Data Collection and Sharing for Pathogen Surveillance: Making Sense of a Fragmented Global System

RAND Europe was commissioned by the Novo Nordisk Foundation to conduct a study on pathogen surveillance and current initiatives. The study aims to provide an overview of the pathogen surveillance space internationally and the stakeholders involved, as well as to understand the strengths and weaknesses of different initiatives, the challenges of pathogen surveillance and how they have been addressed, and how data has been used to inform public health decision making. To do this, a scoping review of pathogen surveillance initiatives was conducted, and ten case studies were developed and selected for further review following a workshop attended by the Novo Nordisk Foundation and RAND Europe study team. Interviews were conducted with individuals involved in pathogen surveillance initiatives to gather additional information to develop case studies, and expert interviews addressed gaps in the pathogen surveillance space and models that would be helpful in filling these gaps.

Enhancing AI-CDSS with U-AnoGAN: Tackling data imbalance

BACKGROUND AND OBJECTIVE

Clinical Decision Support Systems (CDSS) have substantially evolved, aiding healthcare professionals in informed patient care decision-making. The integration of AI, encompassing machine learning and natural language processing, has notably enhanced the capabilities of CDSS. However, a significant challenge remains in addressing data imbalance and the black box nature of AI algorithms, particularly for rare diseases or underrepresented demographic groups. This study aims to propose a model, U-AnoGAN, designed to overcome these hurdles and augment the diagnostic accuracy of AI-integrated CDSS.

METHODS

The U-AnoGAN was trained using masks derived from normal data, focusing on the Covid-19 and pneumonia datasets. Anomaly scores were calculated to assess the model's performance compared to existing AnoGAN-related algorithms. The study also evaluated the model's interpretability through the visualization of abnormal regions.

RESULTS

The results indicated that U-AnoGAN surpassed its counterparts in performance and interpretability. It effectively addressed the data imbalance problem by necessitating only normal data and showcased enhanced diagnostic accuracy. Precision, sensitivity, and specificity values reflected U-AnoGAN's superior capability in accurate disease prediction, diagnosis, treatment recommendations, and adverse event detection.

CONCLUSIONS

U-AnoGAN significantly bolsters the predictive power of AI-integrated CDSS, enabling more precise and timely diagnoses while providing better visualization to potentially overcome the black box problem. This model presents tremendous potential in elevating patient care with advanced AI tools and fostering more accurate and effective decision-making in healthcare environments. As the healthcare sector grapples with escalating data complexity and volume, the importance of models like U-AnoGAN in enhancing CDSS cannot be overstated.

Data Distribution: Normal or Abnormal?

Determining if the frequency distribution of a given data set follows a normal distribution or not is among the first steps of data analysis. Visual examination of the data, commonly by Q-Q plot, although is acceptable by many scientists, is considered subjective and not acceptable by other researchers. One-sample Kolmogorov-Smirnov test with Lilliefors correction (for a sample size ≥ 50) and Shapiro-Wilk test (for a sample size < 50) are common statistical tests for checking the normality of a data set quantitatively. As parametric tests, which assume that the data distribution is normal (Gaussian, bell-shaped), are more robust compared to their non-parametric counterparts, we commonly use transformations (e.g., log-transformation, Box-Cox transformation, etc.) to make the frequency distribution of non-normally distributed data close to a normal distribution. Herein, I wish to reflect on presenting how to practically work with these statistical methods through examining of real data sets.

Machine Learning Approach for Predicting Systemic Lupus Erythematosus in an Oman-Based Cohort

Objectives

This study aimed to design a machine learning-based prediction framework to predict the presence or absence of systemic lupus erythematosus (SLE) in a cohort of Omani patients.

Methods

Data of 219 patients from 2006 to 2019 were extracted from Sultan Qaboos University Hospital's electronic records. Among these, 138 patients had SLE, while the remaining 81 had other rheumatologic diseases. Clinical and demographic features were analysed to focus on the early stages of the disease. Recursive feature selection was implemented to choose the most informative features. The CatBoost classification algorithm was utilised to predict SLE, and the SHAP explainer algorithm was applied on top of the CatBoost model to provide individual prediction reasoning, which was then validated by rheumatologists.

Results

CatBoost achieved an area under the receiver operating characteristic curve score of 0.95 and a sensitivity of 92%. The SHAP algorithm identified four clinical features (alopecia, renal disorders, acute cutaneous lupus and haemolytic anaemia) and the patient's age as having the greatest contribution to the prediction.

Conclusion

An explainable framework to predict SLE in patients and provide reasoning for its prediction was designed and validated. This framework enables clinicians to implement early interventions that will lead to positive healthcare outcomes.

Recommended Implementation of Quantitative Susceptibility Mapping for Clinical Research in The Brain: A Consensus of the ISMRM Electro-Magnetic Tissue Properties Study Group

This article provides recommendations for implementing quantitative susceptibility mapping (QSM) for clinical brain research. It is a consensus of the ISMRM Electro-Magnetic Tissue Properties Study Group. While QSM technical development continues to advance rapidly, the current QSM methods have been demonstrated to be repeatable and reproducible for generating quantitative tissue magnetic susceptibility maps in the brain. However, the many QSM approaches available give rise to the need in the neuroimaging community for guidelines on implementation. This article describes relevant considerations and provides specific implementation recommendations for all steps in QSM data acquisition, processing, analysis, and presentation in scientific publications. We recommend that data be acquired using a monopolar 3D multi-echo GRE sequence, that phase images be saved and exported in DICOM format and unwrapped using an exact unwrapping approach. Multi-echo images should be combined before background removal, and a brain mask created using a brain extraction tool with the incorporation of phase-quality-based masking. Background fields should be removed within the brain mask using a technique based on SHARP or PDF, and the optimization approach to dipole inversion should be employed with a sparsity-based regularization. Susceptibility values should be measured relative to a specified reference, including the common reference region of whole brain as a region of interest in the analysis, and QSM results should be reported with - as a minimum - the acquisition and processing specifications listed in the last section of the article. These recommendations should facilitate clinical QSM research and lead to increased harmonization in data acquisition, analysis, and reporting.

