Gold Standard Data Research Articles

Evaluating survey techniques in wastewater-based epidemiology for accurate COVID-19 incidence estimation

Wastewater-based epidemiology (WBE) requires high-quality survey methods to determine the incidence of infections in wastewater catchment areas. In this study, the wastewater survey methods necessary for comprehending the incidence of infection by WBE are clarified. This clarification is based on the correlation with the number of confirmed coronavirus disease 2019 (COVID-19) cases, considering factors such as handling non-detect data, calculation method for representative values, analytical sensitivity, analytical reproducibility, sampling frequency, and survey duration. Data collected from 15 samples per week for two and a half years using a highly accurate analysis method were regarded as gold standard data, and the correlation between severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) RNA concentrations in wastewater and confirmed COVID-19 cases was analyzed by Monte Carlo simulation under the hypothetical situation where the quality of the wastewater survey method was reduced. Regarding data handling, it was appropriate to replace non-detect data with estimates based on distribution, and to use geometric means to calculate representative values. For the analysis of SARS-CoV-2 RNA in samples, using a highly sensitive and reproducible method (non-detect rates of <40 %; ≤0.4 standard deviation) and surveying at least three samples, preferably five samples, per week were considered desirable. Furthermore, conducting the survey over a period of time that included at least 50 weeks was necessary. A WBE that meets these survey criteria is sufficient for the determination of the COVID-19 infection incidence in the catchment. Furthermore, WBE can offer additional insights into infection rates in the catchment, such as the estimated 48 % decrease in confirmed COVID-19 cases visiting a clinic following a COVID-19 legal reclassification in Japan.

Variational Bayes latent class analysis for EHR-based phenotyping with large real-world data

IntroductionBayesian approaches to patient phenotyping in clinical observational studies have been limited by the computational challenges associated with applying the Markov Chain Monte Carlo (MCMC) approach to real-world data. Approximate Bayesian inference via optimization of the variational evidence lower bound, variational Bayes (VB), has been successfully demonstrated for other applications.MethodsWe investigate the performance and characteristics of currently available VB and MCMC software to explore the practicability of available approaches and provide guidance for clinical practitioners. Two case studies are used to fully explore the methods covering a variety of real-world data. First, we use the publicly available Pima Indian diabetes data to comprehensively compare VB implementations of logistic regression. Second, a large real-world data set, Optum™ EHR with approximately one million diabetes patients extended the analysis to large, highly unbalanced data containing discrete and continuous variables. A Bayesian patient phenotyping composite model incorporating latent class analysis (LCA) and regression was implemented with the second case study.ResultsWe find that several data characteristics common in clinical data, such as sparsity, significantly affect the posterior accuracy of automatic VB methods compared with conditionally conjugate mean-field methods. We find that for both models, automatic VB approaches require more effort and technical knowledge to set up for accurate posterior estimation and are very sensitive to stopping time compared with closed-form VB methods.DiscussionOur results indicate that the patient phenotyping composite Bayes model is more easily usable for real-world studies if Monte Carlo is replaced with VB. It can potentially become a uniquely useful tool for decision support, especially for rare diseases where gold-standard biomarker data are sparse but prior knowledge can be used to assist model diagnosis and may suggest when biomarker tests are warranted.

Smartphone automated motor and speech analysis for early detection of Alzheimer's disease and Parkinson's disease: Validation of TapTalk across 20 different devices.

Smartphones are proving useful in assessing movement and speech function in Alzheimer's disease and other neurodegenerative conditions. Valid outcomes across different smartphones are needed before population-level tests are deployed. This study introduces the TapTalk protocol, a novel app designed to capture hand and speech function and validate it in smartphones against gold-standard measures. Twenty different smartphones collected video data from motor tests and audio data from speech tests. Features were extracted using Google Mediapipe (movement) and Python audio analysis packages (speech). Electromagnetic sensors (60Hz) and a microphone acquired simultaneous movement and voice data, respectively. TapTalk video and audio outcomes were comparable to gold-standard data: 90.3% of video, and 98.3% of audio, data recorded tapping/speech frequencies within±1Hz of the gold-standard measures. Validation of TapTalk across a range of devices is an important step in the development of smartphone-based telemedicine and was achieved in this study. TapTalk evaluates hand motor and speechfunctions across a wide range of smartphones.Data showed 90.3% motor and 98.3% speech accuracy within +/-1Hz of gold standards.Validation advances smartphone-based telemedicine for neurodegenerative diseases.

