Anonymized Dataset Research Articles

Can Nationwide Central Appointment Systems Reduce Waiting Times in Turkish Public Hospitals?

Health service delivery is an important component of a well-functioning health system. To achieve a competitive healthcare system, healthcare should be provided on time to avoid delays that could harm the patient. In this respect, patient wait time is an important indicator of health services delivery performance. Countries develop national or regional booking systems to manage patient wait times and healthcare delivery resources more efficiently. In this study; the effect of a national booking system on outpatient wait times in Turkish public healthcare settings was investigated. This study was conducted using anonymized national outpatient data sets for the year 2016 with permission from the Republic of Turkey Ministry of Health. The data was analyzed after pre-processing and transformation. As a result of the study, patient wait times were calculated, analyzed and evaluated according to province and facility care levels (secondary or tertiary). Results showed that in 2016, only 30.69% of outpatient visits had appointments and that the average waiting time for patients with appointments was 11.18% shorter than for patients without appointments. The results of this study fill a gap in the literature as almost no studies have investigated this issue due to a lack of administrative and clinical data and will provide important evidence for the improvement of health service delivery systems.

Responsible data sharing: Identifying and remedying possible re-identification of human participants.

Open data collected from research participants creates a tension between scholarly values of transparency and sharing, on the one hand, and privacy and security, on the other hand. A common solution is to make data sets anonymous by removing personally identifying information (e.g., names or worker IDs) before sharing. However, ostensibly anonymized data sets may be at risk of re-identification if they include demographic information. In the present article, we provide researchers with broadly applicable guidance and tangible tools so that they can engage in open science practices without jeopardizing participants' privacy. Specifically, we (a) review current privacy standards, (b) describe computer science data protection frameworks and their adaptability to the social sciences, (c) provide practical guidance for assessing and addressing re-identification risk, (d) introduce two open-source algorithms developed for psychological scientists-MinBlur and MinBlurLite-to increase privacy while maintaining the integrity of open data, and (e) highlight aspects of ethical data sharing that require further attention. Ultimately, the risk of re-identification should not dissuade engagement with open science practices. Instead, technical innovations should be developed and harnessed so that science can be as open as possible to promote transparency and sharing and as closed as necessary to maintain privacy and security. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Generative adversarial networks for anonymous acneic face dataset generation.

It is well known that the performance of any classification model is effective if the dataset used for the training process and the test process satisfy some specific requirements. In other words, the more the dataset size is large, balanced, and representative, the more one can trust the proposed model's effectiveness and, consequently, the obtained results. Unfortunately, large-size anonymous datasets are generally not publicly available in biomedical applications, especially those dealing with pathological human face images. This concern makes using deep-learning-based approaches challenging to deploy and difficult to reproduce or verify some published results. In this paper, we propose an efficient method to generate a realistic anonymous synthetic dataset of human faces, focusing on attributes related to acne disorders at three distinct levels of severity (Mild, Moderate, and Severe). Notably, our approach initiates from a small dataset of facial acne images, leveraging generative techniques to augment and diversify the dataset, ensuring comprehensive coverage of acne severity levels while maintaining anonymity and realism in the synthetic data. Therefore, a specific hierarchy StyleGAN-based algorithm trained at distinct levels is considered. Moreover, the utilization of generative adversarial networks for augmentation offers a means to circumvent potential privacy or legal concerns associated with acquiring medical datasets. This is attributed to the synthetic nature of the generated data, where no actual subjects are present, thereby ensuring compliance with privacy regulations and legal considerations. To evaluate the performance of the proposed scheme, we consider a CNN-based classification system, trained using the generated synthetic acneic face images and tested using authentic face images. Consequently, we show that an accuracy of 97.6% is achieved using InceptionResNetv2. As a result, this work allows the scientific community to employ the generated synthetic dataset for any data processing application without restrictions on legal or ethical concerns. Moreover, this approach can also be extended to other applications requiring the generation of synthetic medical images.

Machine learning evaluation for identification of M-proteins in human serum.

