Computer methods and programs in biomedicine update最新文献

Background

As more low and middle-income countries (LMICs) implement electronic health record systems (EHRs), informatics has become an important component of global health. OpenMRS is a popular open-source EHR that has been implemented in over 60 countries. As in high income countries, interoperability and research capabilities remain a challenge. The Observational Medical Outcomes Partnership (OMOP) is one of the most relevant common data models (CDM) to support EHR-based research and data sharing, but its adoption has been limited in LMICs. To address this gap, we developed an OpenMRS to OMOP extract, transform, and load (ETL) tool using Talend.

Methods

We built on existing documentation to develop a comprehensive concept map from OpenMRS to OMOP. The OMOP domains were reviewed for overlapping concepts in OpenMRS, and a core set of tables were selected for ETL development. Specific variables were then identified from OpenMRS tables which mapped to OMOP domain fields. Afterwards, the ETL tool was developed using MySQL Workbench, PostgreSQL, and Talend.

Results

Seven of 14 OMOP domains were selected for ETL pipeline development . The location, person, and provider domains required the least amount of Talend job components, which involved ≤2 tDBInputs, 1 tMap, and 1 tDBOutput. Care_site, observation_period, observation, and person_death all required additional Talend components to properly transform the respective data fields. It took 15 min to transform 9,932 OpenMRS observation records to OMOP.

Conclusions

It is feasible to develop a free, open-source ETL pipeline to transform clinical data in OpenMRS instances into OMOP. Processing large datasets is swift and scalable with potential for more improvement. Using this tool alongside OpenMRS can dramatically increase the potential for global health informatics collaborations and building local infrastructure and research capacity. Further testing and development will be required prior to widespread dissemination, along with appropriate documentation and training resources.

Background

Developing clinical thinking competence (CTC) is crucial for physicians, but effective methods for cultivation and evaluation are a significant challenge. Classroom teaching and paper-and-pencil tests are insufficient, and clinical field learning is difficult to implement, especially during the COVID-19 pandemic. Simulation learning is a useful alternative, but existing methods, e.g., OSCE, 3D AR/VR, and SimMan, have limitations in terms of time, space, and cost.

Objective

This study aims to present the design and development of an Otolaryngology Mobile Tele-education System (OMTS) to facilitate CTC learning, and to evaluate the system's usability with senior otolaryngology experts.

Methods

The OMTS system utilizes the convenience of mobile learning and the touch function of mobile devices to assist users (medical students or post-graduate physicians) in learning CTC remotely. Clinical knowledge and system functions in the OMTS system are defined by senior experts based on required CTC learning cases. Through simulated clinical case scenarios, users can engage in interactive clinical inquiry, practice required physical and laboratory examinations, make treatment decisions based on simulated responses, and understand and correct learning problems through a diagnostic report for effective learning. Usability testing of the OMTS system was evaluated by three senior otolaryngology experts using measurements of content validity, system usability, and mental workload during their available time and location.

Results

Statistical results of experts' evaluation showed that the OMTS system has good content validity, marginal-to-acceptable system usability, and moderate mental workload. Experts agreed that the system was efficient, professional, and usable for learning, although the practicality of the clinical inquiry and hands-on practice functions could be improved further.

Conclusions

Based on the OMTS system, users can efficiently hands-on practice and learn clinical cases in otolaryngology, and understand and correct their problems according to the diagnostic report. Therefore, the OMTS system can be expected to facilitate CTC learning according to experts’ evaluation.

Background

Nurses’ health literacy (HL) and ehealth literacy (eHL) knowledge and skills are crucial for patient care. There is evidence that skills and knowledge on how HL and eHL can facilitate the provision of care, is low among nurses. Especially in the care of older adults with an increased risk of falls and infections, or poor adherence to pharmacotherapy, nurses could increase patient safety by assessing and supporting older people’ HL and eHL. This study aims to present the findings of an e-Delphi survey, which was implemented within the framework of the development of a course focusing on the enhancement of HL and eHL assessment and intervention competencies for nurses.

