Manual Review Research Articles

Reducing manual targeted LC-MS/MS peak integration using a supervised learning peak evaluation and automated review tool

BackgroundOur laboratory analyses over 100,000 samples for multiple analytes yearly. Correctly annotating and quantifying analytes using targeted MS/MS analysis is crucial for most laboratories. Therefore, mass spectrometric instrument manufacturers include software capable of automatic chromatographic peak detection and integration, which is used for most routine applications. Although this generally works well, mistakes such as accidentally selecting a nearby matrix peak or drawing an incorrect baseline are relatively common. Especially when results are to be used for enforcement, a time-consuming manual review of each integrated peak is still required to obtain reliable results. This work aims to provide a tool that can significantly reduce the manual workload of reviewing peak integration, thereby reducing the time used for manual review while ensuring that errors made by automatic integration can be corrected by human experts. ResultsPeak Evaluation and Automated Review tool, or PEAR review, is a machine learning-type tool that can read automatic integrations from various brands of MS equipment and compare them with a set of examples stored in a database of correct peak integrations provided by analysts that are relevant for the type of analysis. Moreover, the automatic review process checks all available ion transitions for a target compound. With those ingredients, the tool can autonomously decide how a peak should be quantified or whether a human expert should review it. The developed tool was tested on routine data processed with a widely used vendor-specific software, and we found that 85% of all chromatograms were handled automatically by this tool. Only the remaining 15% needed a ‘conventional’ manual review. The qualitative and quantitative performance of the PEAR tool was found to be equivalent to that of expert human integration, underlining its reliability. SignificanceOur findings indicate that 85% of all manual integration checks can be skipped using PEAR. This reduces the often tedious workload of reviewing all peaks in multiple chromatograms while offering the same quality as full human intervention.

Outpatient Medications Deimplemented by the AAP Bronchiolitis Guidelines: An Umbrella Review of Meta-Analyses.

The 2014 AAP Bronchiolitis Guidelines deimplemented or recommended against the routine therapeutic use of albuterol, hypertonic saline, and epinephrine for infants and children presenting in the outpatient setting. Our objective was to perform an umbrella review of all meta-analyses that included outpatient subanalyses or network meta-analyses with medication treatment comparisons to study the clinical benefits of these deimplemented medications in the outpatient (i.e., primary care, urgent care, and emergency department) setting. Searches were performed in the databases PubMed and Scopus and the Web search engine Google Scholar on the following three topics: albuterol and bronchiolitis, epinephrine and bronchiolitis, and hypertonic saline and bronchiolitis. Article types were limited to systematic reviews and meta-analyses with outpatient subanalyses, with English language and age restrictions. The search strategy was based on the Population, Intervention, Comparator, Outcomes, Studies (PICOS) framework. The studies were uploaded to Rayyan, a Web-based platform for managing articles of systematic reviews. Citation tracking and manual review of references were performed for the included studies. The meta-analyses and network meta-analyses were reviewed for outpatient subanalyses focused on clinical responses and risk of hospital admission. A total of 6 meta-analyses for albuterol, 4 meta-analyses for epinephrine, and 11 meta-analyses for hypertonic saline were included. Our review identified evidence from predominantly low and moderate evidence meta-analyses (assessed using GRADE and AMSTAR 2) indicating that all three deimplemented medications exhibit one or more therapeutic effects and benefits for infants with the bronchiolitis syndrome in the outpatient setting. Effect sizes ranged from medium to near medium. These clinical benefits include decreased hospital admissions and lower clinical severity scores. Given the heterogeneity of patients under the umbrella term "acute bronchiolitis" and the potential for some patients to respond clinically to albuterol, hypertonic saline, and epinephrine, current evidence supports conducting therapeutic trials in infants with acute bronchiolitis in outpatient settings. However, further well-designed and adequately powered randomized controlled trials and high-quality meta-analyses are still needed.

The added value of including thyroid nodule features into large language models for automatic ACR TI-RADS classification based on ultrasound reports.