Decreasing Admissions but Increasing Readmissions for Mental Health in-Patient Treatment in Scotland, UK

We have analysed mental health data for in-patient admissions from 1997 to 2021 in Scotland. The number of patient admissions for mental health patients is declining despite population numbers increasing. This is driven by the adult population; child and adolescent numbers are consistent. We find that mental health in-patients are more likely to be from deprived areas: 33 % of patients are from the most deprived areas, compared to only 11 % from the least deprived. The average length of stay for a mental health in-patient is decreasing, with a rise in stays lasting less than a day. The number of mental health patients who have been readmitted within a month fell from 1997 to 2011, then increased to 2021. Despite the average stay length decreasing, the number of overall readmissions is increasing, suggesting patients are having more, shorter stays.

Data Collection and Analysis Methods for Smart Nudging to Promote Physical Activity: Protocol for a Mixed Methods Study

New digital technologies like activity trackers, nudge concepts, and approaches can inspire and improve personal health. There is increasing interest in employing such devices to monitor people's health and well-being. These devices can continually gather and examine health-related information from people and groups in their familiar surroundings. Context-aware nudges can assist people in self-managing and enhancing their health. In this protocol paper, we describe how we plan to investigate what motivates people to engage in physical activity (PA), what influences them to accept nudges, and how participant motivation for PA may be impacted by technology use.

Greek Hospital Data Mining and Analysis

Monitoring the performance of hospitals is a crucial issue related both with the quality of healthcare services and with country's economy. An easy and trustful way of evaluating health systems is through key performance indicators (KPIs). Such indicators are widely used for the identification of gaps in the quality or efficiency of the services provided. The main aim of this study is the analysis of the financial and operational indicators at hospitals in the 3rd and 5th Healthcare Regions of Greece. In addition, through cluster analysis and data visualization we attempt to uncover hidden patterns that may lie within our data. The results of the study support the need for re-evaluation of the assessment methodology of Greek hospitals to identify the weaknesses in the system, while evidently unsupervised learning exposes the potential of group-based decision making.

Usability Engineering of Dynamic Biosignal Displays Using Ventilation Data

The aim of this work is to develop and evaluate a multi-stage procedure model for the identification of use problems and optimization of usability using biosignal data. The concept is divided into 5 steps: 1. static analysis of data to identify use problems; 2. conducting interviews within the context of use and requirements analysis to investigate problems in more detail; 3. developing new interface concepts to implement the requirements and a prototype of an interface including dynamic visualization of data; 4. formative evaluation using an unmoderated remote usability test; 5. usability test with realistic scenarios and influencing factors in the simulation room. The concept was evaluated in the ventilation setting as an example. The procedure allowed the identification of use problems in the ventilation of patients as well as the development of suitable concepts and their evaluation to counteract use problems. To relieve users, ongoing analyses of biosignals with respect to the use problem are to be carried out. To overcome technical barriers, further development is needed in this area.

Usability Testing of a Multi-Level Modeling Framework for Just-in-Time Adaptive Interventions (JITAIs) in Mobile Health

The JITAI is an intervention design to support health behavior change. We designed a multi-level modeling framework for JITAIs and developed a proof-of-concept prototype (POC). This study aimed at investigating the usability of the POC by conducting two usability tests with students. We assessed the usability and the students' workload and success in completing tasks. In the second usability test, however, they faced difficulties in completing the tasks. We will work on hiding the complexity of the framework as well as improving the frontend and the instructions.

Outcomes of Delay of Care After the Onset of COVID-19 for Patients Managing Multiple Chronic Conditions

PURPOSE

Many patients delayed health care during COVID-19. We assessed the extent to which patients managing multiple chronic conditions (MCC) delayed care in the first months of the pandemic, reasons for delay, and impact of delay on patient-reported physical and behavioral health (BH) outcomes.

METHODS

As part of a large clinical trial conducted April 2016-June, 2021, primary care patients managing MCC were surveyed about physical and behavioral symptoms and functioning. Surveys administered between September 3, 2020, and March 16, 2021, included questions about the extent of and reasons for any delayed medical and BH care since COVID-19. Multivariable linear regression was used to assess health outcomes as a function of delay of care status.

RESULTS

Among patients who delayed medical care, 58% delayed more than once. Among those who delayed behavioral health care, 63% delayed more than once. Participants who delayed multiple times tended to be younger, female, unmarried, and reported food, financial, and housing insecurities and worse health. The primary reasons for delaying care were lack of availability of in-person visits and perceived lack of urgency. Participants who delayed care multiple times had significantly worse outcomes on nearly every measure of physical and mental health, compared with participants who delayed care once or did not delay.

CONCLUSIONS

Delay of care was substantial. Patients who delayed care multiple times were in poorer health and thus in need of more care. Effective strategies for reengaging patients in deferred care should be identified and implemented on multiple levels.

TRIAL REGISTRATION

ClinicalTrials.gov NCT02868983. Registered on August 16, 2016.

The role of models in the covid-19 pandemic

Mathematical and statistical models have played an important role in the analysis of data from COVID-19. They are important for tracking the progress of the pandemic, for understanding its spread in the population, and perhaps most significantly for forecasting the future course of the pandemic and evaluating potential policy options. This article describes the types of models that were used by research teams in Israel, presents their assumptions and basic elements, and illustrates how they were used, and how they influenced decisions. The article grew out of a “modelists’ dialog” organized by the Israel National Institute for Health Policy Research with participation from some of the leaders in the local modeling effort.

Growing Need for Primary Care Physicians Caring for Cancer Survivors

BACKGROUND

A rising population of cancer survivors is accompanied by a shortage of oncologists for continuity of care. This study examined the physicians who provided most of the care for cancer survivors, along with written information provided to the survivors before transfer of care.

METHODS

Data were collected through the CDC-sponsored Behavioral Risk Factor Surveillance System. Our analysis involved states whose respondents completed a cancer survivorship module from 2016 to 2020. Primary measures were the proportions of physician specialists who provided most of their subsequent health care and the proportions of survivors who received written summaries of their care and instructions.