Serum electrophoresis (SPEP) is a method used to analyze the distribution of the most important proteins in the blood. The major clinical question is the presence of monoclonal fraction(s) of antibodies (M-protein/paraprotein), which is essential for the diagnosis and follow-up of hematological diseases, such as multiple myeloma. Recent studies have shown that machine learning can be used to assess protein electrophoresis by, for example, examining protein glycan patterns to follow up tumor surgery. In this study we compared 26 different decision tree algorithms to identify the presence of M-proteins in human serum by using numerical data from serum protein capillary electrophoresis. For the automated detection and clustering of data, we used an anonymized data set consisting of 67,073 samples. We found five methods with superior ability to detect M-proteins: Extra Trees (ET), Random Forest (RF), Histogram Grading Boosting Regressor (HGBR), Light Gradient Boosting Method (LGBM), and Extreme Gradient Boosting (XGB). Additionally, we implemented a game theoretic approach to disclose which features in the data set that were indicative of the resulting M-protein diagnosis. The results verified the gamma globulin fraction and part of the beta globulin fraction as the most important features of the electrophoresis analysis, thereby further strengthening the reliability of our approach. Finally, we tested the algorithms for classifying the M-protein isotypes, where ET and XGB showed the best performance out of the five algorithms tested. Our results show that serum capillary electrophoresis combined with decision tree algorithms have great potential in the application of rapid and accurate identification of M-proteins. Moreover, these methods would be applicable for a variety of blood analyses, such as hemoglobinopathies, indicating a wide-range diagnostic use. However, for M-protein isotype classification, combining machine learning solutions for numerical data from capillary electrophoresis with gel electrophoresis image data would be most advantageous.

Short-term trajectory prediction for individual metro passengers based on multi-level periodicity mining from semantic trajectory

Short-term trajectory prediction for individual metro passengers is an essential component of many location-based applications, such as bottleneck identification, anomalous trajectory detection, and emergency evacuation. Accurate and interpretable trajectory prediction in short-term remains challenging since the trajectory of individual metro passengers is influenced by time-varying and periodic mobility patterns. Moreover, it is still difficult for deep learning-based prediction approaches to intuitively explain the individual periodic mobility patterns captured by the black-box model. In this study, a long short-term memory method based on multi-level periodicity mining (LSTM-MP) from semantic trajectory is developed to achieve accurate and interpretable short-term trajectory prediction for individual metro passengers. The newly defined semantic trajectories of metro passengers that combine the semantic information about the geometric and operational characteristics of the metro with the raw trajectory data are constructed to help obtain fine-grained periodic mobility patterns of individual passengers. Multi-level periodicity which includes long-range, weekly-interval, and daily-interval periodicities, is calculated with variable time granularity and its polymerization from the constructed semantic trajectories. Finally, the characteristics of multi-level periodicity are considered by developing a principal trajectory attention mechanism to select the most relevant trajectories for prediction in the LSTM-MP framework. The prediction performance of LSTM-MP is compared with 2-order Markov chain and Naïve-LSTM algorithms using the large real-world anonymous passenger WiFi probe dataset. Experimental results show that LSTM-MP can predict when and where an individual will be in the future time series with an average accuracy of 82.3%. This demonstrates the effectiveness of LSTM-MP in the short-term trajectory prediction for individual metro passengers. Potential applications of the short-term trajectory prediction for individual metro passengers are also identified.

Applicability of artificial intelligence-based computer-aided detection (AI–CAD) for pulmonary tuberculosis to community-based active case finding