The ACR Thyroid Imaging, Reporting, and Data System (TI-RADS) uses a score based on ultrasound (US) imaging to stratify the risk of nodule malignancy and recommend appropriate follow-up. This study aims to analyze US reports and explore how Natural Language Processing (NLP) leveraging Transformers models can classify ACR TI-RADS from text reports using the description of thyroid nodule features. This retrospective study evaluated 16,847 thyroid-free text reports from our institution. An automated system, followed by manual review by a radiologist, established baseline annotations by assigning ACR TI-RADS categories from 1 to 5. Two types of systems were evaluated and compared in the dataset. The first by performing a multiclass classification to detect the associated ACR TI-RADS, and the second by extracting thyroid nodule features from the textual reports and incorporating them into the classifier. Our study showed that models enhanced with specific features systematically outperformed those without. Particularly, the BERTIN model, to which additional features were added, achieved the highest level of accuracy, with a score of 0.8426. Moreover, we found a correlation between the presence of punctate echogenic foci, a feature often linked to malignant thyroid lesions, and increased ACR TI-RADS scores. The features of the thyroid nodules described in thyroid US reports, such as composition, echogenicity, shape, margin or echogenic foci, help the NLP classifier to predict the associated ACR TI-RADS most accurately.

Comparing commercial pharmacogenetic testing results and recommendations for antidepressants with established CPIC guidelines

IntroductionIncreasingly, pharmacogenetic testing helps providers with medication selection based upon patient-specific DNA results. While several government-funded organizations work towards consensus and standardization for testing and interpretation, compliance to these best practices remains inconsistent. Pharmacogenetic testing companies often develop proprietary practices for interpreting and reporting, which can lead to incongruency of reported results among companies and potential discrepancies in interpretation.MethodsTo identify the differences of commercial pharmacogenetic testing vendors’ interpretation of genotype-to-phenotype translations and medication recommendations from the Clinical Pharmacogenetic Implementation Consortium (CPIC) guidelines, a retrospective manual chart review was completed in a large rural healthcare system that utilizes two institution-approved pharmacogenetic vendors. One hundred patients were evaluated: 50 who completed testing through Company A and 50 who completed testing through Company B. Genes of interest for genotype-to-phenotype translation included CYP2B6, CYP2C19, and CYP2D6. Comparison of medication recommendations for drug-gene pairs sertraline (CYP2B6 and/or CYP2C19), escitalopram (CYP2C19), and paroxetine (CYP2D6) were compared with recommendations from CPIC, with consideration of the CPIC Serotonin Reuptake Inhibitor Antidepressants (SSRI) guideline 2023 update. This was accomplished via a novel binning process to enable comparison of company-provided binned medication recommendations with CPIC guideline recommendations. Briefly, the binning system included three categorizations based upon the relevant CPIC guideline recommendations–no action needed (green), recommend monitoring (yellow) and therapeutic intervention or alternative recommended (red).ResultsThere were 32/250 (12.8%) genotype-to-phenotype translation discrepancies from CPIC guidelines, all from Company A. Of 266 evaluated binned medication recommendations, there were 114 (42.9%) discrepancies between the pharmacogenetic testing companies (Company A: 93 discrepancies, Company B: 21 discrepancies) and CPIC’s guideline based upon comparison with the novel binning system.DiscussionSignificant differences were observed between testing companies’ interpretations and recommendations, which is concerning as these discrepancies could lead to providers making medication decisions that are not supported by CPIC’s clinical practice guidelines. This may result in suboptimal outcomes for patients, leading to patient and provider dissatisfaction and erosion of trust with pharmacogenetic testing. A proposed resolution for the discrepancies in company-to-company interpretation is adherence to the CPIC guidelines and transparency in interpretation practices.

Extraction of clinical data on major pulmonary diseases from unstructured radiologic reports using a large language model.

