: The paper proposes a hybrid method based on group-conditioned models obtained with machine learning techniques. Patients are grouped into three clusters based on an initial length of stay estimation. On each group (grouped by first days of stay), we calculate the group-conditioned length of stay value to know the predicted date of discharge, then, after a given number of days, another group-conditioned prediction model must be used to calculate the days to discharge in order to obtain a more accurate prediction of the number of remaining days. The study is performed with the eICU database, a public dataset of USA patients admitted to intensive care units between 2014 and 2015. Three machine learning methods (i.e., Random Forest, XGBoost, and lightGBM) are used to generate length of stay and days to discharge predictive models for each group. Results: Random Forest is the algorithm that obtains the best days to discharge predictors. The proposed hybrid method achieves a root mean square error (RMSE) and mean average error (MAE) below one day on the eICU dataset for the last six days of stay. Conclusions: Machine learning models improve quality of predictions for the days to discharge and length of stay for intensive care unit patients. The results demonstrate that the hybrid model, based on Random Forest, improves the accuracy for predicting length of stay at the start and days to discharge at the end of the intensive care unit stay. Implementing these prediction models may help in the accurate estimation of bed occupancy at intensive care units, thus improving the planning for these limited and critical health-care resources.

OBJECTIVE:The study has dual objectives. Our first objective (1) is to develop a community-of-practice-based evaluation methodology for knowledge-intensive computational methods. We target a whitebox analysis of the computational methods to gain insight on their functional features and inner workings. In more detail, we aim to answer evaluation questions on (i) support offered by computational methods for functional features within the application domain; and (ii) in-depth characterizations of the underlying computational processes, models, data and knowledge of the computational methods. Our second objective (2) involves applying the evaluation methodology to answer questions (i) and (ii) for knowledge-intensive clinical decision support (CDS) methods, which operationalize clinical knowledge as computer interpretable guidelines (CIG); we focus on multimorbidity CIG-based clinical decision support (MGCDS) methods that target multimorbidity treatment plans.MATERIALS AND METHODS:Our methodology directly involves the research community of practice in (a) identifying functional features within the application domain; (b) defining exemplar case studies covering these features; and (c) solving the case studies using their developed computational methods-research groups detail their solutions and functional feature support in solution reports. Next, the study authors (d) perform a qualitative analysis of the solution reports, identifying and characterizing common themes (or dimensions) among the computational methods. This methodology is well suited to perform whitebox analysis, as it directly involves the respective developers in studying inner workings and feature support of computational methods. Moreover, the established evaluation parameters (e.g., features, case studies, themes) constitute a re-usable benchmark framework, which can be used to evaluate new computational methods as they are developed. We applied our community-of-practice-based evaluation methodology on MGCDS methods.RESULTS:Six research groups submitted comprehensive solution reports for the exemplar case studies. Solutions for two of these case studies were reported by all groups. We identified four evaluation dimensions: detection of adverse interactions, management strategy representation, implementation paradigms, and human-in-the-loop support. Based on our whitebox analysis, we present answers to the evaluation questions (i) and (ii) for MGCDS methods.DISCUSSION:The proposed evaluation methodology includes features of illuminative and comparison-based approaches; focusing on understanding rather than judging/scoring or identifying gaps in current methods. It involves answering evaluation questions with direct involvement of the research community of practice, who participate in setting up evaluation parameters and solving exemplar case studies. Our methodology was successfully applied to evaluate six MGCDS knowledge-intensive computational methods. We established that, while the evaluated methods provide a multifaceted set of solutions with different benefits and drawbacks, no single MGCDS method currently provides a comprehensive solution for MGCDS.CONCLUSION:We posit that our evaluation methodology, applied here to gain new insights into MGCDS, can be used to assess other types of knowledge-intensive computational methods and answer other types of evaluation questions. Our case studies can be accessed at our GitHub repository (https://github.com/william-vw/MGCDS).

In health-care, there is a need to quantify medical errors. Among these errors, we observe wrong dose prescriptions. Drug dose titration (DT) is the process by which dosage is progressively adjusted to the patient till a steady dose is reached. Depending on the clinical disease, drug, and patient condition, dose titration can follow different procedures. Once modeled, these procedures can serve for clinical homogenization, standardization, decision support and retrospective analysis. Here, we propose a language to model dose titration procedures. The language was used to formalize one- and two-drug titration of chronic and acute cases, and to perform retrospective analysis of the drug titration processes on 253 patients diagnosed of diabetes mellitus type 2 and treated with metformin, 321 patients treated of chonic heart failure with furosemide, 155 patients with hyperuricemia treated with allopurinol as initial drug and febuxostat as alternative drug, and 187 hyperuricemia patients with primary drug allopurinol and supplementary drug probenecid, in order to identify different types of drug titration deviations from standard DT methods.