RESULTS

The 36,737 cancer survivor respondents came from 33 states. Most of their health care came from primary care physicians [family physicians (42.3%, 95% CI: 41.3-43.2%) and general internists (26.0%, 95% CI: 25.2-26.9%)]. When seen by primary care physicians rather than subspecialists, a lower proportion of patients recalled receiving summaries of either their cancer treatments (44.3%, 95% CI: 42.5 to 46.2 vs 50.5%, 95% CI: 49.4 to 51.7%) or follow-up instructions (69.9%, 95% CI: 68.8 to 71.0% vs 78.7%, 95%CI 77.1 to 80.2%), regardless of their cancer type.

CONCLUSIONS

Regardless of their cancer type, two-thirds of survivors received most of their health care from primary care physicians. Collaborative community-based care within a shared decision-making framework is essential to prioritize and individualize patients' understandings and needs in this growing population.

Cancer Risk Score Prediction Based on a Single-Nucleotide Polymorphism Network

Objectives

Genome-wide association studies (GWAS) are performed to study the associations between genetic variants with respect to certain phenotypic traits such as cancer. However, the method that is commonly used in GWAS assumes that certain traits are solely affected by a single mutation. We propose a network analysis method, in which we generate association networks of single-nucleotide polymorphisms (SNPs) that can differentiate case and control groups. We hypothesize that certain phenotypic traits are attributable to mutations in groups of associated SNPs.

Methods

We propose a method based on a network analysis framework to study SNP-SNP interactions related to cancer incidence. We employed logistic regression to measure the significance of all SNP pairs from GWAS for the incidence of colorectal cancer and computed a cancer risk score based on the generated SNP networks.

Results

We demonstrated our method in a dataset from a case-control study of colorectal cancer in the South Sulawesi population. From the GWAS results, 20,094 pairs of 200 SNPs were created. We obtained one cluster containing four pairs of five SNPs that passed the filtering threshold based on their p-values. A locus on chromosome 12 (12:54410007) was found to be strongly connected to the four variants on chromosome 1. A polygenic risk score was computed from the five SNPs, and a significant difference in colorectal cancer risk was obtained between the case and control groups.

Conclusions

Our results demonstrate the applicability of our method to understand SNP-SNP interactions and compute risk scores for various types of cancer.

Towards Adaptability of Just-in-Time Adaptive Interventions

Just-in-time adaptive interventions (JITAIs) can promote behavior change in patients. It was the aim of our study to make JITAIs adaptable, i.e., to configure JITAIs for different purposes and to personalize them for different participants, whilst enabling central maintenance and integrated data analysis across deployments and individuals. We present a concept for adaptable JITAIs that was created following a design science approach. It builds on multi-level conceptual modeling and knowledge graphs and will be evaluated in user studies.

General Strategies for Glycan Microarray Data Processing and Analysis

Glycan microarrays provide a high-throughput technology for rapidly profiling interactions between carbohydrates and glycan-binding proteins (GBPs). Use of glycan microarrays involves several general steps, including construction of the microarray, carrying out the assay, detection of binding events, and analysis of the results. While multiple platforms have been developed to construct microarrays, most utilize fluorescence for detection of binding events. This chapter describes methods to acquire and process microarray images, including generating GAL files, imaging of the slide, aligning the grid, detecting problematic spots, and evaluating the quality of the data. The chapter focuses on processing our neoglycoprotein microarrays, but many of the lessons we have learned are applicable to other array formats.

Time series analysis and predicting COVID-19 affected patients by ARIMA model using machine learning

The spread of a respiratory syndrome known as Coronavirus Disease 2019 (COVID-19) quickly took on pandemic proportions, affecting over 192 countries. An emergency of the health system was obligated for the response to this epidemic. Although containment measures in China reduced new cases by more than 90 %, the levels of reduction were not the same in other countries. So, the question that arises is: what the world will see this pandemic, and how many patients can be affected? The response would be helpful and supportive of the authority and the community to prepare for the coming days. In this study, the Autoregressive Integrated Moving Average (ARIMA) model was employed to analyze the temporal dynamics of the worldwide spread of COVID-19 in the time window from January 22, 2020 to April 7, 2020. The cumulative number of confirmed Covid-19-affected patients forecasted over the three months was between 9,189,262 – 14,906,483 worldwide. This prediction value of Covid 19-affected patients will be valid only if the situation remains unchanged, and the epidemic spreads according to the previous nature worldwide in these three months.

Predicting Hospital Readmission in Heart Failure Patients in Iran: A Comparison of Various Machine Learning Methods

Objectives

Heart failure (HF) is a common disease with a high hospital readmission rate. This study considered class imbalance and missing data, which are two common issues in medical data. The current study’s main goal was to compare the performance of six machine learning (ML) methods for predicting hospital readmission in HF patients.

Methods

In this retrospective cohort study, information of 1,856 HF patients was analyzed. These patients were hospitalized in Farshchian Heart Center in Hamadan Province in Western Iran, from October 2015 to July 2019. The support vector machine (SVM), least-square SVM (LS-SVM), bagging, random forest (RF), AdaBoost, and naïve Bayes (NB) methods were used to predict hospital readmission. These methods’ performance was evaluated using sensitivity, specificity, positive predictive value, negative predictive value, and accuracy. Two imputation methods were also used to deal with missing data.

Results

Of the 1,856 HF patients, 29.9% had at least one hospital readmission. Among the ML methods, LS-SVM performed the worst, with accuracy in the range of 0.57–0.60, while RF performed the best, with the highest accuracy (range, 0.90–0.91). Other ML methods showed relatively good performance, with accuracy exceeding 0.84 in the test datasets. Furthermore, the performance of the SVM and LS-SVM methods in terms of accuracy was higher with the multiple imputation method than with the median imputation method.

Conclusions

This study showed that RF performed better, in terms of accuracy, than other methods for predicting hospital readmission in HF patients.

Estimating the Optimal Dexketoprofen Pharmaceutical Formulation with Machine Learning Methods and Statistical Approaches

Objectives

Orally disintegrating tablets (ODTs) can be utilized without any drinking water; this feature makes ODTs easy to use and suitable for specific groups of patients. Oral administration of drugs is the most commonly used route, and tablets constitute the most preferable pharmaceutical dosage form. However, the preparation of ODTs is costly and requires long trials, which creates obstacles for dosage trials. The aim of this study was to identify the most appropriate formulation using machine learning (ML) models of ODT dexketoprofen formulations, with the goal of providing a cost-effective and time-reducing solution.