Despite significant strides in big data technology, extracting information from unstructured clinical data remains a formidable challenge. This study investigated the utility of large language models (LLMs) for extracting clinical data from unstructured radiological reports without additional training. In this retrospective study, 1800 radiologic reports, 600 from each of the three university hospitals, were collected, with seven pulmonary outcomes defined. Three pulmonology-trained specialists discerned the presence or absence of diseases. Data extraction from the reports was executed using Google Gemini Pro 1.0, OpenAI's GPT-3.5, and GPT-4. The gold standard was predicated on agreement between at least two pulmonologists. This study evaluated the performance of the three LLMs in diagnosing seven pulmonary diseases (active tuberculosis, emphysema, interstitial lung disease, lung cancer, pleural effusion, pneumonia, and pulmonary edema) utilizing chest radiography and computed tomography scans. All models exhibited high accuracy (0.85-1.00) for most conditions. GPT-4 consistently outperformed its counterparts, demonstrating a sensitivity of 0.71-1.00; specificity of 0.89-1.00; and accuracy of 0.89 and 0.99 across both modalities, thus underscoring its superior capability in interpreting radiological reports. Notably, the accuracy of pleural effusion and emphysema on chest radiographs and pulmonary edema on chest computed tomography scans reached 0.99. The proficiency of LLMs, particularly GPT-4, in accurately classifying unstructured radiological data hints at their potential as alternatives to the traditional manual chart reviews conducted by clinicians.

The Los Angeles Motor Scale is independently associated with cerebral blood flow < 30% volume in large vessel occlusions.

Mechanical thrombectomy (MT) is the treatment standard for large vessel occlusion (LVO) stroke. Under current guidelines, only patients with smaller ischemic core volumes (ICV) are eligible for MT. Thus, it is of interest to quickly estimate ICV in stroke patients. The Los Angeles Motor Scale (LAMS) is a validated tool used to assess stroke severity directly in the field. This study aims to determine whether LAMS score is also associated with ICV, as defined by the CBF <30% volume on CT perfusion imaging. We performed a retrospective, multicenter cohort study of consecutive patients presenting with LVO stroke from 9/1/2017 to 10/1/2023. The inclusion criteria were patients with (1) stroke caused by large vessel occlusion confirmed on CTA and (2) diagnostically adequate, multimodal pretreatment CT imaging. Demographic, clinical, and imaging data was collected through manual chart review. Both univariate and multivariate analyses were applied to assess associations. A p-value <.05 was considered significant. A total of 283 patients (median age: 69, IQR: 61-78) were included in the study. On multivariate logistic regression analysis, we found that lower admission LAMS score (adjusted OR: 0.511, 95% CI: 0.313-0.834, p = .007) was independently associated with a CBF <30% volume of less than 50cc. Admission LAMS is an independent predictor of a CBF <30% volume of less than 50cc. This demonstrates that LAMS can be used to estimate ICV, which will aid in the early triaging of LVO patients to thrombectomy-capable centers.

Evolution-Aware Constraint Derivation Approach for Software Remodularization

Existing software clustering techniques tend to ignore prior knowledge from domain experts, leading to results (suggested big-bang remodularization actions) that cannot be acceptable to developers. Incorporating domain experts knowledge or constraints during clustering ensures the obtained modularization aligns with developers’ perspectives, enhancing software quality. However, manual review by knowledgeable domain experts for constraint generation is time-consuming and labor-intensive. In this article, we propose an evolution-aware constraint derivation approach, Escort , which automatically derives clustering constraints based on the evolutionary history from the analyzed software. Specifically, Escort can serve as an alternative approach to derive implicit and explicit constraints in situations where domain experts are absent. In the subsequent constrained clustering process, Escort can be considered as a framework to help supplement and enhance various unconstrained clustering techniques to improve their accuracy and reliability. We evaluate Escort based on both quantitative and qualitative analysis. In quantitative validation, Escort , using generated clustering constraints, outperforms seven classic unconstrained clustering techniques. Qualitatively, a survey with developers from five IT companies indicates that 89% agree with Escort ’s clustering constraints. We also evaluate the utility of refactoring suggestions from our constrained clustering approach, with 54% acknowledged by project developers, either implemented or planned for future releases.

Machine learning-assisted literature screening for a medication-use process-related systematic review.

