Artificial Intelligence-Supported Development of Health Guideline Questions.

BACKGROUND: Guideline questions are typically proposed by experts. OBJECTIVE: To assess how large language models (LLMs) can support the development of guideline questions, providing insights on approaches and lessons learned. DESIGN: Two approaches for guideline question generation were assessed: 1) identification of questions conveyed by online search queries and 2) direct generation of guideline questions by LLMs. For the former, the researchers retrieved popular queries on allergic rhinitis using Google Trends (GT) and identified those conveying questions using both manual and LLM-based methods. They then manually structured as guideline questions the queries that conveyed relevant questions. For the second approach, they tasked an LLM with proposing guideline questions, assuming the role of either a patient or a clinician. SETTING: Allergic Rhinitis and its Impact on Asthma (ARIA) 2024 guidelines. PARTICIPANTS: None. MEASUREMENTS: Frequency of relevant questions generated. RESULTS: The authors retrieved 3975 unique queries using GT. From these, they identified 37 questions, of which 22 had not been previously posed by guideline panel members and 2 were eventually prioritized by the panel. Direct interactions with LLMs resulted in the generation of 22 unique relevant questions (11 not previously suggested by panel members), and 4 were eventually prioritized by the panel. In total, 6 of 39 final questions prioritized for the 2024 ARIA guidelines were not initially thought of by the panel. The researchers provide a set of practical insights on the implementation of their approaches based on the lessons learned. LIMITATION: Single case study (ARIA guidelines). CONCLUSION: Approaches using LLMs can support the development of guideline questions, complementing traditional methods and potentially augmenting questions prioritized by guideline panels. PRIMARY FUNDING SOURCE: Fraunhofer Cluster of Excellence for Immune-Mediated Diseases.

Benchmarking table recognition performance on biomedical literature on neurological disorders.

MOTIVATION: Table recognition systems are widely used to extract and structure quantitative information from the vast amount of documents that are increasingly available from different open sources. While many systems already perform well on tables with a simple layout, tables in the biomedical domain are often much more complex. Benchmark and training data for such tables are however very limited. RESULTS: To address this issue, we present a novel, highly curated benchmark dataset based on a hand-curated literature corpus on neurological disorders, which can be used to tune and evaluate table extraction applications for this challenging domain. We evaluate several state-of-the-art table extraction systems based on our proposed benchmark and discuss challenges that emerged during the benchmark creation as well as factors that can impact the performance of recognition methods. For the evaluation procedure, we propose a new metric as well as several improvements that result in a better performance evaluation. AVAILABILITY AND IMPLEMENTATION: The resulting benchmark dataset (https://zenodo.org/record/5549977) as well as the source code to our novel evaluation approach can be openly accessed. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Common data model for COVID-19 datasets.

MOTIVATION: A global medical crisis like the coronavirus disease 2019 (COVID-19) pandemic requires interdisciplinary and highly collaborative research from all over the world. One of the key challenges for collaborative research is a lack of interoperability among various heterogeneous data sources. Interoperability, standardization and mapping of datasets are necessary for data analysis and applications in advanced algorithms such as developing personalized risk prediction modeling. RESULTS: To ensure the interoperability and compatibility among COVID-19 datasets, we present here a common data model (CDM) which has been built from 11 different COVID-19 datasets from various geographical locations. The current version of the CDM holds 4639 data variables related to COVID-19 such as basic patient information (age, biological sex and diagnosis) as well as disease-specific data variables, for example, Anosmia and Dyspnea. Each of the data variables in the data model is associated with specific data types, variable mappings, value ranges, data units and data encodings that could be used for standardizing any dataset. Moreover, the compatibility with established data standards like OMOP and FHIR makes the CDM a well-designed CDM for COVID-19 data interoperability. AVAILABILITY AND IMPLEMENTATION: The CDM is available in a public repo here: https://github.com/Fraunhofer-SCAI-Applied-Semantics/COVID-19-Global-Model. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

The COVID-19 Ontology.

