Informatics in Medicine Unlocked最新文献

Different pathologies of the hip are characterized by the abnormal shape of the bony structures of the joint, namely the femur and the acetabulum. Three-dimensional (3D) models of the hip can be used for diagnosis, biomechanical simulation, and planning of surgical treatments. These models can be generated by building 3D surfaces of the joint's structures segmented on magnetic resonance (MR) images. Deep learning can avoid time-consuming manual segmentations, but its performance depends on the amount and quality of the available training data. Data augmentation and transfer learning are two approaches used when there is only a limited number of datasets. In particular, data augmentation can be used to artificially increase the size and diversity of the training datasets, whereas transfer learning can be used to build the desired model on top of a model previously trained with similar data. This study investigates the effect of data augmentation and transfer learning on the performance of deep learning for the automatic segmentation of the femur and acetabulum on 3D MR images of patients diagnosed with femoroacetabular impingement. Transfer learning was applied starting from a model trained for the segmentation of the bony structures of the shoulder joint, which bears some resemblance to the hip joint. Our results suggest that data augmentation is more effective than transfer learning, yielding a Dice similarity coefficient compared to ground-truth manual segmentations of 0.84 and 0.89 for the acetabulum and femur, respectively, whereas the Dice coefficient was 0.78 and 0.88 for the model based on transfer learning. The Accuracy for the two anatomical regions was 0.95 and 0.97 when using data augmentation, and 0.87 and 0.96 when using transfer learning. Data augmentation can improve the performance of deep learning models by increasing the diversity of the training dataset and making the models more robust to noise and variations in image quality. The proposed segmentation model could be combined with radiomic analysis for the automatic evaluation of hip pathologies.

Background and aim

The aim of the study is to identify statistical patterns in patients with myocardial infarction (MI) during hospitalization that allow predicting the development of acute conditions (recurrent myocardial infarction, cardiac death).

Methods

We identified 3471 episodes of patients treated with a diagnosis acute MI in Almazov National Medical Research Centre. For modelling we selected episodes with acute MI with cardiac surgery operations. Classical machine learning models were chosen as forecasting models: decision trees and ensembles based on them, logistic regression and support vector machine.

Results

The important signs for predicting recurrent MI were the minimum values of hemoglobin, the echocardiography parameters end systolic volume and pulmonary regurgitation, and the minimum value of leukocyte level. Predictors of lethal outcome during hospitalization were advanced age, high values of leukocytes, low values of hemoglobin, high values of alanine aminotransferase.

Conclusion

The obtained results make it possible to predict the development of a lethal outcome and re-infarction based on simple parameters that are easily available in clinical practice.

The role of aldose reductase (ALR), the key enzyme of the polyol pathway, has been firmly established in hyperglycemia-induced diabetic complications. Therefore, the present study focused on the screening of phytochemicals reported in Coccinia grandis against ALR using in-silico methodologies encompassing molecular docking, pharmacokinetics, molecular dynamic simulation, free energy calculation (MMGBSA), and quantum mechanics. A comprehensive array of 101 compounds from C. grandis documented in IMPPAT database and different literatures have been selected in this study. These compounds were meticulously docked with the ALR (PDB ID: 1EL3), yielding docking scores spanning a range of −5.8 to −11.0 kcal/mol compared to the positive control epalrestat with a score of −7.9kcal/mol. Among them, four compounds have been emerged as the most promising ALR inhibitors: tiliroside, lukianol B, formononetin, and trachelogenin, with docking scores of −11.0, −10.7, −10.4, and −10.2, respectively. Importantly, these compounds exhibited notable stability throughout 100 ns dynamic simulations compared to the control drug, aligning with Lipinski's rule of 5, standard ADME properties, and evincing an absence of anomalous toxic effects. Therefore, these compounds hold great promise as leads to the development of potent ALR inhibitors; however, further studies are needed to warrant their uses in ameliorating diabetic complications.

Background

Electronic prescribing is vital in healthcare systems, providing an efficient alternative to manual prescriptions and addressing issues like errors in writing. This study evaluates Iran's Computerized Physician Order Entry (CPOE) system due to its significant role in the health system.