Estimation and Removal of Residual Motion Artifact in Retrospectively Motion-Corrected fMRI Data: A Comparison of Intervolume and Intravolume Motion Using Gold Standard Simulated Motion Data

Residual head motion artifact in motion-corrected resting-state (rs-) functional MRI (fMRI) and fMRI datasets reduces the temporal signal-to-noise ratio and leaves non-neuronal signal components in the data, which can induce false findings in these studies. While various residual motion nuisance regressors have been proposed to regress out residual motion artifact after motion correction, these validations have typically been conducted empirically in in vivo data, since realistic head motion–corrupted MR data are not available. Here, we generated motion-corrupted MR data by altering imaging plane coordinates before each volume and slice acquisition from an ex vivo brain phantom using the simulated prospective acquisition correction (SIMPACE) sequence. Testing SIMPACE motion-corrupted data with various intervolume motion patterns, we first investigated the mechanism of the residual motion signal after motion correction and also proposed a voxel-wise motion nuisance regressor, called the partial volume (PV) regressor. We also modified the slice-oriented motion-correction method (SLOMOCO) pipeline with 6 volume-wise rigid intervolume motion parameters (Vol-mopa), 6 slice-wise rigid intravolume motion parameters (Sli-mopa), and the proposed PV motion nuisance regressor. We then compared the residual signal after application of the modified SLOMOCO (mSLOMOCO) pipeline with two other methods: intervolume motion-correction method (VOLMOCO), and the original SLOMOCO (oSLOMOCO). We found that mSLOMOCO with 12 Vol-/Sli-mopa and PV regressors outperformed both VOLMOCO with 6 Vol-mopa and PV regressors and oSLOMOCO with 14 voxel-wise regressors. In tests of the 10 different motion patterns of SIMPACE datasets with 1× and 2× amplified intravolume motion, mSLOMOCO with 12 Vol-/Sli-mopa and PV regressors pipeline produced the average standard deviation (SD) of the residual time series signals in the gray matter (GM) smaller by 29% (1× amplified intravolume motion) and 45% (2× amplified intravolume motion) than VOLMOCO with 6 Vol-mopa and PV regressors pipeline. Also, mSLOMOCO with 12 Vol-/Sli-mopa and PV regressors pipeline outperformed oSLOMOCO with 14 voxelwise regressors pipeline, generating the average SD in GM smaller by 28% (1× amplified intravolume motion) and 31% (2× amplified intravolume motion) than oSLOMOCO with 14 voxel-wise regressors pipeline. The novel PV regressor also effectively reduced residual motion artifact as a motion nuisance regressor after both VOLMOCO and mSLOMOCO.

Early detection of N, P, K deficiency in Choy Sum using hyperspectral imaging-based spatial spectral feature mining

Leafy vegetables are widely consumed around the world for their rich nutritional qualities. To ensure a reliable and cost-effective supply of leafy vegetables in the future, advancements in their production are essential. Deficiencies of nitrogen (N), phosphorus (P), and potassium (K) impair growth of leafy vegetables and the ensuing visual symptoms make the plants unmarketable. We studied the use of non-contact large area hyperspectral imaging (HSI) for early detection of N, P and K deficiencies in the leafy vegetable, Choy Sum, before the appearance of visual symptoms. The wide spectral data of 500–900 nm extracted from the plants were subjected to advanced feature mining, facilitating the creation of novel spectral indices tailored to each vital nutrient by leveraging the Pearson’s correlations of 0.85 for N, 0.64 for P, and 0.68 for K with gold standard elemental concentration data. Early detection of deficiencies and timely replenishment of macronutrient(s) can prevent the development of obvious symptoms and thus maintain the visual quality of Choy Sum. These newly created spectral indices hold the potential to provide non-destructive estimation of nutrient content in plants, offering a promising avenue for future advancements in precision agriculture and resource-efficient crop management.