MOTIVATION: The COVID-19 pandemic has prompted an impressive, worldwide response by the academic community. In order to support text mining approaches as well as data description, linking and harmonization in the context of COVID-19, we have developed an ontology representing major novel coronavirus (SARS-CoV-2) entities. The ontology has a strong scope on chemical entities suited for drug repurposing, as this is a major target of ongoing COVID-19 therapeutic development. RESULTS: The ontology comprises 2270 classes of concepts and 38 987 axioms (2622 logical axioms and 2434 declaration axioms). It depicts the roles of molecular and cellular entities in virus-host interactions and in the virus life cycle, as well as a wide spectrum of medical and epidemiological concepts linked to COVID-19. The performance of the ontology has been tested on Medline and the COVID-19 corpus provided by the Allen Institute. AVAILABILITYAND IMPLEMENTATION: COVID-19 Ontology is released under a Creative Commons 4.0 License and shared via https://github.com/covid-19-ontology/covid-19. The ontology is also deposited in BioPortal at https://bioportal.bioontology.org/ontologies/COVID-19. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Core Set of Patient-reported Outcomes in Pancreatic Cancer (COPRAC): An International Delphi Study Among Patients and Health Care Providers.

OBJECTIVE: To establish an international core set of patient-reported outcomes (PROs) selected by both patients and healthcare providers (HCPs) from the United States (US), Europe, and Asia. SUMMARY BACKGROUND DATA: PROs are increasingly recognized in pancreatic cancer studies. There is no consensus on which of the many available PROs are most important. METHODS: A multicenter Delphi study among patients with pancreatic cancer (curative- and palliative-setting) and HCPs in 6 pancreatic centers in the US (Baltimore, Boston), Europe (Amsterdam, Verona), and Asia (Mumbai, Seoul) was performed. In round 1, participants rated the importance of 56 PROs on a 1 to 9 Likert scale. PROs rated as very important (scores 7-9) by the majority (≥80%) of curative- and/or palliative-patients as well as HCPs were included in the core set. PROs not fulfilling these criteria were presented again in round 2, together with feedback on individual and group ratings. Remaining PROs were ranked based on the importance ratings. RESULTS: In total 731 patients and HCPs were invited, 501 completed round 1, and 420 completed both rounds. This included 204 patients in curative-setting, 74 patients in palliative-setting, and 142 HCPs. After 2 rounds, 8 PROs were included in the core set: general quality of life, general health, physical ability, ability to work/do usual activities, fear of recurrence, satisfaction with services/care organization, abdominal complaints, and relationship with partner/family. CONCLUSIONS: This international Delphi study among patients and HCPs established a core set of PROs in pancreatic cancer, which should facilitate the design of future pancreatic cancer trials and outcomes research.

Intravenous calcium infusion in a calving protocol disrupts calcium homeostasis compared with an oral calcium supplement.