Method

Conducted as a cross-sectional case study in 2023, this research targeted physicians and outpatient unit users in three hospitals affiliated with Kashan University of Medical Sciences. User satisfaction was assessed using the QUIS Questionnaire for user interaction satisfaction and the System Usability Scale (SUS) for overall usability. Statistical analysis included descriptive statistics, independent-sample t-tests, one-way ANOVA, and SUS questionnaire calculation via SPSS software.

Result

The QUIS and SUS questionnaires revealed an overall user satisfaction range of 4.65 out of 9 for physicians and 5.73 out of 9 for outpatient unit users. The SUS questionnaire scored the CPOE system at 72 out of 100 for physicians and 76 out of 100 for outpatient unit users, indicating good usability.

Conclusion

Iran's CPOE system received positive feedback, emphasizing ease of use, learnability, control, stimulation, and flexibility to user needs. While the evaluation was generally positive, there are areas for improvement. Future versions should address user demands, incorporate human-computer interaction principles, and rectify identified shortcomings for enhanced competency. Authorities should prioritize user-centric updates in the continuous development of the Iranian CPOE system.

Motility and morphology are crucial factors in determining male fertility. The current gold standard defined by the World Health Organization (WHO) mandates that semen analysis be performed by trained technicians. Despite strict standardization and technical guidelines set by the WHO, variability in semen analysis results remains prevalent owing to human subjectivity. Computer-Aided Sperm Analysis presents a further challenge because of its poor agreement with human analysis. This study aimed to enhance the accuracy of automated semen analysis by introducing a new method for expressing sperm motion and investigating advanced deep neural network architectures to estimate motility and morphology. Initially, we extracted motion information from the VISEM dataset using our novel motion representation technique called MotionFlow, along with shape information. Consequently, motility and morphology neural networks were constructed to exploit transfer learning in other fields to improve performance. These networks ingested motion and shape features and made separate predictions for motility and morphology. The evaluation process utilized a K-Fold cross-validation scheme to determine the mean absolute error (MAE) and maintain objectivity throughout the analysis. The proposed method achieved a greater level of performance than the current methods by attaining MAE of 6.842% and 4.148% for motility and morphology estimation, respectively. The improvement accomplished by this research may pave the way toward a fully automated human sperm quality assessment.

In the realm of medicine, artificial intelligence (AI) has emerged as a transformative force, harnessing the power to convert raw data into meaningful insights. Rather than supplanting the discernment of physicians, AI serves as an unprecedented enabler, equipping them with unimaginable tools. Its far-reaching applications encompass drug discovery, disease diagnosis, prognosis, treatment optimization, and outcome prediction. This technological revolution owes much to the prowess of machine learning algorithms, which adeptly process multifaceted data. Consequently, AI is poised to become an integral pillar of digital health systems, shaping and bolstering the realm of personalized medicine. The current landscape is abuzz with AI’s exponential growth, fueling a surge of research ventures aimed at enhancing medical practices. By delving into the realm of precision medicine, this paper endeavors to scrutinize and evaluate recent advancements in healthcare pertaining to the utilization of machine learning (ML) and deep learning (DL) algorithms. This systematic review comprehensively encompasses previously published works, dissecting key concepts, innovations, significant contributions, and pivotal enabling techniques. Aspiring to equip readers with a profound understanding and invaluable insights, this paper proves indispensable to those dedicated to exploring the state-of-the-art and contributing to future literature in this domain.

Diagnostic systems of cardiac arrhythmias face early and accurate detection challenges due to the overlap of electrocardiogram (ECG) patterns. Additionally, these systems must manage a huge number of features. This paper proposes a new classifier Kullback-Leibler classifier (KLC) that combines feature optimization and probabilistic Kullback-Leibler (KL) divergence. Particle swarm optimization (PSO) is used for optimizing the features of ECG data, while KL divergence counts the variance between training and testing probability distributions. The proposed framework led the new classifier to distinguish normal and abnormal rhythms accurately. MIT-BIH Standard Arrhythmia Dataset (MIT-BIH) is used to test the validity of the proposed model. The experimental results show the proposed classifier achieves results in precision (86.67%), recall (86.67%), and F1_Score (86.5%).