Methods

This research utilized nonlinear regression models, including the k-nearest neighborhood (k-NN), support vector regression (SVR), classification and regression tree (CART), bootstrap aggregating (bagging), random forest (RF), gradient boosting machine (GBM), and extreme gradient boosting (XGBoost) methods, as well as the t-test, to predict the quantity of various components in the dexketoprofen formulation within fixed criteria.

Results

All the models were developed with Python libraries. The performance of the ML models was evaluated with R² values and the root mean square error. Hardness values of 0.99 and 2.88, friability values of 0.92 and 0.02, and disintegration time values of 0.97 and 10.09 using the GBM algorithm gave the best results.

Conclusions

In this study, we developed a computational approach to estimate the optimal pharmaceutical formulation of dexketoprofen. The results were evaluated by an expert, and it was found that they complied with Food and Drug Administration criteria.

Pipeline for Annual Averaged Wind Power Output Generation Prediction of Wind Turbines Based on Large Wind Speed Data Sets and Power Curve Data

In this article, an abstract framework for annual averaged wind power output generation prediction of wind turbines is presented which is heavily based on large wind speed data sets and power curve data of wind turbines due to the rising interest in wind energy as one main future renewable energy source. As combinations of arbitrary power curve modeling techniques and arbitrary wind speed distributions based on wind speed data are seldom combined, the abstract combination of these two aspects in wind power output generation prediction in one pipeline is thoroughly described here. Conclusively, one detailed example wind speed data set from a weather station situation in Bremen, Germany illustrates applicability of the presented framework.

Methodology for using a Bayesian nonparametric model to uncover universal patterns in color naming

Language is an integral part of society which enables communication among its members. To shed light on how words gain their meaning and how their meaning evolves over time, color naming is often used as a case study. The color domain can be defined by a physical space, making it a useful concept for studying denotation of meaning. Though humans can distinguish millions of colors, language provides us with a small, manageable set of terms for categorizing the space. Partitions of the color space vary across different language groups and evolve over time (e.g. new color terms may enter a language). Investigating universal patterns in color naming provides insight into the mechanisms that give rise to the observed data. Recently, computational techniques have been utilized to study this phenomenon. Here, we develop a methodology for transforming a color naming data set-namely, the World Color Survey-which is based on constraints imposed by the stimulus space. This transformed data is used to initialize a nonparametric Bayesian machine learning model in order to implement a culture and theory-independent study of universal color naming patterns across different language groups. All of the methods described are executed by our Python software package called ColorBBDP. • Data from the World Color Survey is transformed from its original format into binary features vectors which can be given as input to the Beta-Bernoulli Dirichlet Process Mixture Model. • This paper provides a specific application of Variational Inference on the Beta-Bernoulli Dirichlet Process Mixture Model towards a color naming data set. • New mathematical measures for performing post-cluster analyses are also detailed in this paper.

An Apple Watch Dashboard for HiGHmed Heart Insufficency Patients

Wearables are commercially available devices allowing continuous monitoring of users' health parameters. Their easy availability, increasing accuracy and functionality render them relevant for medical practice, specifically for longitudinal monitoring. There are clear benefits for the health care system, such as the opportunity of timely interventions by monitoring a patient during his daily life, resulting in a cost reduction in medical care and improved patient well-being. However, some tools are essential to enable the application of wearables in medical daily practice. For example, there is a need for software solutions that allow clinicians to quickly and easily analyze data from devices of their patients. The goal of this study was to develop a dashboard for physicians, which allows rapid data interpretation of longitudinal data from the Apple Watch. The prototype dashboard is an interactive web-based visualization platform utilizing Plotly. The dashboard displays the most important parameters like heart rate, steps per day, activity, exercise collected by the Apple Watch in a user-friendly and accessible way. Clear visualization makes it easy to identify trends or deviations in the data and see how these changes in daily behaviour affect patients' health. Our software is a key component to monitor patients with heart failure who participate in the HiGHmed use case cardiology project.

Current use of effect size or confidence interval analyses in clinical and biomedical research

The isolated use of the statistical hypothesis testing for two group comparison has limitations, and its combination with effect size or confidence interval analysis as complementary statistical tests is recommended. In the present work, we estimate the use of these complementary statistical tests (i.e. effect size or confidence interval) in recently published in research articles in clinical and biomedical areas. Methods: The ProQuest database was used to search published studies in academic journals between 2019 and 2020. The analysis was carried out using terms that represent five areas of clinical and biomedical research: "brain", "liver", "heart", "dental", and "covid-19". A total of 119,558 published articles were retrieved. Results: The relative use of complementary statistical tests in clinical and biomedical publications was low. The highest frequency usage of complementary statistical tests was among articles that also used statistical hypothesis testing for two-sample comparison. Publications with the term "covid-19" showed the lowest usage rate of complementary statistical tests when all article were analyzed but presented the highest rate among articles that used hypothesis testing. Conclusion: The low use of effect size or confidence interval in two-sample comparison suggests that coordinate measures should be taken in order to increase the use of this analysis in clinical and biomedical research. Their use should be emphasized in statistical disciplines for college and graduate students, become a routine procedure in research laboratories, and recommended by reviewers and editors of scientific journals.

Supplementary Information

The online version contains supplementary material available at 10.1007/s11192-021-04150-3.

Impact of the Choice of Cross-Validation Techniques on the Results of Machine Learning-Based Diagnostic Applications

Objective

With advances in data availability and computing capabilities, artificial intelligence and machine learning technologies have evolved rapidly in recent years. Researchers have taken advantage of these developments in healthcare informatics and created reliable tools to predict or classify diseases using machine learning-based algorithms. To correctly quantify the performance of those algorithms, the standard approach is to use cross-validation, where the algorithm is trained on a training set, and its performance is measured on a validation set. Both datasets should be subject-independent to simulate the expected behavior of a clinical study. This study compares two cross-validation strategies, the subject-wise and the record-wise techniques; the subject-wise strategy correctly mimics the process of a clinical study, while the record-wise strategy does not.