Hypocalcemia is a common postpartum condition in dairy cows, which negatively affects health and production. Intravenous Ca infusions are commonly included in calving protocols to prevent or mitigate the effect of hypocalcemia in multiparous cows. Thus, we sought to contrast the effect of intravenous Ca infusion against voluntary oral Ca intake on Ca metabolism. Serum total Ca (tCa) and whole-blood ionized Ca (iCa) were monitored in 24 multiparous Holstein cows after parturition. Precalving diets were formulated with a positive dietary cation-anion difference of 172 mEq/kg of DM and contained 4.1 g of Ca/kg of DM. At parturition, cows were blocked by calving sequence and calcemic status as either normocalcemic (cutoff threshold of iCa ≥1.10 mmol/L) or hypocalcemic (cutoff threshold of iCa <1.10 mmol/L). Cows in each block were randomly assigned to 1 of 2 treatments: either an oral source of Ca (Ca-Oral; n = 12) or an intravenous source of Ca (Ca-IV; n = 12). Cows in the Ca-Oral group were offered a 20-L commercial Ca suspension (48 g of Ca) for voluntary consumption. The supplement contained Ca carbonate, Ca formate, Ca propionate, and other minerals and vitamins (Farm-O-San Reviva, Trouw Nutrition, Amersfoort, the Netherlands). Cows in the Ca-IV group received a 450-mL intravenous Ca solution (13 g of Ca) that contained 298 mg/mL of Ca gluconate, 33 mg/mL of magnesium chloride, and 82 mg/mL of boric acid (AmosCAL, Kommer-Biopharm BV, Heiloo, the Netherlands). Both treatments were initiated within 25 ± 10 min after calving. The oral Ca suspension was offered to cows in a 25-L bucket and was available for 10 min. All cows in the Ca-Oral group voluntarily consumed the entire 20 L of the Ca suspension within 5 min. Blood samples for Ca analyses were collected at 0 (before treatment initiation), 1, 3, 10, and 18 h relative to treatment, and at 0700 and 1900 h for the next 2 consecutive days, to represent the 24-, 36-, 48-, and 60-h sampling time points. In Ca-IV cows, both iCa and tCa concentrations peaked at 1 h (1.54 mmol/L for iCa and 2.85 mmol/L for tCa) and declined to a nadir at 24 h following treatment initiation (0.94 mmol/L for iCa and 1.74 mmol/L for tCa). Although whole-blood iCa and serum tCa were higher at 1 and 3 h in Ca-IV cows, concentrations of iCa were greater for Ca-Oral cows at 18, 24, and 36 h and for tCa at 24 and 36 h. Our data indicate that intravenous Ca infusion immediately induced a state of hypercalcemia followed by lower whole-blood iCa and serum tCa concentrations 24 h later compared with oral Ca.

ROCK inhibitors 2. Improving potency, selectivity and solubility through the application of rationally designed solubilizing groups.

Solubilizing groups have been frequently appended to kinase inhibitor drug molecules when solubility is insufficient for pharmaceutical development. Such groups are usually located at substitution sites that have minimal impact on target activity. In this report we describe the incorporation of solubilizing groups in a class of Rho kinase (ROCK) inhibitors that not only confer improved solubility, but also enhance target potency and selectivity against a closely related kinase, PKA.

ROCK inhibitors 3: Design, synthesis and structure-activity relationships of 7-azaindole-based Rho kinase (ROCK) inhibitors.

Rho kinase (ROCK) inhibitors are potential therapeutic agents for the treatment of a variety of disorders including hypertension, glaucoma and erectile dysfunction. Here we disclose a series of potent and selective ROCK inhibitors based on a substituted 7-azaindole scaffold. Substitution of the 3-position of 7-azaindole led to compounds such as 37, which possess excellent ROCK inhibitory potency and high selectivity against the closely related kinase PKA.

The effect of spiritual interventions addressing existential themes using a narrative approach on quality of life of cancer patients: a systematic review and meta-analysis.

OBJECTIVE: The aim of this study was to examine the effect of spiritual interventions on quality of life of cancer patients. METHODS: We conducted our search on June 6, 2014 in Medline, PsycINFO, Embase, and PubMed. All clinical trials were included that compared standard care with a spiritual intervention that addressed existential themes using a narrative approach. Study quality was evaluated by the Cochrane Risk of Bias Tool. RESULTS: A total of 4972 studies were identified, of which 14 clinical trials (2050 patients) met the inclusion criteria, and 12 trials (1878 patients) were included in the meta-analysis. The overall risk of bias was high. When combined, all studies showed a moderate effect (d) 0.50 (95% CI = 0.20-0.79) 0-2 weeks after the intervention on overall quality of life in favor of the spiritual interventions. Meta-analysis at 3-6 months after the intervention showed a small insignificant effect (0.14, 95% CI = -0.08 to 0.35). Subgroup analysis including only the western studies showed a small effect of 0.17 (95% CI = 0.05-0.29). Including only studies that met the allocation concealment criteria showed an insignificant effect of 0.14 (95% CI = -0.05 to 0.33). CONCLUSIONS: Directly after the intervention, spiritual interventions had a moderate beneficial effect in terms of improving quality of life of cancer patients compared with that of a control group. No evidence was found that the interventions maintained this effect up to 3-6 months after the intervention. Further research is needed to understand how spiritual interventions could contribute to a long-term effect of increasing or maintaining quality of life.