BackgroundArtificial intelligence-based computer-aided detection (AI–CAD) for tuberculosis (TB) has become commercially available and several studies have been conducted to evaluate the performance of AI–CAD for pulmonary tuberculosis (TB) in clinical settings. However, little is known about its applicability to community-based active case-finding (ACF) for TB.MethodsWe analysed an anonymized data set obtained from a community-based ACF in Cambodia, targeting persons aged 55 years or over, persons with any TB symptoms, such as chronic cough, and persons at risk of TB, including household contacts. All of the participants in the ACF were screened by chest radiography (CXR) by Cambodian doctors, followed by Xpert test when they were eligible for sputum examination. Interpretation by an experienced chest physician and abnormality scoring by a newly developed AI–CAD were retrospectively conducted for the CXR images. With a reference of Xpert-positive TB or human interpretations, receiver operating characteristic (ROC) curves were drawn to evaluate the AI–CAD performance by area under the ROC curve (AUROC). In addition, its applicability to community-based ACFs in Cambodia was examined.ResultsTB scores of the AI–CAD were significantly associated with the CXR classifications as indicated by the severity of TB disease, and its AUROC as the bacteriological reference was 0.86 (95% confidence interval 0.83–0.89). Using a threshold for triage purposes, the human reading and bacteriological examination needed fell to 21% and 15%, respectively, detecting 95% of Xpert-positive TB in ACF. For screening purposes, we could detect 98% of Xpert-positive TB cases.ConclusionsAI–CAD is applicable to community-based ACF in high TB burden settings, where experienced human readers for CXR images are scarce. The use of AI–CAD in developing countries has the potential to expand CXR screening in community-based ACFs, with a substantial decrease in the workload on human readers and laboratory labour. Further studies are needed to generalize the results to other countries by increasing the sample size and comparing the AI–CAD performance with that of more human readers.

Split-Boost Neural Networks

The calibration and training of a neural network is a complex and time-consuming procedure that requires significant computational resources to achieve satisfactory results. Key obstacles are a large number of hyperparameters to select and the onset of overfitting in the face of a small amount of data. In this framework, we propose an innovative training strategy for feed-forward architectures - called split-boost - that improves performance and automatically includes a regularizing behaviour without modeling it explicitly. Such a novel approach ultimately allows us to avoid explicitly modeling the regularization term, decreasing the total number of hyperparameters and speeding up the tuning phase. The proposed strategy is tested on a real-world (anonymized) dataset within a benchmark medical insurance design problem.

Effectiveness of patients’ involvement in a medical and nursing pain education programme: a protocol for an open-label randomised controlled trial including qualitative data

IntroductionPain is a multidimensional experience that varies among individuals and has a significant impact on their health. A biopsychosocial approach is recommended for effective pain management; however, health professionals’ education...

Influence of perioperative step volume on complication rate and length of hospital stay after colorectal cancer surgery (IPOS trial): study protocol for a randomised controlled single-centre trial at a German university hospital

BackgroundPerioperative mobilisation and physical activity are critical components of postoperative rehabilitation. Physical inactivity is a significant risk factor for complications and prolonged hospitalisation. However, specific recommendations for preoperative and postoperative...

A Neighborhood-Similarity-Based Imputation Algorithm for Healthcare Data Sets: A Comparative Study

The increasing computerisation of medical services has highlighted inconsistencies in the way in which patients’ historic medical data were recorded. Differences in process and practice between medical services and facilities have led to many incomplete and inaccurate medical histories being recorded. To create a single point of truth going forward, it is necessary to correct these inconsistencies. A common way to do this has been to use imputation techniques to predict missing data values based on the known values in the data set. In this paper, we propose a neighborhood similarity measure-based imputation technique and analyze its achieved prediction accuracy in comparison with a number of traditional imputation methods using both an incomplete anonymized diabetes medical data set and a number of simulated data sets as the sources of our data. The aim is to determine whether any improvement could be made in the accuracy of predicting a diabetes diagnosis using the known outcomes of the diabetes patients’ data set. The obtained results have proven the effectiveness of our proposed approach compared to other state-of-the-art single-pass imputation techniques.

Utilizing random forest algorithm for early detection of academic underperformance in open learning environments

Background One of the primary benefits of Open Learning Environments (OLEs) is their scalability. OLEs provide flexible and accessible learning opportunities to a large number of students, often on a global scale. This scalability has led to the development of OLEs that cover a wide range of subjects and disciplines, from computer science and engineering to humanities and social sciences. However, the scalability of OLEs also presents some challenges i.e., it can be too difficult to provide personalized support and feedback to individuals. Early prediction of student performance can improve the learning experience of students by providing early interventions and support. Method The specific objective of this study was to build a model that identifies at-risk students and allows for timely interventions to promote their academic achievement. The random forest classifier model has been used for analyzing anonymized large datasets available from Open University Learning Analytics (OULAD) to identify patterns and relationships among various factors that contribute to student success or failure. Results The findings of this study suggest that this algorithm achieved 90% accuracy in identifying students who may be at risk and providing them with the necessary support to succeed.