Methods

We started by creating a dataset of smartphone audio recordings of subjects diagnosed with and without Parkinson's disease. This dataset was then divided into training and holdout sets using subject-wise and the record-wise divisions. The training set was used to measure the performance of two classifiers (support vector machine and random forest) to compare six cross-validation techniques that simulated either the subject-wise process or the record-wise process. The holdout set was used to calculate the true error of the classifiers.

RESULTS

The record-wise division and the record-wise cross-validation techniques overestimated the performance of the classifiers and underestimated the classification error.

Conclusions

In a diagnostic scenario, the subject-wise technique is the proper way of estimating a model's performance, and record-wise techniques should be avoided.

Digitizing Data Management for Intraoperative Neuromonitoring

Intraoperative neurophysiological monitoring (IOM) enables a function-preserving surgical strategy for surgeries of brain or spinal cord pathologies by neurophysiological measurements. However, the IOM data management at neurosurgical institutions are often either not digitized or inefficient in terms of collecting, storing and processing of IOM data. Here, we describe the development of a web application, called IOM-Manager, as a first step towards the complete digitization of the IOM workflow. The web application is used for structured protocoling based on standardized protocol entry catalog, data archiving, and data analysis. These functionalities are based on the results of the requirement engineering of a process analysis, a survey with potential users and a market analysis. A usability test with one IOM team indicated the IOM-Manager and its other components can in fact solve many problems of existing solutions.

Optode Based Chemical Imaging-Possibilities, Challenges, and New Avenues in Multidimensional Optical Sensing

Seeing is believing, as the saying goes, and optical sensors (so-called optodes) are tools that can make chemistry visible. Optodes react reversibly and quickly (seconds to minutes) to changing analyte concentrations, enabling the spatial and temporal visualization of an analyte in complex environments. By being available as planar sensor foils or in the form of nano- or microparticles, optodes are flexible tools suitable for a wide array of applications. The steadily grown applications of in particular oxygen (O₂) and pH optodes in fields as diverse as medical, environmental, or material sciences is proof for the large demand of optode based chemical imaging. Nevertheless, the full potential of this technology is not exhausted yet, challenges have to be overcome, and new avenues wait to be taken. Within this Perspective, we look at where the field currently stands, highlight several successful examples of optode based chemical imaging and ask what it will take to advance current state-of-the-art technology. It is our intention to point toward some potential blind spots and to inspire further developments.

FhirExtinguisher: A FHIR Resource Flattening Tool Using FHIRPath

Data analysis with popular statistical toolchains like R usually needs to be performed on "flat tables" (so-called dataframes). However, data exchange is often done with FHIR, a format that is based on a hierarchical data model. In this paper, we want to present our tool FhirExtinguisher, which tackles the problems of loading FHIR data into statistical tools by extending the FHIRSearch API with an additional projection layer using FHIRPath. This projection layer can be used to select the data elements of interest and create a CSV file, which can be easily read as dataframe by almost any statistical toolchain.

Artificial Intelligence and Machine Learning to Predict Student Performance during the COVID-19

Artificial intelligence is based on algorithms that enable machines to make decisions instead of humans. This technology improves user experiences in a variety of areas. In this paper we discuss an intelligent solution to predict the performance of Moroccan students in the region of Guelmim Oued Noun through a recommendation system using artificial intelligence techniques during the COVID-19.

First Steps to Evaluate an NLP Tool's Medication Extraction Accuracy from Discharge Letters

INTRODUCTION

The aim of this study is to evaluate the use of a natural language processing (NLP) software to extract medication statements from unstructured medical discharge letters.

METHODS

Ten randomly selected discharge letters were extracted from the data warehouse of the University Hospital Erlangen (UHE) and manually annotated to create a gold standard. The AHD NLP tool, provided by MIRACUM's industry partner was used to annotate these discharge letters. Annotations by the NLP tool where then compared to the gold standard on two levels: phrase precision (whether or not the whole medication statement has been identified correctly) and token precision (whether or not the medication name has been identified correctly within correctly discovered medication phrases).

RESULTS

The NLP tool detected medication related phrases with an overall F-measure of 0.852. The medication name has been identified correctly with an overall F-measure of 0.936.

DISCUSSION

This proof-of-concept study is a first step towards an automated scalable evaluation system for MIRACUM's industry partner's NLP tool by using a gold standard. Medication phrases and names have been correctly identified in most cases by the NLP system. Future effort needs to be put into extending and validating the gold standard.

The Uptake and Factors Associated with Mastectomy among Chinese Women with Breast Cancer: A Retrospective Observational Study

OBJECTIVE

There are limited data concerning the use of mastectomy and associated factors in China in recent years. This study aimed to investigate the uptake of mastectomy and determine the associations between patients' characteristics and mastectomy among Chinese women with breast cancer.

METHODS

A retrospective analysis of female breast cancer cases from 1st January 2015 to 31st December 2019 from a tertiary hospital was conducted. Socio-demographic data, clinical data, and surgery types were collected by reviewing the medical record system. Chi-squared test, Fisher's exact test and multivariate logistic regression analysis were used to determine any correlations of patients' characteristics with mastectomy.

RESULTS

A total of 1,171 women with breast cancer were identified, and 76.60% of them underwent a mastectomy. The mastectomy rates showed an increase from 70.62% in 2015 to 86.87% in 2017 and then dropped to 71.91% in 2019. Women undergoing mastectomy were older and were more likely to be married and have at least one child. They had an advanced cancer stage, larger tumour size, and more lymph node invasion and were positive for HER-2 overexpression. Older age, larger tumour size (2-5 cm), higher cancer stages (stage 2- stage 3) and being positive for HER-2 were the four independent variables that significantly predicted the uptake of mastectomy.

CONCLUSIONS

Our results showed a wide application of mastectomy in China and uncovered the factors associated with mastectomy uptake from a single-centre experience. Findings suggested the potential overuse of mastectomy among women with early-stage breast cancer, and highlighted the significance of promoting cancer screening in China. Findings could be also used to develop relevant provisions and interventions to facilitate breast cancer treatment decision-making and screening planning.
.