Quality of patient-reported outcome reporting across cancer randomized controlled trials according to the CONSORT patient-reported outcome extension: A pooled analysis of 557 trials.

BACKGROUND: The main objectives of this study were to identify the number of randomized controlled trials (RCTs) including a patient-reported outcome (PRO) endpoint across a wide range of cancer specialties and to evaluate the completeness of PRO reporting according to the Consolidated Standards of Reporting Trials (CONSORT) PRO extension. METHODS: RCTs with a PRO endpoint that had been performed across several cancer specialties and published between 2004 and 2013 were considered. Studies were evaluated on the basis of previously defined criteria, including the CONSORT PRO extension and the Cochrane Collaboration's tool for assessing the risk of bias of RCTs. Analyses were also conducted by the type of PRO endpoint (primary vs secondary) and by the cancer disease site. RESULTS: A total of 56,696 potentially eligible records were scrutinized, and 557 RCTs with a PRO evaluation, enrolling 254,677 patients overall, were identified. PROs were most frequently used in RCTs of breast (n = 123), lung (n = 85), and colorectal cancer (n = 66). Overall, PROs were secondary endpoints in 421 RCTs (76%). Four of 6 evaluated CONSORT PRO items were documented in less than 50% of the RCTs. The level of reporting was higher in RCTs with a PRO as a primary endpoint. The presence of a supplementary report was the only statistically significant factor associated with greater completeness of reporting for both RCTs with PROs as primary endpoints (ß = .19, P = .001) and RCTs with PROs as secondary endpoints (ß = .30, P < .001). CONCLUSIONS: Implementation of the CONSORT PRO extension is equally important across all cancer specialties. Its use can also contribute to revealing the robust PRO design of some studies, which might be obscured by poor outcome reporting.

Preclinical activity of VX-787, a first-in-class, orally bioavailable inhibitor of the influenza virus polymerase PB2 subunit.

VX-787 is a novel inhibitor of influenza virus replication that blocks the PB2 cap-snatching activity of the influenza viral polymerase complex. Viral genetics and X-ray crystallography studies provide support for the idea that VX-787 occupies the 7-methyl GTP (m(7)GTP) cap-binding site of PB2. VX-787 binds the cap-binding domain of the PB2 subunit with a KD (dissociation constant) of 24 nM as determined by isothermal titration calorimetry (ITC). The cell-based EC50 (the concentration of compound that ensures 50% cell viability of an uninfected control) for VX-787 is 1.6 nM in a cytopathic effect (CPE) assay, with a similar EC50 in a viral RNA replication assay. VX-787 is active against a diverse panel of influenza A virus strains, including H1N1pdm09 and H5N1 strains, as well as strains with reduced susceptibility to neuraminidase inhibitors (NAIs). VX-787 was highly efficacious in both prophylaxis and treatment models of mouse influenza and was superior to the neuraminidase inhibitor, oseltamivir, including in delayed-start-to-treat experiments, with 100% survival at up to 96 h postinfection and partial survival in groups where the initiation of therapy was delayed up to 120 h postinfection. At different doses, VX-787 showed a 1-log to >5-log reduction in viral load (relative to vehicle controls) in mouse lungs. Overall, these favorable findings validate the PB2 subunit of the viral polymerase as a drug target for influenza therapy and support the continued development of VX-787 as a novel antiviral agent for the treatment of influenza infection.