Introduction

Patient safety reporting systems (PSRS) play a crucial role in hospitals by collecting patient safety data, primarily from nurses. Identifying the factors influencing nurses' safety reporting behaviors provides safety managers with insights to encourage reporting. This study aims to identify the key factors impacting nurses’ acceptance of PSRS.

Methods

This cross-sectional study, conducted in 2022, enrolled 249 nurses from 14 teaching and non-teaching hospitals in Tehran, Iran. A questionnaire was developed on the basis of the unified theory of acceptance and use of technology (UTAUT) constructs, encompassing actual use, behavioral intention to use, facilitation conditions, effort expectancy, performance expectancy, and social influence. Additional constructs such as perceived positive outcomes, perceived negative outcomes, management support, and trust were also included. Data analysis comprised linear regression and Partial Least Squares-Structural Equation Modeling (PLS-SEM). The reliability and validity of the measurement model were assessed by using metrics like Cronbach's alpha, composite reliability, Rho_A, average variance extracted, and Heterotrait-Monotrait Ratio of Correlations before calculating path coefficients, the coefficient of determination, effect size, and predictive relevance of influencing factors.

Results

The study indicated favorable attitudes among nurses toward PSRS. Significant relationships were observed between behavioral intention (β = 0.379) and facilitation conditions with actual use (β = 0.228). Additionally, effort expectancy (β = 0.101), management support (β = 0.268), and performance expectancy (β = 0.180) demonstrated significant associations with behavioral intention. The R² values for behavioral intention and actual use were 0.198 and 0.246, respectively.

Conclusion

Simplifying reporting systems to reduce nurses’ reporting burden, providing effective facilitation within hospitals, enhancing perceived benefits associated with reporting systems for nurses, and ensuring robust managerial support are pivotal strategies that can significantly boost the acceptance of PSRS among nursing staff.

Clinical laboratories have evolved with technological advancements through integrating various subsystems into Health Information Systems (HIS), particularly the Laboratory Information System (LIS). The LIS automates processes, manages results, and interfaces with healthcare information sources. Challenges include workflow inefficiencies and data interpretation issues. Despite increased data accessibility, managing clinical data across systems remains complex. Integrating laboratory machines into LIS is essential for optimizing healthcare delivery, requiring effective integration technologies. This study aims to synthesize the existing empirical studies on the utilization of integration technologies for Software-to-Software (S2S) communication in automating clinical laboratory processes. This study systematically examined integration technologies in LIS using PubMed and following PRISMA 2020 guidelines. The three-phase methodology included a scoping analysis, methodological analysis, and a gap analysis, focusing on S2S communication, interoperability frameworks, data standards, communication protocols, and challenges in LIS integration technologies. Analysis of 28 sample studies revealed a complex landscape in LIS integration shaped by end-users, care providers, and researchers. Clinical laboratories prioritize integration, focusing on patient data and sustainability. Standards like HL7 and FHIR ensure interoperability. Eleven methodologies highlight system development in Health Information Systems (HIS). Interoperability is a common objective, with 22 out of 28 studies achieving success. Challenges include limited generalizability, poor validation, and post-implementation modifications. Issues like security, data incompatibility, and evolving standards persist.

Background

Clustering on projected data is common in biomedical research analysis. Principal component analysis (PCA) is widely used for projection, focusing on data dispersion (variance), while clustering identifies data concentrations (neighborhood). These are conflicting aims. This study re-evaluates combinations of PCA and other projection methods with common clustering algorithms.

Methods

Six projection methods (PCA, ICA, isomap, MDS, t-SNE, UMAP) were combined with five clustering algorithms (k-means, k-medoids, single link, Ward's method, average link). Projections and clusterings were evaluated using a numerical criterion for evaluating clustering performance and a visual criterion based on plotting the projected data on a Voronoi tessellation plane with class-wise coloring. Nine artificial and five real biomedical datasets were analyzed.

Results

No combination consistently captured prior classifications in projections and clusters. Visual inspection proved essential. PCA was often but not always outperformed or equaled by neighborhood-based methods (UMAP, t-SNE) and manifold learning techniques (isomap).

Conclusions

The results dissaprove PCA as a standard projection method prior to clustering. Therefore, method selection should be data specific as a tailored approach to data projection and clustering in biomedical analysis. To aid this process, we propose a novel visualization technique that combines Voronoi tessellation with color coding.