Method

A modified e-Delphi study was conducted in five countries from September 2020 to January 2021. Initially, a 19-item questionnaire on HL and eHL skills and competencies was developed by the research team based on literature review. Twenty experts from five countries (Cyprus, Czech Republic, Greece, Lithuania, Poland) participated in two e-Delphi rounds. The research team met to reach consensus on the final version of the modules.

Results

Four modules were derived from the Delphi survey: 1) Introduction to HL and eHL 2) Communication skills in practice 3) eHealth challenges: Feasibility and readability issues, and 4) HL/eHL and patient safety.

Conclusions

Raising awareness on HL and eHL skills in nurses and nursing students is considered a priority, especially during the COVID-19 era. The common effort among five academic institutions to develop an HL and eHL course targeting nurses and nursing students is considered an important step towards this direction.

Generally, Melanoma, Merkel cell cancer, Squamous cell carcinoma, and Basal cell carcinoma, are the four major categories of skin cancers. In contrast to other cancer types, melanoma, a type of skin cancer, affects a lot of people. Early identification and prediction of this skin cancer can avoid the risk of spreading to another part of the body which can be treated and cured effectively. The advancing machine learning and deep learning approaches create an efficient computerized diagnosis system that can assist physicians to predict the disease in a much faster way, and enable the affected person to identify it skillfully. The existing models either rely on machine learning models which are limited to feature selection or deep learning-based methods that learn features from full images. The proposed hybrid pre-trained convolutional neural network and machine learning classifiers are used for feature extraction and classification. This kind of approach improves the model's accuracy. Here the hybrid VGG16 and XGBoost is used as feature extraction and as a classifier, this integration obtains maximum accuracy of 99.1%, which is higher accuracy compared to other works represented in the literature survey.

The pandemic of coronavirus disease (COVID-19) caused the world to face an existential health crisis. COVID-19 lesions segmentation from CT images is nowadays an essential step to assess the severity of the disease and the amount of damage to the lungs. Deep learning has brought about a breakthrough in medical image segmentation where U-Net is the most prominent deep network. However, in this study, we argue that its architecture still lacks in certain aspects. First, there is an incompatibility in the U-Net skip connection between the encoder and decoder features which adversely affects the final prediction. Second, it lacks capturing multiscale context information and ignores the contribution of all semantic information through the segmentation process. Therefore, we propose a model named MCA-Unet, a novel multiscale deep learning segmentation model, which proposes some modifications to improve upon the U-Net model. MCA-Unet is integrated with a multiscale context aggregation module which is constituted of two blocks; a context embedding block (CEB) and a cascaded dilated convolution block (CDCB). The CEB aims at reducing the semantic gap between the concatenated features along the U-Net skip connections, it enriches the low-level encoder features with rich semantics inherited from the subsequent higher-level features, to reduce the semantic gap between the low-processed encoder features and the highly-processed decoder features, thus ensuring effectual concatenation. The CDCB is integrated to address the variability in shape and size of the COVID-19 lesions, it captures global context information by gradually expanding the receptive field, then operates reversely to capture the small fine details that might be scattered by enlarging the receptive field. To validate the robustness of our model, we tested it on a publicly available dataset of 1705 axial CT images with different types of COVID-19 infection. Experimental results show that MCA-Unet has attained a remarkable gain in performance in comparison with the basic U-Net and its variant. It achieved high performance using different evaluation metrics showing 88.6% Dice similarity coefficient, 85.4% Jaccard index, and 93.5% F-score measure. This outperformance shows great potential to help physicians during their examination and improve the clinical workflow.