Isosteric replacements of the carboxylic acid of drug candidate VX-787: Effect of charge on antiviral potency and kinase activity of azaindole-based influenza PB2 inhibitors.

VX-787 is a first in class, orally bioavailable compound that offers unparalleled potential for the treatment of pandemic and seasonal influenza. As a part of our routine SAR exploration, carboxylic acid isosteres of VX-787 were prepared and tested against influenza A. It was found that the negative charge is important for maintaining potency and selectivity relative to kinase targets. Neutral carboxylic acid replacements generally resulted in compounds that were significantly less potent and less selective relative to the charged species.

Chips in black boxes? Convenience life span, parafood, brandwidth, families, and co-creation.

Any consumer who opens a bag of potato or corn chips (or crisps in the UK) knows there is no time to waste to enjoy or share them. The convenience life span of chips is limited: it is the shelf or storage life and a very limited time once outside the bag. Many technologies converge to generate the desired effect as a black box, not only of the packaging but also of the chips themselves. The concept of paratext can be applied to printed messages on the package, including the brand name and other texts like advertising (epitexts), which can be expanded into the concept of parafood. These concepts help to discuss technological developments and interpret why this has recently become a negotiation zone for co-creation (see the Do us a flavor campaigns). They are symptoms of changing relations between production, research and development, marketing, and consumption. This paper pays special attention to back stories, underdog brand biographies and narratives about origin. The concept of brandwidth is introduced to sensitize about the limits of combining different stories about chips. A recent brand biography, a family history and a cookery book are used to discuss the phenomenon of cooking with Fritos. Together with the concepts of parafood, brandwidth and black boxes, more reflection and dialogue about the role of history and heritage in marketing put new challenging perspectives on the agenda.

Current state of data stewardship tools in life science.

In today's data-centric landscape, effective data stewardship is critical for facilitating scientific research and innovation. This article provides an overview of essential tools and frameworks for modern data stewardship practices. Over 300 tools were analyzed in this study, assessing their utility, relevance to data stewardship, and applicability within the life sciences domain.

Impact of Short-Course Palliative Radiation Therapy on Pancreatic Cancer-Related Pain: Prospective Phase 2 Nonrandomized PAINPANC Trial.

PURPOSE: Clinical evidence is limited regarding palliative radiation therapy for relieving pancreatic cancer-related pain. We prospectively investigated pain response after short-course palliative radiation therapy in patients with moderate-to-severe pancreatic cancer-related pain. METHODS AND MATERIALS: In this prospective phase 2 single center nonrandomized trial, 30 patients with moderate-to-severe pain (5-10, on a 0-10 scale) of pancreatic cancer refractory to pain medication, were treated with a short-course palliative radiation therapy; 24 Gy in 3 weekly fractions (2015-2018). Primary endpoint was defined as a clinically relevant average decrease of ≥2 points in pain severity, compared with baseline, within 7 weeks after the start of treatment. Secondary endpoint was global quality of life (QoL), with a clinically relevant increase of 5 to 10 points (0-100 scale). Pain severity reduction and QoL were assessed 9 times using the Brief Pain Inventory and European Organization for Research and Treatment of Cancer Quality of Life Questionnaire-C15-PAL, respectively. Both outcomes were analyzed using joint modeling. In addition, acute toxicity based on clinician reporting and overall survival (OS) were assessed. RESULTS: Overall, 29 of 30 patients (96.7%) received palliative radiation therapy. At baseline, the median oral morphine equivalent daily dose was 129.5 mg (range, 20.0-540.0 mg), which decreased to 75.0 mg (range, 15.0-360.0 mg) after radiation (P = .021). Pain decreased on average 3.15 points from baseline to 7 weeks (one-sided P = .045). Patients reported a clinically relevant mean pain severity reduction from 5.9 to 3.8 points (P = .011) during the first 3 weeks, which further decreased to 3.2 until week 11, ending at 3.4 (P = .006) in week 21 after the first radiation therapy fraction. Global QoL significantly improved from 50.5 to 60.8 during the follow-up period (P = .001). Grade 3 acute toxicity occurred in 3 patients and no grade 4 to 5 toxicity was observed. Median OS was 11.8 weeks, with a 13.3% 1-year actuarial OS rate. CONCLUSIONS: Short-course palliative radiation therapy for pancreatic cancer-related pain was associated with rapid, clinically relevant reduction in pain severity, and clinically relevant improvement in global QoL, with mostly mild toxicity.