Transferability across countries of equations developed using milk mid-infrared spectroscopy to estimate daily body condition score change in dairy cows

Routine milk samples are commonly subjected to spectroscopic analysis within the mid-infrared (MIR) region of the electromagnetic spectrum to estimate macro-constituents of milk like fat, protein, lactose, and urea content. These spectra, however, can also be used to predict other traits, such as daily body condition score (BCS) change. The objective of the present study was to assess the transferability across countries of equations to predict daily body condition score change (ΔBCS) developed using milk MIR data collected in Ireland and in Canada. Body condition was scored on a scale from 1 (emaciated) to 5 (obese) in both countries. A total of 347,254 BCS records from 80,400 Canadian cows were available along with 73,193 BCS records from 6,572 Irish cows. Partial least squares regression (PLSR) and neural networks (NN) were separately used to predict daily ΔBCS. Two scenarios were studied 1) using Canadian and Irish data combined as the calibration data set to predict daily ΔBCS in Canada and in Ireland separately, and 2) Canadian and Irish data used separately to predict daily ΔBCS in each country separately. These prediction methods were applied to data with and without pretreatment (i.e., first derivative of the spectrum) as well as with and without standardizing daily ΔBCS across countries. For all the scenarios investigated, the correlation between actual and predicted daily ΔBCS when calibrated and validated (using cross-validation) in the same country ranged from 0.92 to 0.94, and from 0.85 to 0.87 for the Canadian and Irish data sets, respectively. When the data from Canada and Ireland were combined in the calibration process to predict daily ΔBCS, the correlations between actual and predicted ΔBCS were ≥ 0.90 and ≥ 0.80 for Canadian and Irish daily ΔBCS, respectively indicating no improvement in predictive ability. Predictive performance when calibrated using just Canadian data and validated using just Irish data was poor, and vice versa. Nonetheless, when developing equations for a country for which a limited database (i.e., 100 records) of gold standard and MIR data were available, predictive performance improved when the limited database was supplemented with the large data set from the other country. In general, for some of the investigated scenarios, standardizing the daily ΔBCS data within country before undertaking the calibration improved prediction accuracy. In conclusion, the benefit of merging data from different countries, at least based on the trait (i.e., daily ΔBCS) and countries (i.e., Ireland and Canada) considered in the present study were limited and, in cases, counter-productive.

Inverse kinematics in cervical spine models: Effects of scaling and model degrees of freedom for extension and flexion movements

Intervertebral kinematics can affect model-predicted loads and strains in the spine; therefore knowledge of expected vertebral kinematics error is important for understanding the limitations of model predictions. This study addressed how different kinematic models of the neck affect the prediction of intervertebral kinematics from markers on the head and trunk. Eight subjects executed head and neck extension-flexion motion with simultaneous motion capture and biplanar dynamic stereo-radiography (DSX) of vertebrae C1-C7. A generic head and neck model in OpenSim was scaled by marker data, and three versions of the model were used with an inverse kinematics solver. The models differed according to the number of independent degrees of freedom (DOF) between the head and trunk: 3 rotational DOF with constraints defining intervertebral kinematics as a function of overall head-trunk motion; 24DOF with 3 independent rotational DOF at each level, skull-T1; 48DOF with 3 rotational and 3 translational DOF at each level. Marker tracking error was lower for scaled models compared to generic models and decreased as model DOF increased. The largest mean absolute error (MAE) was found in extension-flexion angle and anterior-posterior translation at C1-C2, and superior-inferior translation at C2-C3. Model scaling and complexity did not have a statistically significant effect on most error metrics when corrected for multiple comparisons, but ranges of motion were significantly different from DSX in some cases. This study evaluated model kinematics in comparison to gold standard radiographic data and provides important information about intervertebral kinematics error that are foundational to model validity.

Exploring Offline Large Language Models for Clinical Information Extraction: A Study of Renal Histopathological Reports of Lupus Nephritis Patients.

