-
Journal of Biomedical Informatics Jan 2023Publicly accessible benchmarks that allow for assessing and comparing model performances are important drivers of progress in artificial intelligence (AI). While recent... (Review)
Review
Publicly accessible benchmarks that allow for assessing and comparing model performances are important drivers of progress in artificial intelligence (AI). While recent advances in AI capabilities hold the potential to transform medical practice by assisting and augmenting the cognitive processes of healthcare professionals, the coverage of clinically relevant tasks by AI benchmarks is largely unclear. Furthermore, there is a lack of systematized meta-information that allows clinical AI researchers to quickly determine accessibility, scope, content and other characteristics of datasets and benchmark datasets relevant to the clinical domain. To address these issues, we curated and released a comprehensive catalogue of datasets and benchmarks pertaining to the broad domain of clinical and biomedical natural language processing (NLP), based on a systematic review of literature and. A total of 450 NLP datasets were manually systematized and annotated with rich metadata, such as targeted tasks, clinical applicability, data types, performance metrics, accessibility and licensing information, and availability of data splits. We then compared tasks covered by AI benchmark datasets with relevant tasks that medical practitioners reported as highly desirable targets for automation in a previous empirical study. Our analysis indicates that AI benchmarks of direct clinical relevance are scarce and fail to cover most work activities that clinicians want to see addressed. In particular, tasks associated with routine documentation and patient data administration workflows are not represented despite significant associated workloads. Thus, currently available AI benchmarks are improperly aligned with desired targets for AI automation in clinical settings, and novel benchmarks should be created to fill these gaps.
Topics: Humans; Artificial Intelligence; Benchmarking; Natural Language Processing
PubMed: 36539106
DOI: 10.1016/j.jbi.2022.104274 -
Frontiers in Cellular and Infection... 2022Cryptosporidiosis is a zoonotic disease caused by Cryptosporidium infection with the main symptom of diarrhea. The present study performed a metaanalysis to determine... (Meta-Analysis)
Meta-Analysis
INTODUCTION
Cryptosporidiosis is a zoonotic disease caused by Cryptosporidium infection with the main symptom of diarrhea. The present study performed a metaanalysis to determine the global prevalence of Cryptosporidium in Equus animals.
METHODS
Data collection was carried out using Chinese National Knowledge Infrastructure (CNKI), VIP Chinese journal database (VIP), WanFang Data, PubMed, and ScienceDirect databases, with 35 articles published before 2021 being included in this systematic analysis. This study analyzed the research data through subgroup analysis and univariate regression analysis to reveal the factors leading to high prevalence. We applied a random effects model (REM) to the metadata.
RESULTS
The total prevalence rate of Cryptosporidium in Equus was estimated to be 7.59% from the selected articles. The prevalence of Cryptosporidium in female Equus was 2.60%. The prevalence of Cryptosporidium in Equus under 1-year-old was 11.06%, which was higher than that of Equus over 1-year-old (2.52%). In the experimental method groups, the positive rate detected by microscopy was the highest (10.52%). The highest Cryptosporidium prevalence was found in scale breeding Equus (7.86%). The horses had the lowest Cryptosporidium prevalence (7.32%) among host groups. C. muris was the most frequently detected genotype in the samples (53.55%). In the groups of geographical factors, the prevalence rate of Cryptosporidium in Equus was higher in regions with low altitude (6.88%), rainy (15.63%), humid (22.69%), and tropical climates (16.46%).
DISCUSSION
The search strategy use of five databases might have caused the omission of some researches. This metaanalysis systematically presented the global prevalence and potential risk factors of Cryptosporidium infection in Equus. The farmers should strengthen the management of young and female Equus animals, improve water filtration systems, reduce stocking densities, and harmless treatment of livestock manure.
Topics: Female; Animals; Horses; Cryptosporidiosis; Cryptosporidium; Prevalence; Risk Factors; Zoonoses
PubMed: 36506009
DOI: 10.3389/fcimb.2022.1072385 -
BMJ Open Nov 2022To support the Zika virus (ZIKV) Individual Participant Data (IPD) Consortium's efforts to harmonise and analyse IPD from ZIKV-related prospective cohort studies and...
OBJECTIVES
To support the Zika virus (ZIKV) Individual Participant Data (IPD) Consortium's efforts to harmonise and analyse IPD from ZIKV-related prospective cohort studies and surveillance-based studies of pregnant women and their infants and children; we developed and disseminated a metadata survey among ZIKV-IPD Meta-Analysis (MA) study participants to identify and provide a comprehensive overview of study-level heterogeneity in exposure, outcome and covariate ascertainment and definitions.
SETTING
Cohort and surveillance studies that measured ZIKV infection during pregnancy or at birth and measured fetal, infant, or child outcomes were identified through a systematic search and consultations with ZIKV researchers and Ministries of Health from 20 countries or territories.