Concepts for the Development of Person-Centered, Digitally Enabled, Artificial Intelligence-Assisted ARIA Care Pathways (ARIA 2024).

The traditional healthcare model is focused on diseases (medicine and natural science) and does not acknowledge patients' resources and abilities to be experts in their own lives based on their lived experiences. Improving healthcare safety, quality, and coordination, as well as quality of life, is an important aim in the care of patients with chronic conditions. Person-centered care needs to ensure that people's values and preferences guide clinical decisions. This paper reviews current knowledge to develop (1) digital care pathways for rhinitis and asthma multimorbidity and (2) digitally enabled, person-centered care.1 It combines all relevant research evidence, including the so-called real-world evidence, with the ultimate goal to develop digitally enabled, patient-centered care. The paper includes (1) Allergic Rhinitis and its Impact on Asthma (ARIA), a 2-decade journey, (2) Grading of Recommendations, Assessment, Development and Evaluation (GRADE), the evidence-based model of guidelines in airway diseases, (3) mHealth impact on airway diseases, (4) From guidelines to digital care pathways, (5) Embedding Planetary Health, (6) Novel classification of rhinitis and asthma, (7) Embedding real-life data with population-based studies, (8) The ARIA-EAACI (European Academy of Allergy and Clinical Immunology) strategy for the management of airway diseases using digital biomarkers, (9) Artificial intelligence, (10) The development of digitally enabled, ARIA person-centered care, and (11) The political agenda. The ultimate goal is to propose ARIA 2024 guidelines centered around the patient to make them more applicable and sustainable.

Systematic review reveals limitations of studies evaluating health-related quality of life after potentially curative treatment for esophageal cancer.

PURPOSE: To (1) assess the quality of studies evaluating Health-Related Quality of Life (HRQL) after potentially curative treatment for esophageal cancer, and (2) to identify high-quality studies that provide robust HRQL results. METHODS: A systematic literature search is to identify studies evaluating HRQL with a validated multidimensional patient-reported outcome measure. Study quality focused on study design, risk of bias, HRQL outcome reporting, and additional issues of reporting and methodology. A study was deemed high-quality if criteria for robust study methodology and robust HRQL outcome reporting were met. RESULTS: Six RCTs, 12 cohort studies, 13 case-series, and 31 cross-sectional studies were identified. Overall risk of bias was high. Reporting of HRQL concept (e.g., a priori hypothesis), methodology (e.g., reasons for missing data), and interpretation (e.g., clinical significance) was often absent or unclear. Additional issues of reporting (e.g., unclear treatment descriptions) and methodology (e.g., no control for multiple testing) were identified. Four studies (6%) met the criteria for robust study methodology, and 26 studies (42%) met the criteria for robust HRQL outcome reporting. We identified three high-quality studies (5%)--two RCTs and one case-series--capable of providing robust results. CONCLUSIONS: Current evidence is restricted in its ability to inform practice on HRQL after potentially curative treatment for esophageal cancer.

Effects of mobile phone electromagnetic fields on brain waves in healthy volunteers.