Open source, lightweight and offline generative large language models (LLMs) hold promise for clinical information extraction due to their suitability to operate in secured environments using commodity hardware without token cost. By creating a simple lupus nephritis (LN) renal histopathology annotation schema and generating gold standard data, this study investigates prompt-based strategies using three state-of-the-art lightweight LLMs, namely BioMistral-DARE-7B (BioMistral), Llama-2-13B (Llama 2), and Mistral-7B-instruct-v0.2 (Mistral). We examine the performance of these LLMs within a zero-shot learning environment for renal histopathology report information extraction. Incorporating four prompting strategies, including combinations of batch prompt (BP), single task prompt (SP), chain of thought (CoT) and standard simple prompt (SSP), our findings indicate that both Mistral and BioMistral consistently demonstrated higher performance compared to Llama 2. Mistral recorded the highest performance, achieving an F1-score of 0.996 [95% CI: 0.993, 0.999] for extracting the numbers of various subtypes of glomeruli across all BP settings and 0.898 [95% CI: 0.871, 0.921] in extracting relational values of immune markers under the BP+SSP setting. This study underscores the capability of offline LLMs to provide accurate and secure clinical information extraction, which can serve as a promising alternative to their heavy-weight online counterparts.

Rapid Whole Genome Characterization of High-Risk Pathogens Using Long-Read Sequencing to Identify Potential Healthcare Transmission.

Routine use of whole genome sequencing (WGS) has been shown to help identify transmission of pathogens causing healthcare-associated infections (HAIs). However, the current gold standard of short-read, Illumina-based WGS is labor and time-intensive. In light of recent improvements in long-read Oxford Nanopore Technologies (ONT) sequencing, we sought to establish a low resource utilization approach capable of providing accurate WGS-based comparisons of HAI pathogens within a time frame allowing for infection prevention and control (IPC) interventions. WGS was prospectively performed on antimicrobial-resistant pathogens at increased risk of potential healthcare transmission using the ONT MinION sequencer with R10.4.1 flow cells and Dorado basecalling algorithm. Potential transmission was assessed via Ridom SeqSphere+ for core genome multilocus sequence typing and MINTyper for reference-based core genome single nucleotide polymorphisms using previously published cut-off values. The accuracy of our ONT pipeline was determined relative to Illumina-based WGS data generated from the same genomic DNA sample. Over a six-month period, 242 bacterial isolates from 216 patients were sequenced by a single operator. Compared to the Illumina gold-standard data, our ONT pipeline achieved a Q score of 60 for assembled genomes, even with a coverage rate of as low as 40X. The mean time from initiating DNA extraction to complete genetic analysis was 2 days (IQR 2-3.25 days). We identified five potential transmission clusters comprising 21 isolates (8.7% of all sequenced strains). Combining ONT WGS data with epidemiological data, >70% (15/21) of the isolates originated from patients with potential healthcare transmission links. Via a stand-alone ONT pipeline, we detected potentially transmitted HAI pathogens rapidly and accurately, aligning closely with epidemiological data. Our low-resource method has the potential to assist in the efficient detection and deployment of preventative measures against HAI transmission.

COfEE: A comprehensive ontology for event extraction from text

Large volumes of data are constantly being published on the web; however, the majority of this data is often unstructured, making it difficult to comprehend and interpret. To extract meaningful and structured information from such data, researchers and practitioners have turned to Information Extraction (IE) methods. One of the most challenging IE tasks is Event Extraction (EE), which involves extracting information related to specific incidents and their associated actors from text. EE has broad applications, including building a knowledge base, information retrieval, summarization, and online monitoring systems. Over the past few decades, various event ontologies, such as ACE, CAMEO, and ICEWS, have been developed to define event forms, actors, and dimensions of events observed in text. However, these ontologies have some limitations, such as covering only a few topics like political events, having inflexible structures in defining argument roles, lacking analytical dimensions, and insufficient gold-standard data. To address these concerns, we propose a new event ontology, COfEE, which integrates expert domain knowledge, previous ontologies, and a data-driven approach for identifying events from text. COfEE comprises two hierarchy levels (event types and event sub-types) that include new categories related to environmental issues, cyberspace, criminal activity, and natural disasters that require real-time monitoring. In addition, dynamic roles are defined for each event sub-type to capture various dimensions of events. The proposed ontology is evaluated on Wikipedia events, and it is shown to be comprehensive and general. Furthermore, to facilitate the preparation of gold-standard data for event extraction, we present a language-independent online tool based on COfEE. A gold-standard dataset annotated by ten human experts consisting of 24,000 news articles in Persian according to the COfEE ontology is also prepared. To diversify the data, news articles from the Wikipedia event portal and the 100 most popular Persian news agencies between 2008 and 2021 were collected. Finally, we introduce a supervised method based on deep learning techniques to automatically extract relevant events and their corresponding actors.