It is well documented that energy balance and other remote sensing-based evapotranspiration (ET) models face greater uncertainty over water-limited tree-grass ecosystems (TGEs), representing nearly 1/6th of the global land surface. Their dual vegetation strata, the grass-dominated understory and tree-dominated overstory, make for distinct structural, physiological and phenological characteristics, which challenge models compared to more homogeneous and energy-limited ecosystems. Along with this, the contribution of grasses and trees to total transpiration (T), along with their different climatic drivers, is still largely unknown nor quantified in TGEs. This study proposes a thermal-based three-source energy balance (3SEB) model, accommodating an additional vegetation source within the well-known two-source energy balance (TSEB) model. The model was implemented at both tower and continental scales using eddy-covariance (EC) TGE sites, with variable tree canopy cover and rainfall (P) regimes and Meteosat Second Generation (MSG) images. 3SEB robustly simulated latent heat (LE) and related energy fluxes in all sites (Tower: LE RMSD similar to 60 W/m(2); MSG: LE RMSD similar to 90 W/m(2)), improving over both TSEB and seasonally changing TSEB (TSEB-2S) models. In addition, 3SEB inherently partitions water fluxes between the tree, grass and soil sources. The modelled T correlated well with EC T estimates (r > .76), derived from a machine learning ET partitioning method. The T/ET was found positively related to both P and leaf area index, especially compared to the decomposed grass understory T/ET. However, trees and grasses had contrasting relations with respect to monthly P. These results demonstrate the importance in decomposing total ET into the different vegetation sources, as they have distinct climatic drivers, and hence, different relations to seasonal water availability. These promising results improved ET and energy flux estimations over complex TGEs, which may contribute to enhance global drought monitoring and understanding, and their responses to climate change feedbacks.

ICU management depends on the level of occupation and the length of stay of the patients. Daily prediction of the days to discharge (DTD) of ICU patients is essential to that management. Previous studies showed a low predictive capability of internists and ML-generated models. Therefore, more elaborated combinations of ML technologies are required. Here, we present four approaches to the analysis of the DTDs of ICU patients from different perspectives: heterogeneity quantification, biomarker identification, phenotype recognition, and prediction. Several ML-based methods are proposed for each approach, which were tested with the data of 3,973 patients of a Spanish ICU. Results confirm the complexity of analyzing DTDs with intelligent data analysis methods.

Patients in intensive care units are heterogeneous and the daily prediction of their days to discharge (DTD) a complex task that practitioners and computers are not always able to solve satisfactorily. In order to make more precise DTD predictors, it is necessary to have tools for the analysis of the heterogeneity of the patients. Unfortunately, the number of publications in this field is almost non-existent. In order to alleviate this lack of tools, we propose four methods and their corresponding measures to quantify the heterogeneity of intensive patients in the process of determining the DTD. These new methods and measures have been tested with patients admitted over four years to a tertiary hospital in Spain. The results deepen the understanding of the intensive patient and can serve as a basis for the construction of better DTD predictors.

Many satellite missions rely on modeling approaches to acquire global or regional evapotranspiration (ET) products. However, a current challenge in ET modeling lies in dealing with sub-pixel heterogeneity, as models often assume homogeneous conditions at the pixel level. This is particularly an issue for heterogeneous landscapes, such as tree-grass ecosystems (TGE). In these areas, while appearing homogeneous at larger spatial scales pertaining to a single land cover type, the separation of the spectral signals of the main landscape features (e.g. trees and grasses) may not be achieved at the conventional satellite sensor resolution (e.g. 10-1000 m). This leads to important heterogeneity within the pixel grid that may not be accounted for in traditional modeling frameworks. This study examined the effect of pixel heterogeneity on ET simulations over a complex TGE in central Spain. High resolution hyperspectral imagery from five airborne campaigns forced the two-source energy balance (TSEB) model at 1.5-1000 m spatial resolutions. Along with this, the sharpened (20 m) and original (1000 m) Sentinels for Evapotranspiration (Sen-ET) products were evaluated over the study site for 2017. Results indicated that TSEB accurately simulated ET (RMSD: -60 W/m2) when the pixel scale was able to robustly discriminate between grass and tree pixels (<5 m). However, model uncertainty drastically increased at spatial resolution greater than 10 m (RMSD: -115 W/m2). Model performance remains relatively constant between 30 and 1000 m spatial resolutions, with within pixel heterogeneity being similar at all these scales. For mixed pixels (>= 30 m), forcing an effective landscape roughness into TSEB (RMSD: -80 W/m2) or applying a seasonally changing TSEB (TSEB-2S; RMSD: -65 W/m2) improved the modeling performance. The Sen-ET products behaved similarly at both scales with RMSD of ET roughly 80 W/m2. The non-linear relationship between input parameters and flux output, along with the poor representation of aerodynamic surface roughness, were the main drivers for the increased uncertainties at coarser scales. These results suggest that care should be taken when using global ET products over TGE and similarly heterogeneous landscapes. The modeling procedure should inherently account for the presence of vastly different vegetation roughness elements within the pixel, to achieve reliable estimates of turbulent fluxes over a TGE.