PARTICIPANTS
Fifty-four cohort or active surveillance studies shared deidentified data for the IPD-MA and completed the metadata survey, representing 33 061 women (11 020 with ZIKV) and 18 281 children.
PRIMARY AND SECONDARY OUTCOME MEASURES
Study-level heterogeneity in exposure, outcome and covariate ascertainment and definitions.
RESULTS
Median study sample size was 268 (IQR=100, 698). Inclusion criteria, follow-up procedures and exposure and outcome ascertainment were highly heterogenous, differing meaningfully across regions and multisite studies. Enrolment duration and follow-up for children after birth varied before and after the declaration of the Public Health Emergency of International Concern (PHEIC) and according to the type of funding received.
CONCLUSION
This work highlights the logistic and statistical challenges that must be addressed to account for the multiple sources of within-study and between-study heterogeneity when conducting IPD-MAs of data collected in the research response to emergent pathogens like ZIKV.
Topics: Child; Female; Humans; Infant; Infant, Newborn; Pregnancy; Metadata; Parturition; Pregnancy Complications, Infectious; Pregnant Women; Prospective Studies; Zika Virus; Zika Virus Infection; Meta-Analysis as Topic
PubMed: 36414312
DOI: 10.1136/bmjopen-2022-064362 -
World Neurosurgery Jan 2023Pediatric brainstem lesions are diagnoses that require tissue sampling to advance our understanding of them and their management. Frameless, robot-assisted biopsy of... (Meta-Analysis)
Meta-Analysis Review
BACKGROUND
Pediatric brainstem lesions are diagnoses that require tissue sampling to advance our understanding of them and their management. Frameless, robot-assisted biopsy of these lesions has emerged as a novel, viable biopsy approach. Correspondingly, the aim of this study was to quantitively and qualitatively summarize the contemporary literature regarding the likelihood of achieving tumor diagnosis and experiencing any postoperative complications.
METHODS
Searches of 7 electronic databases from inception to September 2022 were conducted following Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. Articles were screened against prespecified criteria. Outcomes were pooled by random-effects meta-analyses of proportions where possible.
RESULTS
A total of 8 cohort studies satisfied all criteria. They described 99 pediatric patients with brainstem lesions in whom frameless, robot-assisted biopsy was involved in their work-up. There were 62 (63%) male and 37 (37%) female patients with a median age of 9 years at time of biopsy. Overall, all patients had sufficient tissue obtained by initial biopsy for evaluation. Pooled estimate of achieving tumor diagnosis was 100% (95% confidence interval [CI] 97%-100%) across all studies with a high degree of certainty. Across all studies, there were no cases of procedure-related mortality. The pooled estimates of transient and permanent complications after biopsy were 10% (95% CI 4%-19%) and 0% (95% CI 0%-2%), respectively, of very low and low degrees of certainty each.
CONCLUSIONS
The contemporary metadata demonstrates the frameless, robot-assisted biopsy of pediatric brainstem lesions is both effective and safe when performed in an experienced setting. Further research is needed to augment robot and automated technologies into workup algorithms.
Topics: Humans; Male; Child; Female; Robotic Surgical Procedures; Biopsy; Brain Stem Neoplasms
PubMed: 36307039
DOI: 10.1016/j.wneu.2022.10.071 -
Journal of Personalized Medicine Oct 2022Predicting tooth loss is a persistent clinical challenge in the 21st century. While an emerging field in dentistry, computational solutions that employ machine learning... (Review)
Review
Predicting tooth loss is a persistent clinical challenge in the 21st century. While an emerging field in dentistry, computational solutions that employ machine learning are promising for enhancing clinical outcomes, including the chairside prognostication of tooth loss. We aimed to evaluate the risk of bias in prognostic prediction models of tooth loss that use machine learning. To do this, literature was searched in two electronic databases (MEDLINE via PubMed; Google Scholar) for studies that reported the accuracy or area under the curve (AUC) of prediction models. AUC measures the entire two-dimensional area underneath the entire receiver operating characteristic (ROC) curves. AUC provides an aggregate measure of performance across all possible classification thresholds. Although both development and validation were included in this review, studies that did not assess the accuracy or validation of boosting models (AdaBoosting, Gradient-boosting decision tree, XGBoost, LightGBM, CatBoost) were excluded. Five studies met criteria for inclusion and revealed high accuracy; however, models displayed a high risk of bias. Importantly, patient-level assessments combined with socioeconomic predictors performed better than clinical predictors alone. While there are current limitations, machine-learning-assisted models for tooth loss may enhance prognostication accuracy in combination with clinical and patient metadata in the future.