Brain tissues have single-voxel signatures in multi-spectral MRI

Since the seminal works by Brodmann and contemporaries, it is well-known that different brain regions exhibit unique cytoarchitectonic and myeloarchitectonic features. Transferring the approach of classifying brain tissues - and other tissues - based on their intrinsic features to the realm of magnetic resonance (MR) is a longstanding endeavor. In the 1990s, atlas-based segmentation replaced earlier multi-spectral classification approaches because of the large overlap between the class distributions. Here, we explored the feasibility of performing global brain classification based on intrinsic MR features, and used several technological advances: ultra-high field MRI, q-space trajectory diffusion imaging revealing voxel-intrinsic diffusion properties, chemical exchange saturation transfer and semi-solid magnetization transfer imaging as a marker of myelination and neurochemistry, and current neural network architectures to analyze the data. In particular, we used the raw image data as well to increase the number of input features. We found that a global brain classification of roughly 97 brain regions was feasible with gross classification accuracy of 60%; and that mapping from voxel-intrinsic MR data to the brain region to which the data belongs is possible. This indicates the presence of unique MR signals of different brain regions, similar to their cytoarchitectonic and myeloarchitectonic fingerprints.

Opportunities and challenges of text mining in aterials research

Research publications are the major repository of scientific knowledge. However, their unstructured and highly heterogenous format creates a significant obstacle to large-scale analysis of the information contained within. Recent progress in natural language processing (NLP) has provided a variety of tools for high-quality information extraction from unstructured text. These tools are primarily trained on non-technical text and struggle to produce accurate results when applied to scientific text, involving specific technical terminology. During the last years, significant efforts in information retrieval have been made for biomedical and biochemical publications. For materials science, text mining (TM) methodology is still at the dawn of its development. In this review, we survey the recent progress in creating and applying TM and NLP approaches to materials science field. This review is directed at the broad class of researchers aiming to learn the fundamentals of TM as applied to the materials science publications.

Reporting Survey Based Studies - a Primer for Authors

The coronavirus disease 2019 (COVID-19) pandemic has led to a massive rise in survey-based research. The paucity of perspicuous guidelines for conducting surveys may pose a challenge to the conduct of ethical, valid and meticulous research. The aim of this paper is to guide authors aiming to publish in scholarly journals regarding the methods and means to carry out surveys for valid outcomes. The paper outlines the various aspects, from planning, execution and dissemination of surveys followed by the data analysis and choosing target journals. While providing a comprehensive understanding of the scenarios most conducive to carrying out a survey, the role of ethical approval, survey validation and pilot testing, this brief delves deeper into the survey designs, methods of dissemination, the ways to secure and maintain data anonymity, the various analytical approaches, the reporting techniques and the process of choosing the appropriate journal. Further, the authors analyze retracted survey-based studies and the reasons for the same. This review article intends to guide authors to improve the quality of survey-based research by describing the essential tools and means to do the same with the hope to improve the utility of such studies.

Timing Uncertainty in Collective Risk Dilemmas Encourages Group Reciprocation and Polarization

Social dilemmas are often shaped by actions involving uncertain returns only achievable in the future, such as climate action or voluntary vaccination. In this context, uncertainty may produce non-trivial effects. Here, we assess experimentally — through a collective risk dilemma — the effect of timing uncertainty, i.e. how uncertainty about when a target needs to be reached affects the participants' behaviors. We show that timing uncertainty prompts not only early generosity but also polarized outcomes, where participants' total contributions are distributed unevenly. Furthermore, analyzing participants' behavior under timing uncertainty reveals an increase in reciprocal strategies. A data-driven game-theoretical model captures the self-organizing dynamics underpinning these behavioral patterns. Timing uncertainty thus casts a shadow on the future that leads participants to respond early, whereas reciprocal strategies appear to be important for group success. Yet, the same uncertainty also leads to inequity and polarization, requiring the inclusion of new incentives handling these societal issues.

•

Timing uncertainty influences experimental observations in the collective risk game

•

It induces subjects to contribute earlier and in a polarized manner

•

Successful players adopt reciprocal strategies, responding in kind to past actions

•

Coordination gets more difficult under high timing uncertainty

The trend of prevalence of pain in Korea from 2005 to 2016

BACKGROUND

Korean society is afflicted with rapid aging. Aging is a risk factor for pain, and pain can reduce patients' quality of life. Thus, adequate management and monitoring of changing trends accompanying the demographic shift are highly valuable. However, this study was conducted because no studies have investigated the recent changes in the prevalence of pain.

METHODS

The extent of the prevalence of pain was determined by questions related to quality of life based on the data derived from the Korea National Health and Nutrition Survey (KNHNS) from 2005 to 2016. The annual frequencies of the pain group and severe pain group were calculated using the survey questionnaire. Multiple logistic regression analysis was performed to determine possible differences in prevalence by year.

RESULTS

The prevalence of pain in all populations was 30.6% in 2005 and 18.9% in 2016. The average prevalence from 2005 to 2016 was 21.9%. A declining trend occurred over time with an odds ratio of 0.929 per year (95% CI: 0.921-0.938). The prevalence of severe pain was 2.35% in 2005 and 1.88% in 2016. Likewise, a decrease was observed over time, with an odds ratio of 0.920 per year at 95% CI 0.901-0.939. The decline in age-/sex-stratified analysis also showed a statistically significant trend in all groups.

CONCLUSIONS

The prevalence of pain in Korean society, based on the KNHNS, has declined since 2005. Such a trend was observed in all ages and sexs, and was most significant in the elderly.

[Letter to the Editor about the Article "Is Medical Education Changing? Five Challenges for the Near Future"]

N/a.

Has the COVID-19 pandemic affected the susceptibility to cyberbullying in India?

Owing to the COVID-19 induced lockdown in India, most people’s internet activity surged, leading to an expected increase in the rate of cybercrimes. This research focuses on analyzing whether the factors significant in cyberbullying susceptibility changed with the lockdown. The study was conducted by surveying 256 students before the pandemic, in October 2019, and 118 students during the lockdown, in June 2020. This included questions about the respondents’ demographics, online presence, experience with offline bullying, perception of other’s opinions, and the instances of cyberbullying that apply to them. The results showed factors important in both timespans, namely (i) experience with offline bullying; (ii) individuals’ perceptiveness to others’ opinions; (iii) frequency of social media posts. Additionally, in the period before lockdown, factors namely (i) tendency to interact with strangers online; (ii) whether they’ve started a relationship online (iii) hours spent on social media; were found significant. Conversely, during the lockdown, additional distinct factors namely (i) being opinionated on public platforms; (ii) preference of Instagram; (iii) preferred gaming platform; (iv) number of games played; (v) sexual orientation; (vi) age were significant. With the change in variables in the two timespans, we can conclude that the pandemic has affected our susceptibility to cyberbullying.