In health-care, medical errors are quantified. Among them, wrong dose prescriptions occur. Drug dose titration (DT) is the process by which dosage is progressively adjusted to the patient till a steady dose is reached. Depending on the clinical disease, drug, and patient, dose titration can follow different procedures. Once modeled, these procedures can serve for clinical homogenization, standardization, decision support and retrospective analysis. Here, we propose a language to model dose titration procedures. The language was used to formalize single-drug titration of chronic and acute cases, and perform retrospective analysis of the drug titration processes on 1,000 cases treated with Bisoprolol and 2,430 cases treated with Ramipril, in order to identify different types of drug titration deviations from standard DT methods.

Mixed vegetation systems such as wood pastures and shrubby pastures are vital for extensive and sustainable livestock production as well as for the conservation of biodiversity and provision of ecosystem services, and are mostly located in areas that are expected to be more strongly affected by climate change. However, the structural characteristics, phenology, and the optical properties of the vegetation in these mixed -ecosystems such as savanna-like ecosystems in the Iberian Peninsula which combines herbaceous and/or shrubby understory with a low density tree cover, constitute a serious challenge for the remote sensing studies. This work combines physical and empirical methods to improve the estimation of essential vegetation variables: leaf area index (LAI, m2 / m2 ), leaf (Cab,leaf, μg / cm2 ) and canopy(Cab,canopy, g / m2 ) chlorophyll content, and leaf (Cm, leaf, g / cm2 ) and canopy (Cm,canopy, g / m2 ) dry matter content in a dehesa ecosystem. For this purpose, a spectral simulated database for the four main phenological stages of the highly dynamic herbaceous layer (summer senescence, autumn regrowth, greenness peak and beginning of senescence), was built by coupling PROSAIL and FLIGHT radiative transfer models. This database was used to calibrate different predictive models based on vegetation indices (VI) proposed in the literature which combine different spectral bands; as well as Partial Least Squares Regression (PLSR) using all bands in the simulated spectral range (400-2500 nm). PLSR models offered greater predictive power (R2 ≥ 0.93, RRMSE ≤ 10.77 %) both for the leaf and canopy- level variables. The results suggest that directional and geometric effects control the relationships between simulated reflectance factors and the foliar parameters. High seasonal variability is observed in the relationship between biophysical variables and IVs, especially for LAI and Cab, which is confirmed in the PLSR analysis. The models developed need to be validated with spectral data obtained either with proximal or remote sensors.

Mixed vegetation systems such as wood pastures and shrubby pastures are vital for extensive and sustainable livestock production as well as for the conservation of biodiversity and provision of ecosystem services, and are mostly located in areas that are expected to be more strongly affected by climate change. However, the structural characteristics, phenology, and the optical properties of the vegetation in these mixed ecosystems such as savanna-like ecosystems in the Iberian Peninsula which combines herbaceous and/or shrubby understory with a low density tree cover, constitute a serious challenge for the remote sensing studies. This work combines physical and empirical methods to improve the estimation of essential vegetation variables: leaf area index (LAI, m(2)/m(2)), leaf (C-ab,C-leaf,C- mu g/cm(2)) and canopy(C-ab,C-canopy g/m(2)) chlorophyll content, and leaf (C-m,C-leaf,C- g/cm(2)) and canopy (C-m,C-canopy, g/m(2)) dry matter content in a dehesa ecosystem. For this purpose, a spectral simulated database for the four main phenological stages of the highly dynamic herbaceous layer (summer senescence, autumn regrowth, greenness peak and beginning of senescence), was built by coupling PROSAIL and FLIGHT radiative transfer models. This database was used to calibrate different predictive models based on vegetation indices (VI) proposed in the literature which combine different spectral bands; as well as Partial Least Squares Regression (PLSR) using all bands in the simulated spectral range (400-2500 nm). PLSR models offered greater predictive power (R-2 >= 0.93, RRMSE <= 10.77 %) both for the leaf and canopy level variables. The results suggest that directional and geometric effects control the relationships between simulated reflectance factors and the foliar parameters. High seasonal variability is observed in the relationship between biophysical variables and IVs, especially for LAI and C-ab , which is confirmed in the PLSR analysis. The models developed need to be validated with spectral data obtained either with proximal or remote sensors.