Leveraging Large Language Models for Knowledge-free Weak Supervision in Clinical Natural Language Processing.

The performance of deep learning-based natural language processing systems is based on large amounts of labeled training data which, in the clinical domain, are not easily available or affordable. Weak supervision and in-context learning offer partial solutions to this issue, particularly using large language models (LLMs), but their performance still trails traditional supervised methods with moderate amounts of gold-standard data. In particular, inferencing with LLMs is computationally heavy. We propose an approach leveraging fine-tuning LLMs and weak supervision with virtually no domain knowledge that still achieves consistently dominant performance. Using a prompt-based approach, the LLM is used to generate weakly-labeled data for training a downstream BERT model. The weakly supervised model is then further fine-tuned on small amounts of gold standard data. We evaluate this approach using Llama2 on three different n2c2 datasets. With no more than 10 gold standard notes, our final BERT models weakly supervised by fine-tuned Llama2-13B consistently outperformed out-of-the-box PubMedBERT by 4.7-47.9% in F1 scores. With only 50 gold standard notes, our models achieved close performance to fully fine-tuned systems.

CelFiE-ISH: a probabilistic model for multi-cell type deconvolution from single-molecule DNA methylation haplotypes

Deconvolution methods infer quantitative cell type estimates from bulk measurement of mixed samples including blood and tissue. DNA methylation sequencing measures multiple CpGs per read, but few existing deconvolution methods leverage this within-read information. We develop CelFiE-ISH, which extends an existing method (CelFiE) to use within-read haplotype information. CelFiE-ISH outperforms CelFiE and other existing methods, achieving 30% better accuracy and more sensitive detection of rare cell types. We also demonstrate the importance of marker selection and of tailoring markers for haplotype-aware methods. While here we use gold-standard short-read sequencing data, haplotype-aware methods will be well-suited for long-read sequencing.

NCI MATCH (EAY131) arm Z1D: Retrospective evaluation of attributions of adverse events experienced on nivolumab in patients with MSI-high tumors.

e14711 Background: Machine learning algorithms for prediction of Adverse Events (AEs) due to immune checkpoint inhibitors (ICIs) rely on clinically annotated data to serve as gold standards. Clinical trials are the best sources of such data, but audits to assess validity are rare, compromising future machine learning efforts. We reviewed AEs reported on the National Cancer Institute Molecular Analysis for Therapy Choice (NCI-MATCH) EAY-131 Arm Z1D. Patients matched to Arm Z1D due to MSI-high status were treated with nivolumab. Routine and serious AEs were reported to NCI. Attribution data from AEs on 29 patients with tumor genomics suitable for machine learning approaches were obtained and reviewed retrospectively. Methods: Documented AEs were sorted by grade via the Common Terminology for Adverse Events (CTCAE) system and categorized as related or unrelated to ICIs. Primary source data reported to NCI were reviewed. If any ambiguity was found, trial sites were queried to provide additional information and documentation. Decisions were made by an experienced NCI medical officer with the benefit of additional data and site responses. All serious AEs and a subset of ambiguous routine AEs were assessed to determine if attribution conformed to expected attribution per the nivolumab package insert. Uncertain attribution required additional queries, including email or phone correspondence. Grade 1 or 2 events were only evaluated to determine if any attribution to ICIs was possible. If a higher grade event was attributed to ICIs, there was no further query for lower grade events on the same patient. Results: A total of 639 Grade 1, 152 Grade 2, 75 Grade 3, 10 Grade 4, and 5 Grade 5 events were reported among the 29 patients. One Grade 5 event, 5 Grade 4 events, 13 Grade 3 Events, 9 Grade 2 events, and 6 Grade 1 events were considered less likely to be attributed to ICIs compared with the initial sponsor or site investigator assessment of possibly, probably, or definitely related. Two Grade 4 events, 1 Grade 2 event, and 8 Grade 1 events were considered to be possibly, probably, or definitely due to ICIs compared with the initial sponsor or site assessment of less likely or unrelated. Site responsiveness to retrospective queries from the sponsor varied. Additional lab values, primary source documentation, and communication with site investigators were necessary to make modifications. A total of 11 AEs required additional queries to site investigators. Conclusions: Retrospective reviews of real-time site and sponsor assessments of attribution exhibit significant modification of initial attribution to ICIs in this review of 29 patients on Arm Z1D of the NCI MATCH trial. Machine learning approaches for prediction of AEs on patients treated with ICIs may require careful assessments of primary source data to develop adequate training sets, given the paucity of gold standard data available.