In an effort to expedite the publication of articles, AJHP is posting manuscripts online as soon as possible after acceptance. Accepted manuscripts have been peer-reviewed and copyedited, but are posted online before technical formatting and author proofing. These manuscripts are not the final version of record and will be replaced with the final article (formatted per AJHP style and proofed by the authors) at a later time. This article summarizes a novel methodology of applying machine learning (ML) algorithms trained with external training data to assist with article screening for 2 annual review series related to the medication-use process (MUP) generally and the MUP in ambulatory care settings (ACMUP) specifically. As the literature review for these 2 series grew over time, it became essential for the authors to develop methods to be efficient while still capturing most of the relevant literature. The ML model can be used to predict whether search results are likely to be relevant or not relevant. Results least likely to be relevant can then be excluded without manual screening, allowing research teams to save time that would otherwise be spent reviewing a portion of the search results for inclusion. ML models require a large training dataset typically derived from the unclassified corpus. In this study, the authors demonstrate the efficacy of training the ML model using external training data, which is possible in scenarios such as a systematic review update or ongoing review series such as those for the MUP and ACMUP. The authors ran 3 simulations using screening decisions from previous publications and in-process manuscripts for the MUP and ACMUP review series to test the efficacy of the approach. The simulations were compared to actual manual screening decisions made by the research teams to include or exclude articles using title and abstract text. For each simulation, the authors developed a training dataset using a sample of screening decisions from previous years to predict article relevance in an "unclassified" corpus. In this case, the screening decisions for the unclassified corpus were actually known, allowing us to calculate recall (percent of relevant articles captured) and time saved using the number of articles that would be excluded without manual review. Combined, the ML approach correctly labeled 187 of 192 relevant studies. The 3 simulations included 17,227 unique studies, and using ML the authors demonstrated that 13,201 studies could have been excluded without manual screening while still maintaining recall of relevant articles of 95% or greater. This novel approach is applicable to systematic reviews and ongoing review series, including those for the MUP and ACMUP. Pharmacists have a duty to review and incorporate best practices into their organizations to improve the efficiency and cost of care, optimally utilize technology, and reduce the potential for medication errors. This methodology will allow evidence syntheses for the MUP and other disciplines in pharmacy practice to be published more expeditiously by saving significant time during the article screening step.

Automatically Counting Florida Manatees (Trichechus manatus latirostris) from Drone Images Using Object-Based Image Analysis

Florida manatees (Trichechus manatus latirostris) require frequent and extensive surveys to inform conservation efforts. Crewed aircraft surveys can be costly, dangerous, and logistically complex. Unoccupied aerial systems (UASs) can assist with these issues. While manual review of UAS imagery can be time- and labor-intensive, automated detection of manatees in aerial survey footage can help. We present an object-based image analysis workflow for the automated detection and count of Florida manatees in Google Earth Engine, a free platform for research that allows for scripts and imagery sharing. Training and testing datasets were built from randomly extracted image frames from two stationary, unoccupied aerial system videos over thermal refugia. The workflow captured most manatees (93.98 to 95.62% recall; 4.38 to 6.03% false negative rate), but also counted many objects as manatees incorrectly (4.24 to 14.77% precision; 998.40 to 3,885.54% false positive over the detectable rate). Sun glint, mud plumes, and water close to shore were common causes of false positives. While the automated count was too high, the workflow lays markers over each detection, allowing for quick manual review for more accurate (semi-automated) counts. This study is an early step in automated detection tools for Florida manatees in a cloud-based platform. Future efforts could explore other platforms or may improve this workflow by including new classes for confounding objects.

Performance of Abbreviated Breast MRI in High-Risk Patients in a Tertiary Care Academic Medical Center.

The development of abbreviated breast MRI (AB-MRI) protocols reduce scan times. This paper reports the performance of AB-MRI at a tertiary care public academic medical center in comparison with established literature. This HIPAA-compliant IRB-approved retrospective study reviewed 413 AB-MRI screenings in high-risk patients from June 2020 to March 2023. Data were collected from 3 databases (MagView, Cerner PowerChart, and Prism Primordial). Demographics and overall BI-RADS assessment were recorded. For all positive (BI-RADS 0, 3, 4, 5) examinations, manual review of each case was performed. Performance metrics (sensitivity, specificity, cancer detection rate [CDR], recall rate, positive predictive value [PPV] 3 and negative predictive value [NPV]) were calculated. PubMed and Google Scholar were used to review similar AB-MRI studies to compare performance metrics. There were 413 AB-MRI examinations from 413 unique patients. The majority of cases were audit-negative BI-RADS 1 or 2 (83.8%, 346/413). There were 67 (16.2%, 67/413) audit-positive cases with 3.6% (15/413) BI-RADS 3, 10.9% (45/413) BI-RADS 4, 0.7% (3/413) BI-RADS 5, and 1.0% (4/413) BI-RADS 0. Performance metrics showed a sensitivity of 100.0% (95% CI, 63.1%-100.0%) and a specificity of 85.7% (95% CI, 81.9%-88.9%). The PPV3 was 14.3% (95% CI, 5.1%-23.5%), and the NPV was 100.0% (95% CI, 99.0%-100.0%). The CDR was 19.4 per 1000 screenings. The results are comparable to prior literature and benchmark data. This study demonstrates high sensitivity (100.0%) and NPV (100.0%) of AB-MRI with comparable specificity (85.7%) and CDR (19.4/1000) to the literature, adding support to the use of AB-MRI. Further research is needed to optimize AB-MRI protocols.