Sphetcher: Spherical Thresholding Improves Sketching of Single-Cell Transcriptomic Heterogeneity

The massive size of single-cell RNA sequencing datasets often exceeds the capability of current computational analysis methods to solve routine tasks such as detection of cell types. Recently, geometric sketching was introduced as an alternative to uniform subsampling. It selects a subset of cells (the sketch) that evenly cover the transcriptomic space occupied by the original dataset, to accelerate downstream analyses and highlight rare cell types. Here, we propose algorithm Sphetcher that makes use of the thresholding technique to efficiently pick representative cells within spheres (as opposed to the typically used equal-sized boxes) that cover the entire transcriptomic space. We show that the spherical sketch computed by Sphetcher constitutes a more accurate representation of the original transcriptomic landscape. Our optimization scheme allows to include fairness aspects that can encode prior biological or experimental knowledge. We show how a fair sampling can inform the inference of the trajectory of human skeletal muscle myoblast differentiation.

•

Sphetcher distils large-scale scRNA-seq data down to a small selection of cells

•

Spheres of small radius around selected cells cover the original transcriptomic space

•

Selection enhances and accelerates downstream analysis such as trajectory inference

•

Sphetcher can leverage existing annotation of known cell types

Generation of Fine Grained Demographic Information for Epidemiological Analysis

Cancer risks may be influenced by local exposures such as working conditions or nuclear waste repositories. To find influences, local accumulations of cancer rates are used, for which finely granulated data should be utilized. In particular, high-resolution demographic data for a reference population are important, but often not available for data protection reasons. Therefore, estimation methods are necessary to approximate small-scale demographic data as accurately as possible. This paper presents an approach to project existing epidemiological and public data to a common granularity with respect to attribute characteristics such as place of residence, age or smoking status to allow for analyses such as local accumulations and consistently falls below an average relative error of 5%.

Is Medical Education Changing? Five Challenges for the Near Future

N/a.

Application of the 2017 KDIGO Guideline for the Evaluation and Care of Living Kidney Donors to Clinical Practice

The Kidney Disease: Improving Global Outcomes (KDIGO) 2017 "Clinical Practice Guideline on the Evaluation and Care of Living Kidney Donors" was developed to assist medical professionals who evaluate living kidney donor candidates and provide care before, during, and after donation. This guideline Work Group concluded that a comprehensive approach to donor candidate risk assessment should replace eligibility decisions on the basis of assessments of single risk factors in isolation. To address all issues important to living donors in a pragmatic and comprehensive guideline, many of the guideline recommendations were on the basis of expert consensus opinion even when no direct evidence was available. To advance available evidence, original data analyses were also undertaken to produce a "proof-of-concept" risk projection model for kidney failure. This was done to illustrate how the community can advance a new quantitative framework of risk that considers each candidate's profile of demographic and health characteristics. A public review by stakeholders and subject matter experts as well as industry and professional organizations informed the final formulation of the guideline. This review highlights the guideline framework, key concepts, and recommendations, and uses five patient scenarios and 12 guideline statements to illustrate how the guideline can be applied to support living donor evaluation and care in clinical practice.

Thyroid Hormone Use in the United States, 1997-2016

BACKGROUND

Thyroid disorders are among the most commonly treated conditions by the United States health care system. The number of patients reporting thyroid hormone use has increased in recent years, but it is unknown if there have been differential increases in the number of treated individuals within different demographic groups. Previous research has also not evaluated how expenditures for different thyroid hormone medications have changed in recent years.

METHODS

Using data from the 1997 through 2016 Medical Expenditure Panel Survey, we calculated the proportion of adults reporting thyroid hormone prescriptions by 3 demographic variables (age, sex, and race) and determined expenditures from thyroid hormone prescriptions by medication type (overall, generic, Synthroid or Cytomel, and other brand).

RESULTS

Between 1997 and 2016, the proportion of adults who reported thyroid hormone use increased from 4.1% (95% CI, 3.7-4.4) to 8.0% (95% CI, 7.5-8.5). Most of the growth in thyroid hormone use occurred among adults aged >65, and use was also more common among females and non-Hispanic whites. Expenditures from thyroid hormones increased from $1.1 billion (95% CI, 0.9-1.3) in 1997 to $3.2 billion dollars (95% CI, 2.9-3.6) in 2016. Generic thyroid hormone prescriptions comprised 18.1% of all thyroid hormone prescriptions in 2004 (95% CI, 15.8-20.4) and 80.8% of all thyroid hormone prescriptions (95% CI, 78.4-83.2) in 2016.

CONCLUSIONS

Thyroid hormone use nearly doubled over the last 20 years, and increased use was associated with being older, female, and non-Hispanic white. During the same time period, thyroid hormone expenditures almost tripled.

Stacking Ensemble Technique for Classifying Breast Cancer

Objectives

Breast cancer is the second most common cancer among Korean women. Because breast cancer is strongly associated with negative emotional and physical changes, early detection and treatment of breast cancer are very important. As a supporting tool for classifying breast cancer, we tried to identify the best meta-learner model in a stacking ensemble when the same machine learning models for the base learner and meta-learner are used.

Methods

We used machine learning models, such as the gradient boosted model, distributed random forest, generalized linear model, and deep neural network in a stacking ensemble. These models were used to construct a base learner, and each of them was used as a meta-learner again. Then, we compared the performance of machine learning models in the meta-learner to determine the best meta-learner model in the stacking ensemble.

Results

Experimental results showed that using the GBM as a meta-learner led to higher accuracy than that achieved with any other model for breast cancer data and using the GLM as a meta learner led to low root-mean-squared error for both sets of breast cancer data.