High-throughput functional mapping of variants in an arrhythmia gene, KCNE1, reveals novel biology

BackgroundKCNE1 encodes a 129-residue cardiac potassium channel (IKs) subunit. KCNE1 variants are associated with long QT syndrome and atrial fibrillation. However, most variants have insufficient evidence of clinical consequences and thus limited clinical utility.MethodsIn this study, we leveraged the power of variant effect mapping, which couples saturation mutagenesis with high-throughput sequencing, to ascertain the function of thousands of protein-coding KCNE1 variants.ResultsWe comprehensively assayed KCNE1 variant cell surface expression (2554/2709 possible single-amino-acid variants) and function (2534 variants). Our study identified 470 loss- or partial loss-of-surface expression and 574 loss- or partial loss-of-function variants. Of the 574 loss- or partial loss-of-function variants, 152 (26.5%) had reduced cell surface expression, indicating that most functionally deleterious variants affect channel gating. Nonsense variants at residues 56–104 generally had WT-like trafficking scores but decreased functional scores, indicating that the latter half of the protein is dispensable for protein trafficking but essential for channel function. 22 of the 30 KCNE1 residues (73%) highly intolerant of variation (with > 70% loss-of-function variants) were in predicted close contact with binding partners KCNQ1 or calmodulin. Our functional assay data were consistent with gold standard electrophysiological data (ρ = − 0.64), population and patient cohorts (32/38 presumed benign or pathogenic variants with consistent scores), and computational predictors (ρ = − 0.62). Our data provide moderate-strength evidence for the American College of Medical Genetics/Association of Molecular Pathology functional criteria for benign and pathogenic variants.ConclusionsComprehensive variant effect maps of KCNE1 can both provide insight into IKs channel biology and help reclassify variants of uncertain significance.

A general framework for developing computable clinical phenotype algorithms.

To present a general framework providing high-level guidance to developers of computable algorithms for identifying patients with specific clinical conditions (phenotypes) through a variety of approaches, including but not limited to machine learning and natural language processing methods to incorporate rich electronic health record data. Drawing on extensive prior phenotyping experiences and insights derived from 3 algorithm development projects conducted specifically for this purpose, our team with expertise in clinical medicine, statistics, informatics, pharmacoepidemiology, and healthcare data science methods conceptualized stages of development and corresponding sets of principles, strategies, and practical guidelines for improving the algorithm development process. We propose 5 stages of algorithm development and corresponding principles, strategies, and guidelines: (1) assessing fitness-for-purpose, (2) creating gold standard data, (3) feature engineering, (4) model development, and (5) model evaluation. This framework is intended to provide practical guidance and serve as a basis for future elaboration and extension.

An Optimization Approach for Creating Application-specific Ultrasound Speckle Tracking Algorithms

ObjectiveUltrasound speckle tracking enables in vivo measurement of soft tissue deformation or strain, providing a non-invasive diagnostic tool to quantify tissue health. However, adoption into new fields is challenging since algorithms need to be tuned with gold-standard reference data that are expensive or impractical to acquire. Here, we present a novel optimization approach that only requires repeated measurements, which can be acquired for new applications where reference data might not be readily available or difficult to get hold of. MethodsSoft tissue motion was captured using ultrasound for the medial collateral ligament (MCL) of three quasi-statically loaded porcine stifle joints, and medial ligamentous structures of a dynamically loaded human cadaveric knee joint. Using a training subset, custom speckle tracking algorithms were created for the porcine and human ligaments using surrogate optimization, which aimed to maximize repeatability by minimizing the normalized standard deviation of calculated strain maps for repeat measurements. An unseen test subset was then used to validate the tuned algorithms by comparing the ultrasound strains to digital image correlation (DIC) surface strains (porcine specimens) and length change values of the optically tracked ligament attachments (human specimens). ResultsAfter 1500 iterations, the optimization routine based on the porcine and human training data converged to similar values of normalized standard deviations of repeat strain maps (porcine: 0.19, human: 0.26). Ultrasound strains calculated for the independent test sets using the tuned algorithms closely matched the DIC measurements for the porcine quasi-static measurements (R > 0.99, RMSE < 0.59%) and the length change between the tracked ligament attachments for the dynamic human dataset (RMSE < 6.28%). Furthermore, strains in the medial ligamentous structures of the human specimen during flexion showed a strong correlation with anterior/posterior position on the ligaments (R > 0.91). ConclusionAdjusting ultrasound speckle tracking algorithms using an optimization routine based on repeatability led to robust and reliable results with low RMSE for the medial ligamentous structures of the knee. This tool may be equally beneficial in other soft-tissue displacement or strain measurement applications and can assist in the development of novel ultrasonic diagnostic tools to assess soft tissue biomechanics.

