ALBERT

Description: Background Occlusions of intravenous (IV) tubing can prevent vital and time-critical medication or solutions from being delivered into the bloodstream of patients receiving IV therapy. At low flow rates (≤ 1 ml/h) the alarm delay (time to an alert to the user) can be up to 2 h using conventional pressure threshold algorithms. In order to reduce alarm delays we developed and evaluated the performance of two new real-time occlusion detection algorithms and one co-occlusion detector that determines the correlation in trends in pressure changes for multiple pumps. Methods Bench-tested experimental runs were recorded in triplicate at rates of 1, 2, 4, 8, 16, and 32 ml/h. Each run consisted of 10 min of non-occluded infusion followed by a period of occluded infusion of 10 min or until a conventional occlusion alarm at 400 mmHg occurred. The first algorithm based on binary logistic regression attempts to detect occlusions based on the pump’s administration rate Q(t) and pressure sensor readings P(t). The second algorithm continuously monitored whether the actual variation in the pressure exceeded a threshold of 2 standard deviations (SD) above the baseline pressure. When a pump detected an occlusion using the SD algorithm, a third algorithm correlated the pressures of multiple pumps to detect the presence of a shared occlusion. The algorithms were evaluated using 6 bench-tested baseline single-pump occlusion scenarios, 9 single-pump validation scenarios and 7 multi-pump co-occlusion scenarios (i.e. with flow rates of 1 + 1, 1 + 2, 1 + 4, 1 + 8, 1 + 16, and 1 + 32 ml/h respectively). Alarm delay was the primary performance measure. Results In the baseline single-pump occlusion scenarios, the overall mean ± SD alarm delay of the regression and SD algorithms were 1.8 ± 0.8 min and 0.4 ± 0.2 min, respectively. Compared to the delay of the conventional alarm this corresponds to a mean time reduction of 76% (P = 0.003) and 95% (P = 0.001), respectively. In the validation scenarios the overall mean ± SD alarm delay of the regression and SD algorithms were respectively 1.8 ± 1.6 min and 0.3 ± 0.2 min, corresponding to a mean time reduction of 77% and 95%. In the multi-pump scenarios a correlation 〉 0.8 between multiple pump pressures after initial occlusion detection by the SD algorithm had a mean ± SD alarm delay of 0.4 ± 0.2 min. In 2 out of the 9 validation scenarios an occlusion was not detected by the regression algorithm before a conventional occlusion alarm occurred. Otherwise no occlusions were missed. Conclusions In single pumps, both the regression and SD algorithm considerably reduced alarm delay compared to conventional pressure limit-based detection. The SD algorithm appeared to be more robust than the regression algorithm. For multiple pumps the correlation algorithm reliably detected co-occlusions. The latter may be used to localize the segment of tubing in which the occlusion occurs. Trial registration Not applicable.

Description: Machine learning and artificial intelligence have entered biomedical decision-making for diagnostics, prognostics, or therapy recommendations. However, these methods need to be interpreted with care because of the severe consequences for patients. In contrast to human decision-making, computational models typically make a decision also with low confidence. Machine learning with abstention better reflects human decision-making by introducing a reject option for samples with low confidence. The abstention intervals are typically symmetric intervals around the decision boundary. In the current study, we use asymmetric abstention intervals, which we demonstrate to be better suited for biomedical data that is typically highly imbalanced. We evaluate symmetric and asymmetric abstention on three real-world biomedical datasets and show that both approaches can significantly improve classification performance. However, asymmetric abstention rejects as many or fewer samples compared to symmetric abstention and thus, should be used in imbalanced data.