The acute respiratory distress syndrome (ARDS) is a frequent type of respiratory failure observed in intensive care units. The Berlin classification identifies three severity levels of ARDS (mild, moderate, and severe), but this classification is under controversy in the medical community because it reflects neither the care requirements nor the expected clinical outcome of the patients. Here, the database MIMIC III (MetaVision) was used to investigate the similarity of patients within each one of the Berlin severity groups. We also ranked the relevance of common ARDS descriptive features and proposed four alternative classifiers to improve Berlin's classification in the prediction of the duration of mechanical ventilation and mortality. One of these classifiers proved to be significantly better than current proposals and, therefore, it can be considered as a robust model to potentially improve health care processes and quality in the management of ARDS patients in Intensive Care Units (ICUs).

In hospital intensive care units (ICU), patients are under continuous evaluation. One of the purposes of this evaluation is to determine the expected number of days to discharge. This value is important to manage ICUs. Some studies show that health care professionals are good at predicting short-term discharge times, but not as good at long-term predictions. Machine learning methods can achieve 1.79-day average prediction error. We performed a study on 3,787 patient-days in the ICU of the Hospital Joan XXIII (Spain) to obtain a data-driven model to predict the discharge time of ICU patients, in a daily basis. Our model, which is based on random forest technology, obtained an error of 1.34 days. We studied the progression of the model as more data are available and predicted that the number of instances required to reduce the error below one day is 4,745. When we trained the model with all the available data, we obtained a mean error of less than half a day with a coefficient of determination (R2) above 97% in their predictions on either ICU survivors and not survivors. Similar results were obtained differentiating by patients’ gender and age, confirming our approach as a good means to achieve optimal performance when more data will be available.

Globe-LFMC is an extensive global database of live fuel moisture content (LFMC) measured from 1,383 sampling sites in 11 countries: Argentina, Australia, China, France, Italy, Senegal, Spain, South Africa, Tunisia, United Kingdom and the United States of America. The database contains 161,717 individual records based on in situ destructive samples used to measure LFMC, representing the amount of water in plant leaves per unit of dry matter. The primary goal of the database is to calibrate and validate remote sensing algorithms used to predict LFMC. However, this database is also relevant for the calibration and validation of dynamic global vegetation models, eco-physiological models of plant water stress as well as understanding the physiological drivers of spatiotemporal variation in LFMC at local, regional and global scales. Globe-LFMC should be useful for studying LFMC trends in response to environmental change and LFMC influence on wildfire occurrence, wildfire behavior, and overall vegetation health.

Agricultural activity is one of the major sources of nitrous oxide (N2O) emission into the natural environment. Yet, due to the soil's spatial heterogeneities it is hard to accurately upscale point N2O emission measurements into the orchard scale. This study aims to introduce a simple, yet robust, way for upscaling point N2O emission measurements into the orchard scale, under drip and micro-sprinkler irrigation systems. Surface point measurements of N2O emissions were performed at five distances from drip and micro-sprinkler emitters in two almond orchards, following irrigation and fertigation events. Principal component analysis (PCA) and linear regression were used to study the correlations between the soil water filled pore-space (WFPS), and subsurface N2O, NO3- and NH4+ concentrations down to depth of 60 cm. The correlation tables indicated that most of the N2O emission resulted from microbial nitrification in the top soil (< 10 cm). However, in many cases the correlations did not provide meaningful explanations to the relations between the subsurface parameters and the surface N2O emission flux. It was suggested that the main limitation of the analysis results from the current soil sampling method, where soil samples are not taken from the profile inside the collar in-order to keep the soil profile undisturbed over long measuring periods. In the microsprinkler irrigation, relative water application depth of the emitter, showed strong positive linear correlation to the soil NH4+ concentration and surface N2O emission flux. These correlations could be used for upscaling the point measurements into the tree and orchard scale. In the drip irrigation, the N2O emission flux was found to follow the water and NH4+ distribution pattern, and could be upscaled to the tree and orchard level using a sinusoidal function using only measurement of the peak emission and the radius of the wetting pattern. These results improve current understanding on the dynamics of N2O production in orchards irrigated with micro-irrigation systems in arid and semi-arid ecosystems and might contribute to modeling of N2O emissions using less rigorous methods.