Comparisons of machine learning techniques for detecting fraudulent criminal identities

This paper focuses on applications of various machine learning techniques on an anonymized policing dataset used in EU SPIRIT Horizon 2020 project to identify fraudulent identities and help Law Enforcement Agencies (LEAs) in their investigation in finding potential criminals and identity resolution. Lack of qualitative data and appropriate methodology to carry out research on criminal fraudulent identities is a common reason for fewer research in this area. Additionally, it is a very sensitive data to work with and minor inaccuracy in prediction of result causes massive impact in the society as genuine people could be questioned whereas criminals could be sent free. Both of these issues are addressed in this paper by application of 39 million records from policing dataset and working towards higher accuracy while building the model. Various machine learning approaches are applied to train the dataset to make predictions and the research focus on being able to predict the 5 suspected fraudulent identities out of 39 million records in the policing dataset. One of the applied machine learning techniques include TensorFlow along with Keras model which has seldomly been applied by researchers in detection of criminal data. To compare the results and test accuracy of TensorFlow model, other machine learning techniques such as Support Vector Machine, Naïve Bayes and K-nearest Neighbours are also applied to have a comparative study on the obtained outcomes from each model. The goal of this research is to find fraudulent IDs amongst all the anonymized IDs in the criminal dataset using TensorFlow and three other machine learning models and select the most optimal model out of them. Since the model is comparing two names so string-matching techniques such as Levenshtein edit distance, Hamming Distance, Jaro-Winkler and Soundex were applied to select an effective approach first before building the model and analysing the results. TensorFlow model demonstrated highest accuracy with relatively least execution time and the only model to successfully predict all the 5 suspects from the policing dataset.

A machine learning framework to assess the impact of the COVID-19 pandemic on electricity consumption patterns

Energy self-sufficiency and resilience are essential elements for long-term urban sustainability. Energy supply in many cities in the United States has been highly dependent on fossil fuels and excessive energy consumption, both at home and outside (i.e., residential and non-residential energy consumption). In this work, we identify electricity consumption spatial trends and profiles across neighborhoods in the city of Chicago using a self-organizing map (SOM) machine learning method as a main tool. Toward this goal, we use an anonymous dataset provided by Commonwealth Edison (ComEd; Chicago's main electricity provider), which includes electricity consumption data of individual residential and non-residential accounts in Chicago. We aggregate electricity consumption to the zip codes level, apply machine learning clustering methods to categorize electricity consumption patterns, and finally couple it with land use variables to find spatial interrelations. The output demonstrates how specific profiles of electricity consumption emerge across zip codes, which is the first step for developing tailored policies to lower energy consumption. The result of this study can help engineers, urban planners, and policymakers develop actionable strategies for electricity consumption management.

Proximal hip fractures in 71,920 elderly patients: incidence, epidemiology, mortality and costs from a retrospective observational study

BackgroundThe risk of proximal femoral fractures increases with aging, causing significant morbidity, disability, mortality and socioeconomic pressure. The aims of the present work are (1) to investigate the epidemiology and incidence of these fractures among the elderly in the Region of Lombardy; (2) to identify the factors influencing survival; (3) to identify the factors influencing hospitalization and post-operative costs.MethodsThe Region of Lombardy provided anonymized datasets on hospitalized patients with a femoral neck fracture between 2011 and 2016, and anonymized datasets on extra-hospital treatments to track the patient history between 2008 and 2019. Statistical evaluations included descriptive statistics, survival analysis, Cox regression and multiple linear models.Results71,920 older adults suffered a femoral fracture in Lombardy between 2011 and 2016. 76.3% of patients were females and the median age was 84. The raw incidence of fractures was stable from year 2011 to year 2016, while the age-adjusted incidence diminished. Pertrochanteric fractures were more spread than transcervical fractures. In patients treated with surgery, receiving treatment within 48 h reduced the hazard of death within the next 24 months. Combined surgical procedures led to increased hazard in comparison with arthroplasty alone, while no differences were observed between different arthroplasties and reduction or fixation. In patients treated conservatively, age and male gender were associated with higher hazard of death. All patients considered, the type of surgery was the main factor determining primary hospitalization costs. A higher number of surgeries performed by the index hospital in the previous year was associated with financial savings. The early intervention significantly correlated with minor costs.ConclusionsThe number of proximal femoral fractures is increasing even if the age-adjusted incidence is decreasing. This is possibly due to prevention policies focused on the oldest cohort of the population. Two policies proved to be significantly beneficial in clinical and financial terms: the centralization of patients in high-volume hospitals and a time limit of 48 h from fracture to surgery.Trial registrationNon applicable.