Abstract 4125252: Pre-Hematopoietic Stem Cell Transplantation Echocardiographic Indices and Post-Transplant Cardiovascular Outcomes

Introduction: Hematopoietic stem cell transplantation (HSCT) is associated with adverse cardiovascular (CV) events including the development of heart failure (HF) and arrythmias. While transthoracic echocardiogram (TTE) is routinely obtained prior to HSCT, its role in predicting the incidence of HSCT related CV events is poorly understood. Methods: We used data from the Cardiovascular Registry in Bone Marrow Transplantation (CARE-BMT) study, a multicenter observational study of adult patients (aged ≥18 years) who underwent autologous/allogeneic HSCT for malignant or nonmalignant bone marrow disorders at the University of Michigan Health System (UMHS) and Rush University Medical Center from 2008-2019. In this analysis, we included patients from UMHS with a baseline TTE. Data on pre-HSCT TTE parameters and post-HSCT CV outcomes were collected through manual chart review. Left ventricular (LV) function and dimensions were categorized into normal, mildly abnormal, and moderately/severely abnormal based on American Society of Echocardiography guidelines. The primary outcomes were new-onset HF and atrial fibrillation/flutter post-HSCT. Analyses were conducted using a Fine-Gray model adjusted for the pre-HSCT CARE-BMT CV risk score. Results: Of the 2071 patients (mean age at HSCT 55.5 + 12.9 years; 59.5% male) with a pre-HSCT TTE (median 25 days pre-HSCT), 116 (5.6%) and 128 (6.2%) patients experienced HF and atrial fibrillation/flutter, respectively, over a median period of 2.2 years. Greater abnormalities in left ventricular internal diameter at end-diastole (LVIDd) and end-systole (LVIDs) were linearly associated with a higher risk of HF (P-trend 0.018 and 0.004, respectively) (Table). Similarly, moderately/severely abnormal LVIDd was associated with a 2.41-fold (95% CI: 1.07, 5.43) increase in risk of atrial fibrillation/flutter (Table). Pre-HSCT ejection fraction (EF) was not associated with either endpoint. Conclusion: LV dilation, even when mild, was notably associated with increased risk of developing new HF or atrial arrythmias post-HSCT, regardless of EF. Whether evidence of LV dilation should prompt the initiation of guideline directed medical therapy to minimize the risk of incident HF warrants further study.

Using bioinformatics and artificial intelligence to map the cyclin-dependent kinase 4/6 inhibitor biomarker landscape in breast cancer.

A cyclin-dependent kinase 4/6 (CDK4/6) inhibitor combined with endocrine therapy is the standard-of-care for patients with hormone receptor-positive/human epidermal growth factor receptor 2-negative advanced breast cancer. However, not all patients respond to the treatment, resistance often occurs and efficacy outcomes from early breast cancer trials have been mixed. To identify biomarkers associated with CDK4/6 inhibitor response or resistance, we combined bioinformatic-database analyses, artificial intelligence-assisted literature review, and manual literature review (Embase and OVID Medline; search window: January 2012-October 2022) to compile data to comprehensively describe the CDK4/6 inhibitor biomarker landscape. Based on these results, and validation by external experts, we identified 15 biomarkers of clinical importance (AR , AURKA, ERBB2, ESR1, CCNE1, CDKN1A/B, CDK2, CDK6, CDK7, CDK9, FGFR1/2, MYC, PIK3CA/AKT, RB1 and STAT3) that could guide future breast cancer research.