Anticipating Job Market Demands—A Deep Learning Approach to Determining the Future Readiness of Professional Skills

Anticipating the demand for professional job market skills needs to consider trends such as automation, offshoring, and the emerging Gig economy, as they significantly impact the future readiness of skills. This article draws on the scientific literature, expert assessments, and deep learning to estimate two indicators of high relevance for a skill’s future readiness: its automatability and offshorability. Based on gold standard data, we evaluate the performance of Support Vector Machines (SVMs), Transformers, Large Language Models (LLMs), and a deep learning ensemble classifier for propagating expert and literature assessments on these indicators of yet unseen skills. The presented approach uses short bipartite skill labels that contain a skill topic (e.g., “Java”) and a corresponding verb (e.g., “programming”) to describe the skill. Classifiers thus need to base their judgments solely on these two input terms. Comprehensive experiments on skewed and balanced datasets show that, in this low-token setting, classifiers benefit from pre-training and fine-tuning and that increased classifier complexity does not yield further improvements.

Online feeding behavior monitoring of individual group-housed grow-finish pigs using a low-frequency RFID electronic feeding system.

Early identification of animals in need of management intervention is critical to maximize animal health and welfare and minimize issues with productivity. Feeding behavior, captured by automated feeding systems, can be used to monitor the health and welfare status of individual pigs. Here, we present a framework for monitoring feeding behavior of grow-finish pigs in real time, using a low-frequency radio frequency identification (RFID) system. Using historical data, an autoregressive linear model for predicting daily time at the feeder was developed and utilized to detect anomalous decreases in feeding behavior associated with health status of the pig. A total of 2,826 pigs were individually monitored with our warning system over the entire grow-finish period, and health warnings were compared to caretaker diagnoses. The system detected 55.7% of the caretaker diagnoses, and on average these events were detected 2.8 d earlier than diagnosis by the caretaker. High numbers of potentially spurious health warnings, generated by the system, can be partly explained by the lack of a reliable and repeatable gold standard reference data set. Results from this work provide a solid basis for monitoring individual animals, but further improvements to the system are necessary for practical implementation.

Identification of patients undergoing chronic kidney replacement therapy in primary and secondary care data: validation study based on OpenSAFELY and UK Renal Registry

ObjectiveTo validate primary and secondary care codes in electronic health records to identify people receiving chronic kidney replacement therapy based on gold standard registry data.DesignValidation study using data from OpenSAFELY...

TacticAI: an AI assistant for football tactics

Identifying key patterns of tactics implemented by rival teams, and developing effective responses, lies at the heart of modern football. However, doing so algorithmically remains an open research challenge. To address this unmet need, we propose TacticAI, an AI football tactics assistant developed and evaluated in close collaboration with domain experts from Liverpool FC. We focus on analysing corner kicks, as they offer coaches the most direct opportunities for interventions and improvements. TacticAI incorporates both a predictive and a generative component, allowing the coaches to effectively sample and explore alternative player setups for each corner kick routine and to select those with the highest predicted likelihood of success. We validate TacticAI on a number of relevant benchmark tasks: predicting receivers and shot attempts and recommending player position adjustments. The utility of TacticAI is validated by a qualitative study conducted with football domain experts at Liverpool FC. We show that TacticAI’s model suggestions are not only indistinguishable from real tactics, but also favoured over existing tactics 90% of the time, and that TacticAI offers an effective corner kick retrieval system. TacticAI achieves these results despite the limited availability of gold-standard data, achieving data efficiency through geometric deep learning.