Toward Global Consensus for MR-Guided Treatment Planning for Pancreatic Tumors on a 1.5 T MR-Linac

MR-guided SBRT with a 1.5 T MR-Linac is a relatively new therapy for pancreatic tumors with varying expertise levels. Moreover, treatment planning in the upper abdomen can be challenging as target coverage is often compromised by dosimetric constraints of abutting bowel structures. This may lead to large differences between centers in protocols, practices. To increase harmonization a worldwide consortium was founded among 1.5 T MR-Linac users. In this work we report on the outcome of the first phase within this collaboration, which is the assessment of the baseline variation between the treatment planning protocols and subsequent dose distributions. Twelve centers across three continents (North America, Europe, and Australia) participated in this consortium. Each center was sent the same two anonymized data sets reflecting two cases of locally advanced pancreatic cancer of different complexity levels. The data sets included a CT scan, a predefined structure set containing the gross target volume (GTV) and the OARs, a brief medical history, tumor motion characteristics, and auxiliary CT and MR imaging. Centers were asked to create an MRgRT treatment plan according to their clinical five-fraction SBRT protocol, using their institutional margin structures, beam setup, target prescriptions, and OAR constraints. Key DVH parameters that were evaluated are D99%, D90%, D50%, D1% for the GTV and D0.5cc for the duodenum, small bowel, and stomach. In general, large variations were observed in planning objectives and machine settings yielding widely varying inhomogeneous dose distributions to both the tumor and organs at risk (Table 1). This was especially manifest for case 2 where the tumor abutted with both the duodenum and small bowel over a trajectory of multiple centimeters. Not only were different trade-offs between target coverage and OAR sparing observed, but also different strategies for optimizing the integral dose to the tumor. These results indicate a large variety in the treatment planning strategies that could well translate to differences in outcome. Based on this first evaluation, the consortium will work towards a collective consensus protocol with a second evaluation round after internal discussions.

Wearable edge machine learning with synthetic photoplethysmograms

Strict privacy regulations pose challenges to the development of machine learning (ML) in the field of health technology where data is particularly sensitive. Gathering and using robust, bias-free, and suitably anonymized datasets required by ML models is difficult, time-consuming, and thus expensive. Parametric synthetic data offers a solution by mimicking real-world processes with easily adjustable parameters that shape the information content of the data as desired. This article presents a system demonstrating how synthetic data can be used in conjunction with wearable edge devices. Importantly, the system preserves privacy as there is no risk of leaking sensitive information from the model or during the use of the wearable device. The system consists of (1) a synthetic photoplethysmogram (PPG) model, (2) convolutional neural network (CNN) models trained with the synthetic signals, (3) a wearable edge device that computes heart rate from real-time PPG signals using the developed CNN models, and (4) an accompanying mobile phone application receiving the results. The synthetic model produces realistic PPG signals together with labels that can be used in CNN model training. The quality of the synthetic data is sufficient to train even a tiny CNN model with only two convolutional layers and 28 parameters to detect PPG waveform feet. The developed wearable device is able to run the model smoothly and the performance of the model is on par with the more complex models and other foot detection algorithms.

Flow-models 2.0: Elephant flows modeling and detection with machine learning

This article presents the new version of the flow-models IP network flow modeling framework. The improved features include flow skipping and counting, flow filtering, IP address anonymization, and time series data calculation. The new version also enables simulation of the first packet mirroring feature and provides tools for modeling the detection of elephant flows. It includes examples of using the scikit-learn library to build machine learning models for elephant flow detection based on the first packet. Furthermore, it provides an anonymized flow dataset, enabling researchers to train and validate machine learning models for traffic analysis in a reproducible and comparable manner.

Predictors and prognostic implications of hospital-acquired pneumonia in patients admitted for acute heart failure.