PubMed: 36294820
DOI: 10.3390/jpm12101682 -
Journal of Neurosurgery. Pediatrics Dec 2022Resection and disconnection surgeries for epilepsy in the pediatric demographic (patients ≤ 18 years of age) are two separate, definitive intervention options in... (Meta-Analysis)
Meta-Analysis
OBJECTIVE
Resection and disconnection surgeries for epilepsy in the pediatric demographic (patients ≤ 18 years of age) are two separate, definitive intervention options in medically refractory cases. Questions remain regarding the role of surgery when seizures persist after an initial incomplete surgery. The aim of this study was to review the contemporary literature and summarize the metadata on the outcomes of repeat surgery in this specific demographic.
METHODS
Searches of seven electronic databases from inception to July 2022 were conducted using PRISMA guidelines. Articles were screened using prespecified criteria. Metadata from the articles were abstracted and pooled by random-effects meta-analysis of proportions.
RESULTS
Eleven studies describing 12 cohorts satisfied all criteria, reporting outcomes of 170 pediatric patients with epilepsy who underwent repeat resection or disconnection surgery. Of these patients, 55% were male, and across all studies, median ages at initial and repeat surgeries were 7.2 and 9.4 years, respectively. The median follow-up duration after repeat surgery was 47.7 months. The most commonly reported etiology for epilepsy was cortical dysplasia. Overall, the estimated incidence of complete seizure freedom (Engel class I) following repeat surgery was 48% (95% CI 40%-56%, p value for heterogeneity = 0.93), and the estimated incidence of postoperative complications following repeat surgery was 25% (95% CI 12%-39%, p = 0.04). There were six cohorts each that described outcomes for repeat resection and repeat disconnection surgeries. There was no statistical difference between these two subgroups with respect to estimated incidence of complete seizure freedom (p value for interaction = 0.92), but postoperative complications were statistically more common following repeat resection (p ≤ 0.01).
CONCLUSIONS
For both resection and disconnection surgeries, repeat epilepsy surgery in children is likely to confer complete seizure freedom in approximately half of the patients who experience unsuccessful initial incomplete epilepsy surgery. More data are needed to elucidate the impact on efficacy based on surgical approach selection. Judicious discussion and planning between the patient, family, and a multidisciplinary team of epilepsy specialists is recommended to optimize expectations and outcomes in this setting.
Topics: Child; Humans; Male; Female; Reoperation; Treatment Outcome; Epilepsy; Malformations of Cortical Development; Postoperative Complications; Retrospective Studies; Drug Resistant Epilepsy; Electroencephalography
PubMed: 36242580
DOI: 10.3171/2022.9.PEDS22344 -
Value in Health : the Journal of the... Mar 2023Real-world evidence (RWE) studies are increasingly being used to support healthcare decisions. Various frameworks, tools, and checklists exist for ensuring quality of... (Review)
Review
Use of Structured Template and Reporting Tool for Real-World Evidence for Critical Appraisal of the Quality of Reporting of Real-World Evidence Studies: A Systematic Review.
OBJECTIVES
Real-world evidence (RWE) studies are increasingly being used to support healthcare decisions. Various frameworks, tools, and checklists exist for ensuring quality of real-world data, designing robust studies, and assessing potential for bias. In January 2021, Structured Template and Reporting Tool for RWE (STaRT-RWE) was released to further reduce ambiguity, assumptions, and misinterpretation while planning, implementing, and reporting RWE studies of the safety and effectiveness of treatments. The objective of this study was to identify gaps in the reporting quality of published RWE studies by using this template for critical appraisal.
METHODS
Two reviewers conducted a keyword search on PubMed for free-full-text research articles using real-world data, RWE design, and safety with or without effectiveness outcomes of a medicinal product or intervention in humans of any age or gender, published in English between January 13, 2021, and January 13, 2022. Assessment of risk of bias was done using Assessment of Real-World Observational Studies critical appraisal tool. Deficiencies in methods and findings as per STaRT-RWE template were reported as frequencies.
RESULTS
A total of 54 of 2374 retrieved studies were included in the review. Based on the STaRT-RWE template, the studies inadequately reported empirically defined covariates, power and sample size calculation, attrition, sensitivity analyses, index date (day 0) defining criterion, predefined covariates, outcome, metadata about data source and software, objective, inclusion and exclusion criteria, analysis specifications, and follow-up.
CONCLUSIONS
The use of STaRT-RWE template along with its tables, design diagram, and library of published studies has a potential of improving robustness of RWE studies.
Topics: Humans; Bias; Checklist
PubMed: 36210293
DOI: 10.1016/j.jval.2022.09.003 -
Birth Defects Research Oct 2022The dynamics and complexities of in utero fetal development create significant challenges in transitioning from lab animal-centric developmental toxicity testing methods...