Description: Background Health systems globally are investing in integrating secure messaging platforms for virtual care in clinical practice. Implementation science is essential for adoption, scale-up, spread and maintenance of complex evidence-based solutions in clinics with evolving priorities. In response, the mobile Health (mHealth) Research Group modified the existing consolidated framework for implementation research (CFIR) to evaluate implementation of virtual health tools in clinical settings. WelTel® is an evidence-based digital health platform widely deployed in various geographical and health contexts. The objective is to identify the facilitators and barriers for implementing WelTel and to assess the application of the mCFIR tool in facilitating focus groups in different geographical and health settings. Methods Both qualitative and descriptive quantitative approaches were employed. Six mCFIR sessions were held in three countries with 51 key stakeholders. The mCFIR tool consists of 5 Domains and 25 constructs and was distributed through Qualtrics Experience Management (XM). “Performance” and “Importance” scores were valued on a scale of 0 to 10 (Mean ± SD). Descriptive analysis was conducted using R computing software. NVivo 12 Pro software was used to analyze mCFIR responses and to generate themes from the participants’ input. Results We observed a parallel trend in the scores of Importance and Performance. Of the five Domains, Domain 4 (End-user Characteristics) and Domain 3 (Inner Settings) scored highest in Importance (8.9 ± 0.5 and 8.6 ± 0.6, respectively) and Performance (7.6 ± 0.7 and 7.2 ± 1.3, respectively) for all sites. Domain 2 (Outer Setting) scored the lowest in both Importance and Performance for all sites (7.6 ± 0.4 and 5.6 ± 1.8). The thematic analysis produced the following themes: for areas of strengths, the themes brought up were timely diagnosis and response, cost-effectiveness, and user-friendliness. As for areas for improvement, the themes discussed were training, phone accessibility, stakeholder engagement, and literacy. Conclusion The mCFIR tool allowed for a comprehensive understanding of the barriers and facilitators to the implementation, reach, and scale-up of digital health tools. Amongst several important findings, we observed the value of bringing the perspectives of both end users (HCPs and patients) to the table across Domains. Trial Registration: NCT02603536 – November 11, 2015: WelTelOAKTREE: Text Messaging to Support Patients With HIV/AIDS in British Columbia (WelTelOAKTREE). NCT01549457 – March 9, 2012: TB mHealth Study—Use of Cell Phones to Improve Compliance in Patients on LTBI Treatment.

Description: Background One of the important components of the health system is the emergency medical services (EMS) system. The EMS system was implemented at Kerman University of Medical Sciences teaching hospitals to communicate the situation of patients being transferred to the hospital by EMS and to provide facilities tailored to the patient's condition. The objective of this study was to investigate the impact of the EMS system on the patient care process and the workflow of users. Methods The hospital information system (HIS) report was used to investigate the impact of the EMS system on the patient care process and a questionnaire was distributed among 244 participants to determine its impact on the workflow of the users. Mann–Whitney U was used to analyze HIS reports, and Chi-square was used to analyze the data collected by questionnaires. Results The EMS system reduced the patient's stay in hospital by an average of 3 h and 45 min. It also increased the number of patients' discharge from the emergency room to 2.2% and reduced the death rate by 1.3% (p 

Description: Background Undernutrition is the main cause of child death in developing countries. This paper aimed to explore the efficacy of machine learning (ML) approaches in predicting under-five undernutrition in Ethiopian administrative zones and to identify the most important predictors. Method The study employed ML techniques using retrospective cross-sectional survey data from Ethiopia, a national-representative data collected in the year (2000, 2005, 2011, and 2016). We explored six commonly used ML algorithms; Logistic regression, Least Absolute Shrinkage and Selection Operator (L-1 regularization logistic regression), L-2 regularization (Ridge), Elastic net, neural network, and random forest (RF). Sensitivity, specificity, accuracy, and area under the curve were used to evaluate the performance of those models. Results Based on different performance evaluations, the RF algorithm was selected as the best ML model. In the order of importance; urban–rural settlement, literacy rate of parents, and place of residence were the major determinants of disparities of nutritional status for under-five children among Ethiopian administrative zones. Conclusion Our results showed that the considered machine learning classification algorithms can effectively predict the under-five undernutrition status in Ethiopian administrative zones. Persistent under-five undernutrition status was found in the northern part of Ethiopia. The identification of such high-risk zones could provide useful information to decision-makers trying to reduce child undernutrition.