Data on predictors and prognosis of hospital acquired pneumonia (HAP) in patients admitted for acute heart failure (AHF) to intensive care units (ICU) are scarce. Better knowledge of these factors may inform management strategies. This study aimed to assess the incidence and predictors of HAP and its impact on management and outcomes in patients hospitalised for AHF in the ICU. this was a retrospective single-centre observational study. Patient-level and outcome data were collected from an anonymized registry-based dataset. Primary outcome was in-hospital all-cause mortality and secondary outcomes included length of stay (LOS), requirement for inotropic/ventilatory support, and prescription patterns of heart failure (HF) drug classes at discharge. Of 638 patients with AHF (mean age, 71.6 ± 12.7 years, 61.9% male), HAP occurred in 137 (21.5%). In multivariable analysis, HAP was predicted by de novo AHF, higher NT proB-type natriuretic peptide levels, pleural effusion on chest x-ray, mitral regurgitation, and a history of stroke, diabetes, and chronic kidney disease. Patients with HAP had a longer LOS, and a greater likelihood of requiring inotropes (adjusted odds ratio, OR, 2.31, 95% confidence interval, CI, 2.16-2.81; p < 0.001) or ventilatory support (adjusted OR 2.11, 95%CI, 1.76-2.79, p < 0.001). After adjusting for age, sex and clinical covariates, all-cause in-hospital mortality was significantly higher in patients with HAP (hazard ratio, 2.10; 95%CI, 1.71-2.84; p < 0.001). Patients recovering from HAP were less likely to receive HF medications at discharge. HAP is frequent in AHF patients in the ICU setting and more prevalent in individuals with de novo AHF, mitral regurgitation, higher burden of comorbidities, and more severe congestion. HAP confers a greater risk of complications and in-hospital mortality, and a lower likelihood of receiving evidence-based HF medications at discharge.

The Impression of Phones and Prosody Choice in the Gibberish Speech of the Virtual Embodied Conversational Agent Kotaro

The number of smart devices is expected to exceed 100 billion by 2050, and many will feature conversational user interfaces. Thus, methods for generating appropriate prosody for the responses of embodied conversational agents will be very important. This paper presents the results of the “Talk to Kotaro” experiment, which was conducted to better understand how people from different cultural backgrounds react when listening to prosody and phone choices for the IPA symbol-based gibberish speech of the virtual embodied conversational agent Kotaro. It also presents an analysis of the responses to a post-experiment Likert scale questionnaire and the emotions estimated from the participants’ facial expressions, which allowed one to obtain a phone embedding matrix and to conclude that there is no common cross-cultural baseline impression regarding different prosody parameters and that similarly sounding phones are not close in the embedding space. Finally, it also provides the obtained data in a fully anonymous data set.

Performance Evaluation of Financial Industry Related Expense Forecasting Using Various Regression Algorithms for Machine Learning

Insurance costs refer to the fees charged by insurance companies to customers to pay for possible risks and losses. Insurance costs are usually based on the personal information of the insured, such as age, gender, occupation, health status and so on. For insurance companies, it is very important to accurately predict insurance costs, because it is directly related to the company's profits and risk control capabilities. The purpose of using regression algorithm to predict insurance expenses is to make insurance companies evaluate customers' risks more accurately and make more reasonable insurance expenses, so as to better manage risks and improve the company's profitability. In addition, for individuals, knowing their own insurance cost forecast results will also help them make better decisions and choose the most suitable insurance products to protect themselves and their families.In order to improve the pricing accuracy and profit rate of insurance companies, this study uses regression algorithm to predict insurance costs. It uses real anonymous data sets, which contain information of the insured from different regions, different ages, different sexes and different smoking status. It uses the comparison algorithm function of regression algorithm, which contains dozens of algorithms and covers all regression algorithms and compare their prediction performance. Our data set takes into account various factors that affect the insurance cost, such as age, gender, body mass index, smoking status and so on. And add them to the model as independent variables. It uses cross-validation to evaluate the generalization ability of the model and R2 index to evaluate the prediction performance. The results show that GBR has the best prediction performance, with R2 of 87%. Our research provides an accurate method for insurance companies to predict insurance costs, which is helpful for insurance companies to formulate more reasonable pricing strategies and improve market competitiveness.