We present the first continental-scale methodology for estimating Live Fuel Moisture Content (FMC) and flammability in Australia using satellite observations. The methodology includes a physically-based retrieval model to estimate FMC from MODIS (Moderate Resolution Imaging Spectrometer) reflectance data using radiative transfer model inversion. The algorithm was evaluated using 363 observations at 33 locations around Australia with mean accuracy for the studied land cover classes (grassland, shrubland and forest) close to those obtained elsewhere (r 2 =0.57, RMSE = 40%) but without site-specific calibration. Logistic regression models were developed to predict a flammability index, trained on fire events mapped in the MODIS burned area product and four predictor variables calculated from the FMC estimates. The selected predictor variables were actual FMC corresponding to the 8-day and 16-day period before burning; the same but expressed as an anomaly from the long-term mean for that date; and the FMC change between the two successive 8-day periods before burning. Separate logistic regression models were developed for grassland, shrubland and forest, obtaining performance metrics of 0.70, 0.78 and 0.71, respectively, indicating reasonable skill in fire risk prediction.

Background : In medical practice, long term interventions are common and they require timely planning of the involved processes. Unfortunately, evidence-based statements about time are hard to find in Clinical Practice Guidelines (CPGs) and in other sources of medical knowledge. At the same time, health care centers use medical records and information systems to register data about clinical processes and patients, including time information about the encounters, prescriptions, and other clinical actions. Consequently, medical records and health care information systems are promising sources of data from which we can detect temporal medical knowledge. Objective : The objectives were to (1) Analyze and classify the sorts of time constraints in medical processes, (2) Propose a formalism to represent these sorts of clinical time constraints, (3) Use these formalisms to enable the automatic generation of temporal models from clinical data, and (4) Study the adherence of these intervention models to CPG recommendations. Methods : In order to achieve these objectives, we carried out four studies: The identification of the sort of times involved in the long-term diagnostic and therapeutic medical procedures of fifty patients, the supervision of the indications about time contained in six CPGs on chronic diseases, the study of the time structures of two standard data models, as well as ten languages to computerize CPGs. Based on the provided studies, we synthesized two representation formalisms: Micro-and macro-temporality. We developed three algorithms for automatic generation of generalized time constraints in the form of micro-and macro-temporalities from clinical databases, which were double tested. Results : A full classification of time constraints for medical procedures is proposed. Two formalisms called micro-and macro-temporality are introduced and validated to represent these time constraints. Time constraints were generated automatically from the data about 8781 Arterial Hypertension (AH) patients. The generated macro-temporalities restricted visits to be between 1-7 weeks, whereas CPGs recommend 2-4 weeks. Micro-temporal constraints on drug-dosage therapies distinguished between the initial dosage and the target dosage, with visits every 1-6 weeks, and 2-5 months, respectively. Our algorithms obtained semi-complete maps of dosage increments and the maximum dosages for 7 drug types. Data-based time limits for lifestyle change counsels and blood pressure (BP) check-ups were fixed to 6 and 3 months, for patients with low- and high-BP, respectively, when CPGs specify a general 3-6 month range. Conclusions : Experience-based temporal knowledge detected using our algorithms complements the evidence-based knowledge about clinical procedures contained in the CPGs. Our temporal model is simple and highly descriptive when dealing with general or specific time constraints' representations, offering temporal knowledge representation of varying detail. Therefore, it is capable of capturing all the temporal knowledge we can find in medical procedures, when dealing with chronic diseases. With our model and algorithms, an adherence analysis emerges naturally to detect CPG-compliant interventions, but also deviations whose causes and possible rationales can call into question CPG recommendations (e.g., our analysis of AH patients showed that the time between visits recommended by CPGs were too long for a proper drug therapy decision, dosage titration, or general follow-up). (C) 2018 Elsevier B.V. All rights reserved.

This paper presents medical knowledge representation of data provided within Clinical Practice Guidelines for Heart Failure. The formalization is provided in order to support taking decisions on an appropriate treatment strategy for a specific patient. An intuitive and efficient mechanism of medical knowledge formalization, called extended Timed Transition Diagram (eTTD), is used to represent acquired medical knowledge. The presented models can be used to help students in their training as well as to support physicians with their decision-making tasks.