BACKGROUND
The dynamics and complexities of in utero fetal development create significant challenges in transitioning from lab animal-centric developmental toxicity testing methods to assessment strategies based on new approach methodologies (NAMs). Nevertheless, considerable progress is being made, stimulated by increased research investments and scientific advances, such as induced pluripotent stem cell-derived models. To help identify developmental toxicity NAMs for toxicity screening and potential funding through the American Chemistry Council's Long-Range Research Initiative, a systematic literature review was conducted to better understand the current landscape of developmental toxicity NAMs.
METHODS
Scoping review tools were used to systematically survey the literature (2010-2021; ~18,000 references identified), results and metadata were then extracted, and a user-friendly interactive dashboard was created.
RESULTS
The data visualization dashboard, developed using Tableau® software, is provided as a free, open-access web tool. This dashboard enables straightforward interactive queries and visualizations to identify trends and to distinguish and understand areas or NAMs where research has been most, or least focused.
CONCLUSIONS
Herein, we describe the approach and methods used, summarize the benefits and challenges of applying the systematic-review techniques, and highlight the types of questions and answers for which the dashboard can be used to explore the many different facets of developmental toxicity NAMs.
Topics: Animals; Software; Toxicity Tests; United States
PubMed: 36205106
DOI: 10.1002/bdr2.2075 -
PeerJ. Computer Science 2022Understanding the complexity of restricted research data is vitally important in the current new era of Open Science. While the FAIR Guiding Principles have been...
Understanding the complexity of restricted research data is vitally important in the current new era of Open Science. While the FAIR Guiding Principles have been introduced to help researchers to make data Findable, Accessible, Interoperable and Reusable, it is still unclear how the notions of FAIR and Openness can be applied in the context of restricted data. Many methods have been proposed in support of the implementation of the principles, but there is yet no consensus among the scientific community as to the suitable mechanisms of making restricted data FAIR. We present here a systematic literature review to identify the methods applied by scientists when researching restricted data in a FAIR-compliant manner in the context of the FAIR principles. Through the employment of a descriptive and iterative study design, we aim to answer the following three questions: (1) What methods have been proposed to apply the FAIR principles to restricted data?, (2) How can the relevant aspects of the methods proposed be categorized?, (3) What is the maturity of the methods proposed in applying the FAIR principles to restricted data?. After analysis of the 40 included publications, we noticed that the methods found, reflect the stages of the Data Life Cycle, and can be divided into the following Classes: Data Collection, Metadata Representation, Data Processing, Anonymization, Data Publication, Data Usage and Post Data Usage. We observed that a large number of publications used 'Access Control' and 'Usage and License Terms' methods, while others such as 'Embargo on Data Release' and the use of 'Synthetic Data' were used in fewer instances. In conclusion, we are presenting the first extensive literature review on the methods applied to confidential data in the context of FAIR, providing a comprehensive conceptual framework for future research on restricted access data.
PubMed: 36091999
DOI: 10.7717/peerj-cs.1038 -
Seizure Oct 2022Multiple hippocampal transection (MHT) is a surgical technique that offers adequate seizure control with minimal perioperative morbidity. However, there is little... (Review)
Review
PURPOSE
Multiple hippocampal transection (MHT) is a surgical technique that offers adequate seizure control with minimal perioperative morbidity. However, there is little evidence available to guide neurosurgeons in selecting this technique for use in appropriate patients. This systematic review analyzes patient-level data associated with MHT for intractable epilepsy, focusing on postoperative seizure control and memory outcomes.
METHODS
The systematic review was conducted in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Relevant articles were identified from 3 databases (PubMed, Medline, Embase) up to August 1, 2021. Inclusion criteria were that the majority of patients had received a diagnosis of intractable epilepsy, the article was written in English, MHT was the primary procedure, and patient-level metadata were included.
RESULTS
Fifty-nine unique patients who underwent MHT were identified across 11 studies. Ten (17%) of 59 patients underwent MHT alone. Forty-three (75%) of 57 patients who had a follow-up 12 months or longer were seizure free at last follow-up. With respect to postoperative verbal memory retention, 9 of 38 (24%) patient test scores did not change, 14 (37%) decreased, and 16 (42%) increased. With respect to postoperative nonverbal memory retention, 12 of 38 (34%) patient test scores did not change, 13 (34%) decreased, and 13 (33%) increased.
CONCLUSION
There are few reported patients analyzed after MHT. Although the neurocognitive benefits of MHT are unproven, this relatively novel technique has shown promise in the management of seizures in patients with intractable epilepsy. However, structured trials assessing MHT in isolation are warranted.
Topics: Drug Resistant Epilepsy; Epilepsy, Temporal Lobe; Hippocampus; Humans; Memory; Postoperative Complications; Seizures; Treatment Outcome
PubMed: 36041364
DOI: 10.1016/j.seizure.2022.08.007