Abstract Su801: Use of Large Language Models to Optimize Clinical Text Analysis for In-Hospital Cardiac Arrest Identification

Introduction/Background: In-hospital cardiac arrest (IHCA) is experienced by approximately 300,000 patients annually in the US. While individual care teams can readily identify IHCA at the bedside, the reporting of these events after the event often lacks consistency and poses a major challenge for hospitals across the US. Accurate and timely reporting of IHCA events is crucial for facilitating QI initiatives, such as CPR quality review, optimizing team performance, and benchmarking IHCA outcomes. Hypothesis: Large language models (LLMs) can be used to identify IHCA events at the patient encounter-level by analyzing targeted note types from the inpatient record with high performance. Methods: Adult (≥18 years old) inpatient encounters at the Hospital of the University of Pennsylvania from 06/2018 to 03/2022 with a reported clinical emergency were identified from a QI database. Discharge summaries and notes associated with the encounters of interest were extracted from the EHR and deidentified using a natural language processing tool. Notes were reviewed by research staff to ascertain true IHCA events. Positive IHCA was defined as the loss of pulses followed by the delivery of CPR. An LLM, GPT4 v. 32K-Chat, was employed on the Penn Medicine Microsoft Azure Databricks environment to compare zero- and few-shot prompting methods to identify IHCA. Model performance was measured using accuracy, precision, recall and F1 scores. Results: Manual chart review determined 96.8% and 3.2% of cases to be IHCA+ and IHCA-, respectively. A balanced pilot dataset of positive and negative cases was created and the accuracy values across strategically sampled note types was checked. Using zero-shot learning with Chain-of-Thought (COT) prompting, the LLM performed with an accuracy of 86%, precision of 77%, recall of 100%, and F1-score of 87%. Using three-shot learning with COT, the LLM performed with an accuracy of 90%, precision of 83%, recall of 100%, and F1-score of 91%. Conclusions: This study demonstrates the potential efficacy of leveraging LLMs to automatically classify IHCA from strategically sampled note types using both zero and few-shot learning approaches. In the future, we will implement temporal sampling of note types and refine model parameters such as output token size and temperature. These findings suggest that such technologies can be effective in real-world clinical settings, providing a scalable solution to improve patient outcomes and quality improvement efforts.

Sialendoscopy as An Alternative Treatment for Sialoliths in Salivary Glands

Introduction: Salivary gland obstruction affects approximately 1% of the general population. Common symptoms include pain and edema that worsen with eating. Sialolithiasis is responsible for 60%-70% of all salivary gland blockages. Sialendoscopy was first introduced in 1988 and is now established for the diagnosis and treatment of salivary duct disorders. Objective: The objective of this review is to evaluate and present the treatment of sialadenitis, including the indications, technique, and results of sialendoscopy. Method: A manual literature review search was carried out in the PubMed and EBSCO databases by two researchers independently, using a combination of keywords and the Boolean terms “Sialolith” AND “Sialendoscope” AND “Sialolithiasis” AND “Salivary gland” Discussion: Sialendoscopy performed under general anesthesia uses small semi-rigid or semi-flexible endoscopes to access the salivary ducts orally and visualize their lumen. Once the sialolith is found, it is trapped with a basket or clamp. Thus avoiding the removal of the affected gland, which reduces the complications associated with it, nerve damage, salivary fistula, sialocele and aesthetic consequences. Conclusion: Sialendoscopy is an innovative and minimally invasive technique that proves to be effective in the diagnosis and treatment of sialolithiasis. Complications derived from sialendoscopy were not significant and this procedure may be superior to sialadenectomy, since it preserves the salivary gland.

Diagnostic Accuracy of Optic Nerve Sheath Diameter Measurement by Ultrasonography for Noninvasive Estimation of Intracranial Hypertension in Traumatic Brain Injury: A Systematic Review and Meta-Analysis.