Conclusions

We compared the performance of every meta-learner model in a stacking ensemble as a supporting tool for classifying breast cancer. The study showed that using specific models as a metalearner resulted in better performance than single classifiers, and using GBM and GLM as a meta-learner is appropriate as a supporting tool for classifying breast cancer data.

Non-compartmental data analysis using SimBiology and MATLAB

MATLAB® is widely used for numerical analysis, modeling, and simulation. One of MATLAB's tools, SimBiology®, is often used for pharmacokinetic, pharmacodynamic model and dynamic systems; however, SimBiology seems to be rarely used for non-compartmental analysis (NCA), and the published official documentation provides a poor description of the analysis algorithm for NCA. Therefore, we conducted NCAs with a hypothetical dataset and some scenarios and compared the results. According to the results of this study, SimBiology estimates parameters using the unweighted linear regression for the terminal slope and linear interpolation method. Moreover, although the documentation describing the actual analysis algorithm used to process non-numeric data is not easily accessible to users, users may introduce numeric data at time zero to perform NCA properly. Using the command window, users can perform analyses more quickly and effectively. If the NCA official documentation were improved, SimBiology might be more widely adopted to perform NCA in clinical pharmacology.

Genomic Common Data Model for Biomedical Data in Clinical Practice

A common data model for clinical NGS panel data that is used in a distributed research network to achieve large scale to make evidence for improving patient care should be developed. This study developed OMOP-CDM extension for NGS panel data and confirmed the feasibility of the model by finding the differences between a database generated by research-purpose and clinical practice. We believe this data model can be used in distributed research model and will facilitate the usage of the clinical NGS data in patient care.

MedEx - Data Analytics for Medical Domain Experts in Real-Time

Translational research in the medical sector is dependent on clear communication between all participants. Visualization helps to represent data from different sources in a comprehensible way across disciplines. Existing tools for clinical data management are usually monolithic and technically challenging to set up, others require a transformation into specific data models while providing mostly non-interactive visualizations or being specialized to very particular use cases. Statistical programming languages (R, Julia) on the other hand offer great flexibility in data analytics, but are harder to access for clinicians with little to no programming expertise. Our software, the Medical Data Explorer (MedEx), aims to fill this gap as light-weight, intuitive, web-based solution with simple data import routes. We couple a modern dynamic web interface with an in-memory database solution for near real-time responsiveness. MedEx provides multiple visualization options (Scatterplot, correlation heatmap, bar chart, grouped boxplot, grouped histogram, coplot) to get an easy overview on the loaded data as well as to perform pattern discovery and elementary statistics. We demonstrate the utility of MedEx, by example, on data from the cardiology research warehouse of Heidelberg University Hospital. In summary, our tool empowers clinicians to conduct their own interactive exploratory data analysis.

Mapping the Hyperlink Structure of Diabetes Online Communities

Diabetes is one of the largest global health emergencies of the 21st century. As a chronic disease, diabetes requires continuous medical care and constant patient self-management. Such care involves several stakeholders to improve health outcome and patient quality of life. This paper makes use of World Wide Web network analysis to highlight how stakeholders, providing information about online diabetes communities, link to each other. To achieve this, we capture the network of diabetes related websites as a digital trace of a non-digital phenomenon. Furthermore, this helps us to understand the current situation of diabetes organizations from a digital perspective. The methodology involves state-of-the-art tools to crawl (Hyphe) and visualize (Gephi) topic-sensitive networks. While neither of these tools is new in itself, their combination provides a promising way to analyze chronic disease stakeholders, organizations and communities, representing a large proportion of the knowledge and support diabetes patients have access to nowadays.

Quality Improvement of Blood Drawing Through Targeted Training Using an Operation Support System

In order to reduce the rate of needle re-insertion (hereafter, retaking) during a blood drawing operation, we proposed a method for training blood drawing staff, based on data analysis. The method is composed of three steps: (1) analysis of collected data for selecting staff to be trained intensively, (2) training, and (3) assessment of training results. We obtained the result that the retaking rate was reduced for the trained staff.

Using Google Trends to understand information-seeking behaviour about throat cancer

OBJECTIVE

Many people seek health information from internet sources. Understanding this behaviour can help inform healthcare delivery. This study aimed to review Google Trends as a method for investigating internet-based information-seeking behaviour related to throat cancer in terms of quantity, content and thematic analysis.

METHOD

Data was collected using Google Trends. Normalised data was created using the search terms 'throat cancer', 'cancer', 'HPV', 'laryngeal cancer' and 'head and neck cancer'. The search data was used to analyse the temporal and geographical interest pattern of these terms from 2004 to 2015.

RESULTS

Three important peaks in searches for 'throat cancer' were identified. The first and greatest increase in interest was in September 2010, and there were also peaks in June 2013 and in October 2011.

CONCLUSION

Internet-search analysis can provide an insight into the information-seeking behaviour of the public. Mass media can hugely affect this information-seeking behaviour. Possessing tools to investigate and understand information-seeking behaviour may be used to improve healthcare delivery.

Automated assessment of biological database assertions using the scientific literature

Background

The large biological databases such as GenBank contain vast numbers of records, the content of which is substantively based on external resources, including published literature. Manual curation is used to establish whether the literature and the records are indeed consistent. We explore in this paper an automated method for assessing the consistency of biological assertions, to assist biocurators, which we call BARC, Biocuration tool for Assessment of Relation Consistency. In this method a biological assertion is represented as a relation between two objects (for example, a gene and a disease); we then use our novel set-based relevance algorithm SaBRA to retrieve pertinent literature, and apply a classifier to estimate the likelihood that this relation (assertion) is correct.

Results

Our experiments on assessing gene–disease relations and protein–protein interactions using the PubMed Central collection show that BARC can be effective at assisting curators to perform data cleansing. Specifically, the results obtained showed that BARC substantially outperforms the best baselines, with an improvement of F-measure of 3.5% and 13%, respectively, on gene-disease relations and protein-protein interactions. We have additionally carried out a feature analysis that showed that all feature types are informative, as are all fields of the documents.

Conclusions

BARC provides a clear benefit for the biocuration community, as there are no prior automated tools for identifying inconsistent assertions in large-scale biological databases.