Introduction

Diabetes Mellitus (DM) is characterized by impaired ability to metabolize glucose for use in cells for energy, resulting in high blood sugar (hyperglycemia). DM impacted 463 million individuals worldwide in 2019, with over four million fatalities documented. Blood glucose levels (BGL) are usually measured, as standard protocols, through invasive procedures. Recently, Artificial Intelligence (AI) based techniques have demonstrated the potential to estimate BGL using data collected by non-invasive Wearable Devices (WDs), thereby, facilitating monitoring and management of diabetics. One of the key aspects of WDs with machine learning (ML) algorithms is to find specific data signatures, called Digital biomarkers, that can be used in classification or gaging the extent of the underlying condition. The use of such biomarkers to monitor glycemic events represents a major shift in technology for self-monitoring and developing digital biomarkers using non-invasive WDs. To do this, it is necessary to investigate the correlations between characteristics acquired from non-invasive WDs and indicators of glycemic health; furthermore, much work is needed to validate accuracy.

Research Design & Methods

The study aimed to investigate performance of AI models in estimating BGL among diabetic patients using non-invasive wearable devices data An open-source dataset was used which provided BGL readings, diabetic status (Diabetic or non-diabetic), heart rate, Blood oxygen level (SPO2), Diastolic Blood pressure, Systolic Blood Pressure, Body temperature, Sweating, and Shivering for 13 participants by age group taken from WDs. Our experimental design included Data Collection, Feature Engineering, ML model selection/development, and reporting evaluation of metrics.

Results

We were able to estimate with high accuracy (RMSE range: 0.099 to 0.197) the relationship between glycemic metrics and features that can be derived from non-invasive WDs when utilizing AI models.

Conclusion

We provide further evidence of the feasibility of using commercially available WDs for the purpose of BGL estimation amongst diabetics.

Background

Machine learning (ML) has demonstrated success in classifying patients’ diagnostic outcomes in free-text clinical notes. However, due to the machine learning model's complexity, interpreting the mechanism behind classification results remains difficult.

Methods

We investigated interpretable representations of text-based machine learning classification models. We created machine learning models to classify temporal bone fractures based on 164 temporal bone computed tomography (CT) text reports. We adopted the XGBoost, Support Vector Machine, Logistic Regression, and Random Forest algorithms. To interpret models, we used two major methodologies: (1) We calculated the average word frequency score (WFS) for keywords. The word frequency score shows the frequency gap between positively and negatively classified cases. (2) We used Local Interpretable Model-Agnostic Explanations (LIME) to show the word-level contribution to bone fracture classification.

Results

In temporal bone fracture classification, the random forest model achieved an average F1-score of 0.93. WFS revealed a difference in keyword usage between fracture and non-fracture cases. Additionally, LIME visualized the keywords' contributions to the classification results. The evaluation of LIME-based interpretation achieved the highest interpreting accuracy of 0.97.

Conclusion

The interpretable text explainer can improve physicians' understanding of machine learning predictions. By providing simple visualization, our model can increase the trust of computerized models. Our model supports more transparent computerized decision-making in clinical settings.

This paper presents a convolutional neural network (CNN) based automated noninvasive voice pathology detection system. The proposed system functions in two steps. First, it discriminates pathological voices from healthy ones, and then, it classifies the discriminated pathological voices into one of the three pathologies. Two CNNs are used for these purposes; one works as a binary classifier to identify pathological voices. The other one works as a multiclass classifier for categorizing the voice pathologies. This work investigates the effectiveness of electroglottographic (EGG) and speech signals to detect and classify pathological voices using sustained vowel ('/a/') samples. EGG signals can assess the vibratory pattern of the vocal folds during voiced sound. On the other hand, the speech signals add spectral color to the EGG signals. Hence, their contributions for pathology identification and segregation differ, as demonstrated in this work. The Saarbrücken Voice Database (SVD) is used in this investigation. The results show that the proposed system achieves a higher accuracy (more than 9%) in identifying pathological voices from healthy ones with speech signals than EGG signals. However, categorizing pathological voices into different pathology types demonstrates higher accuracy (more than 12%) with EGG signals than speech signals. A comparative performance analysis of the proposed system is presented with these two signals in terms of clinical and statistical measures. The obtained results of this work are also compared with those of other related published works.