Intracranial hypertension (IH) is associated with an unfavorable outcome in traumatic brain injury (TBI), and management strategies guided by intracranial pressure monitoring improve prognosis. Owing to the limitations of using invasive devices, measurement of optic nerve sheath diameter (ONSD) by ultrasonography is an alternative noninvasive method. However, its accuracy has not been validated in patients with TBI, so we aim to determine the diagnostic accuracy of measuring ONSD by ultrasonography in patients with TBI to estimate IH, compared with invasive monitoring. Systematic review of electronic databases and manual literature review from inception to June 2023. The analysis included diagnostic accuracy studies of ultrasonographic measurement of ONSD compared with invasive monitoring published in any language and with patients of any age. A qualitative synthesis was performed describing the clinical and methodological characteristics, strengths, limitations, and quality of evidence. In addition, a bivariate random effects model meta-analysis and a hierarchical summary receiver operating characteristics model were performed for the pediatric and adult population separately. Five hundred and forty eight patients of 688 in 16 eligible studies were adults and 120 were children. Pooled sensitivity and specificity of ONSD measurement by ultrasonography were 84% (95% CI, 76%-89%) and 83% (95% CI, 73%-90%), respectively. During the sensitivity analysis, these parameters exhibited consistent values. Pooled area under the curve was 0.91 for adults and 0.76 for children. Optimal threshold for estimating IH was 5.76 mm for adults and 5.78 mm for children. Measurement of ONSD by ultrasonography is a reliable, low-cost, and safe alternative for the estimation of IH with TBI in adults. More robust studies are needed to overcome the high risk of bias and heterogeneity for this analysis.

Research on Behavior Recognition and Online Monitoring System for Liaoning Cashmere Goats Based on Deep Learning.

Liaoning Cashmere Goats are a high-quality dual-purpose breed valued for both their cashmere and meat. They are also a key national genetic resource for the protection of livestock and poultry in China, with their intensive farming model currently taking shape. Leveraging new productivity advantages and reducing labor costs are urgent issues for intensive breeding. Recognizing goatbehavior in large-scale intelligent breeding not only improves health monitoring and saves labor, but also improves welfare standards by providing management insights. Traditional methods of goat behavior detection are inefficient and prone to cause stress in goats. Therefore, the development of a convenient and rapid detection method is crucial for the efficiency and quality improvement of the industry. This study introduces a deep learning-based behavior recognition and online detection system for Liaoning Cashmere Goats. We compared the convergence speed and detection accuracy of the two-stage algorithm Faster R-CNN and the one-stage algorithm YOLO in behavior recognition tasks. YOLOv8n demonstrated superior performance, converging within 50 epochs with an average accuracy of 95.31%, making it a baseline for further improvements. We improved YOLOv8n through dataset expansion, algorithm lightweighting, attention mechanism integration, and loss function optimization. Our improved model achieved the highest detection accuracy of 98.11% compared to other state-of-the-art (SOTA) target detection algorithms. The Liaoning Cashmere Goat Online Behavior Detection System demonstrated real-time detection capabilities, with a relatively low error rate compared to manual video review, and can effectively replace manual labor for online behavior detection. This study introduces detection algorithms and develops the Liaoning Cashmere Goat Online Behavior Detection System, offering an effective solution for intelligent goat management.

Automated Medical Records Review for Mild Cognitive Impairment and Dementia.

Objectives : Unstructured and structured data in electronic health records (EHR) are a rich source of information for research and quality improvement studies. However, extracting accurate information from EHR is labor-intensive. Here we introduce an automated EHR phenotyping model to identify patients with Alzheimer's Disease, related dementias (ADRD), or mild cognitive impairment (MCI). Methods : We assembled medical notes and associated International Classification of Diseases (ICD) codes and medication prescriptions from 3,626 outpatient adults from two hospitals seen between February 2015 and June 2022. Ground truth annotations regarding the presence vs. absence of a diagnosis of MCI or ADRD were determined through manual chart review. Indicators extracted from notes included the presence of keywords and phrases in unstructured clinical notes, prescriptions of medications associated with MCI/ADRD, and ICD codes associated with MCI/ADRD. We trained a regularized logistic regression model to predict the ground truth annotations. Model performance was evaluated using area under the receiver operating curve (AUROC), area under the precision-recall curve (AUPRC), accuracy, specificity, precision/positive predictive value, recall/sensitivity, and F1 score (harmonic mean of precision and recall). Results : Thirty percent of patients in the cohort carried diagnoses of MCI/ADRD based on manual review. When evaluated on a held-out test set, the best model using clinical notes, ICDs, and medications, achieved an AUROC of 0.98, an AUPRC of 0.98, an accuracy of 0.93, a sensitivity (recall) of 0.91, a specificity of 0.96, a precision of 0.96, and an F1 score of 0.93 The estimated overall accuracy for patients randomly selected from EHRs was 99.88%. Conclusion : Automated EHR phenotyping accurately identifies patients with MCI/ADRD based on clinical notes, ICD codes, and medication records. This approach holds potential for large-scale MCI/ADRD research utilizing EHR databases.