The interaction between biological tissue and electromagnetic fields (EMF) is a topic of increasing interest due to the rising prevalence of background EMF in the past decades. Previous studies have attempted to measure the effects of EMF on brainwaves using EEG recordings, but are typically hampered by experimental and environmental factors. In this study, we present a framework for measuring the impact of EMF on EEG while controlling for these factors. A Bayesian statistical approach is employed to provide robust statistical evidence of the observed EMF effects. This study included 32 healthy participants in a double-blinded crossover counterbalanced design. EEG recordings were taken from 63 electrodes across 6 brain regions. Participants underwent a measurement protocol comprising two 18-min sessions with alternating blocks of eyes open (EO) and eyes closed (EC) conditions. Group 1 (n = 16) had EMF during the first session and sham during the second session; group 2 (n = 16) had the opposite. Power spectral density plots were generated for all sessions and brain regions. The Bayesian analysis provided statistical evidence for the presence of an EMF effect in the alpha band power density in the EO condition. This measurement protocol holds potential for future research on the impact of novel transmission protocols.

Design of the formalized and integrated Alzheimer's Disease Ontology and its application in retrieving textual data via text mining.

As one of the leading causes for dementia in the population, it is imperative that we discern exactly why Alzheimer's disease (AD) has a strong molecular association with beta-amyloid and tau. Although a clear understanding about etiology and pathogenesis of AD remains unsolved, scientists worldwide have dedicated significant efforts to discovering the molecular interactions linked to the pathological characteristics and potential treatments. Knowledge representations, such as domain ontologies encompassing our current understanding about AD, could greatly assist and contribute to disease research. This paper describes the construction and application of the integrated Alzheimer's Disease Ontology (ADO), combining selected concepts from the former version of the ADO and the Alzheimer's Disease Mapping Ontology (ADMO). In addition to the existing entities available from these knowledge models, essential knowledge about AD from public sources, such as newly discovered risk factor genes and novel treatments, was also integrated. The ADO can also be leveraged in text mining scenarios given that it is conceptually enriched with domain-specific knowledge as well as their relations. The integrated ADO consists of 39 855 total axioms. The ontology covers many aspects of the AD domain, including risk factor genes, clinical features, treatments and experimental models. The ontology complies with the Open Biological and Biomedical Ontology principles and was accepted by the foundry. In this paper, we illustrate the role of the presented ontology in extracting textual information from the SCAIView database and key measures in an ADO-based corpus. Database URL:  https://academic.oup.com/database.

Automated pose estimation reveals walking characteristics associated with lameness in broilers.

Walking ability of broilers can be improved by selective breeding, but large-scale phenotypic records are required. Currently, gait of individual broilers is scored by trained experts, however, precision phenotyping tools could offer a more objective and high-throughput alternative. We studied whether specific walking characteristics determined through pose estimation are linked to gait in broilers. We filmed male broilers from behind, walking through a 3 m × 0.4 m (length × width) corridor one by one, at 3 time points during their lifetime (at 14, 21, and 33 d of age). We used a deep learning model, developed in DeepLabCut, to detect and track 8 keypoints (head, neck, left and right knees, hocks, and feet) of broilers in the recorded videos. Using the keypoints of the legs, 6 pose features were quantified during the double support phase of walking, and 1 pose feature was quantified during steps, at maximum leg lift. Gait was scored on a scale from 0 to 5 by 4 experts, using the videos recorded on d 33, and the broilers were further classified as having either good gait (mean gait score ≤2) or suboptimal gait (mean gait score >2). The relationship of pose features on d 33 with gait was analyzed using the data of 84 broilers (good gait: 57.1%, suboptimal gait: 42.9%). Birds with suboptimal gait had sharper hock joint lateral angles and lower hock-feet distance ratios during double support on d 33, on average. During steps, relative step height was lower in birds with suboptimal gait. Step height and hock-feet distance ratio showed the largest mean deviations in broilers with suboptimal gait compared to those with good gait. We demonstrate that pose estimation can be used to assess walking characteristics during a large part of the productive life of broilers, and to phenotype and monitor broiler gait. These insights can be used to understand differences in the walking patterns of lame broilers, and to build more sophisticated gait prediction models.