Identifying genomic data use with the Data Citation Explorer

Increases in sequencing capacity, combined with rapid accumulation of publications and associated data resources, have increased the complexity of maintaining associations between literature and genomic data. As the volume of literature and data have exceeded the capacity of manual curation, automated approaches to maintaining and confirming associations among these resources have become necessary. Here we present the Data Citation Explorer (DCE), which discovers literature incorporating genomic data that was not formally cited. This service provides advantages over manual curation methods including consistent resource coverage, metadata enrichment, documentation of new use cases, and identification of conflicting metadata. The service reduces labor costs associated with manual review, improves the quality of genome metadata maintained by the U.S. Department of Energy Joint Genome Institute (JGI), and increases the number of known publications that incorporate its data products. The DCE facilitates an understanding of JGI impact, improves credit attribution for data generators, and can encourage data sharing by allowing scientists to see how reuse amplifies the impact of their original studies.

Corporate governance characteristics and involvement in ESG activities: current trends and research directions

PurposeThis study aims to conduct a comprehensive literature review to examine the relationship between corporate governance characteristics and firms’ engagement in environmental, social and governance (ESG) activities. The review focuses specifically on academic papers published in ranked accounting and finance journals.Design/methodology/approachThe analysis combines a structured literature review with citation analysis, topic modeling using a machine learning (ML) approach and a manual review of selected articles published between 2000 and 2021.FindingsThis paper contributes to corporate governance and ESG literature by conducting an in-depth review, offering a comprehensive analysis of the existing findings and identifying future research directions. From the reviewed literature, this paper proposes the following thematic areas: board characteristics, ownership structure and their impact on a company’s engagement in ESG activities; CEO characteristics and their influence on a company’s involvement in ESG activities; corporate governance and ESG as sources for transparency and legitimacy; internal and external assurance of a company’s involvement in ESG activities; and gender diversity and a company’s involvement in ESG activities.Originality/valueThe study provides a comprehensive understanding of corporate governance and ESG literature. The innovative combination of methods, including ML and manual techniques, enhances the ability to identify key research topics and uncover research directions in the field. Moving forward, this paper suggests several promising directions for future research, including examining the influence of emerging technologies on ESG reporting and assessing the impact of regulatory changes and context on the link between corporate governance and firms’ involvement in ESG practices.

Development and validation of a novel AI framework using NLP with LLM integration for relevant clinical data extraction through automated chart review.

The accurate extraction of surgical data from electronic health records (EHRs), particularly operative notes through manual chart review (MCR), is complex, crucial, and time-intensive, limited by human error due to fatigue and the level of training. This study aimed to develop and validate a novel Natural Language Processing (NLP) algorithm integrated with a Large Language Model (LLM; GPT4-Turbo) to automate the extraction of spinal surgery data from EHRs. The algorithm employed a two-stage approach. Initially, a rule-based NLP framework reviewed and classified candidate segments from the text, preserving their reference segments. These segments were then verified in the second stage through the LLM. The primary outcomes of this study were the accurate extraction of surgical data, including the type of surgery, levels operated, number of disks removed, and presence of intraoperative incidental durotomies. Secondary objectives explored time efficiency, tokenization lengths, and costs. The performance of the algorithm was assessed across two validation databases, analyzing metrics such as accuracy, sensitivity, discrimination, F1-score, and precision, with 95% confidence intervals calculated using percentile-based bootstrapping. The NLP + LLM algorithm markedly outperformed all performance metrics, demonstrating significant improvements in time and cost efficiency. These results suggest the potential for widespread adoption of this technology.