Description: Purpose Some predictive systems using machine learning models have been developed to predict sepsis; however, they were mostly built with a low percent of missing values, which does not correspond with the actual clinical situation. In this study, we developed a machine learning model with a high rate of missing and erroneous data to enable prediction under missing, noisy, and erroneous inputs, as in the actual clinical situation. Materials and methods The proposed artificial neural network model was implemented using the MATLAB ANN toolbox, based on stochastic gradient descent. The dataset was collected over the past decade with approval from the appropriate institutional review boards, and the sepsis status was identified and labeled using Sepsis-3 clinical criteria. The imputation method was built by last observation carried forward and mean value, aimed to simulate clinical situation. Results The mean area under the receiver operating characteristic (ROC) curve (AUC) of classifying sepsis and nonsepsis patients was 0.82 and 0.786 at 0 h and 40 h prior to onset, respectively. The highest model performance was found for one-hourly data, demonstrating that our ANN model can perform adequately with limited hourly data provided. Conclusions Our model has the moderate ability to predict sepsis up to 40 h in advance under simulated clinical situation with real-world data.

Description: Background Despite the improvements in the knowledge and understanding of the role of health information in the global health system, the quality of data generated by a routine health information system is still very poor in low and middle-income countries. There is a paucity of studies as to what determines data quality in health facilities in the study area. Therefore, this study was aimed to assess the quality of routine health information system data and associated factors in public health facilities of Harari region, Ethiopia. Methods A cross-sectional study was conducted in all public health facilities in the Harari region of Ethiopia. The department-level data were collected from respective department heads through document reviews, interviews, and observation checklists. Descriptive statistics were used to data quality and multivariate logistic regression was run to identify factors influencing data quality. The level of significance was declared at P value 

Description: Background Strengthening surveillance systems to collect near-real-time case-based data plays a fundamental role in achieving malaria elimination in the Greater Mekong Subregion (GMS). With the advanced and widespread use of digital technology, mHealth is increasingly taking a prominent role in malaria surveillance systems in GMS countries, including Myanmar. In Myanmar’s malaria elimination program, an mHealth system called Malaria Case-based Reporting (MCBR) has been applied for case-based reporting of malaria data by integrated community malaria volunteers (ICMVs). However, the sustainability of such mHealth systems in the context of existing malaria elimination programs in Myanmar is unknown. Methods Focus group discussions were conducted with ICMVs and semi-structured in-depth interviews were conducted with malaria program stakeholders from Myanmar’s Ministry of Health and Sports and its malaria program implementing partners. Thematic (deductive followed by inductive) analysis was undertaken using a qualitative descriptive approach. Results Technological and financial constraints such as inadequate internet access, software errors, and insufficient financial resources to support mobile phone-related costs have hampered users’ access to MCBR. Poor system integrity, unpredictable reporting outcomes, inadequate human resources for system management, and inefficient user support undermined the perceived quality of the system and user satisfaction, and hence its sustainability. Furthermore, multiple parallel systems with functions overlapping those of MCBR were in use. Conclusions Despite its effectiveness and efficiency in malaria surveillance, the sustainability of nationwide implementation of MCBR is uncertain. To make it sustainable, stakeholders should deploy a dedicated human workforce with the necessary technical and technological capacities; secure sustainable, long-term funding for implementation of MCBR; find an alternative cost-effective plan for ensuring sustainable system access by ICMVs, such as using volunteer-owned mobile phones for reporting rather than supporting new mobile phones to them; and find a solution to the burden of multiple parallel systems. Trial registration Not applicable.

Description: Charging according to disease is an important way to effectively promote the reform of medical insurance mechanism, reasonably allocate medical resources and reduce the burden of patients, and it is also an important direction of medical development at home and abroad. The cost forecast of single disease can not only find the potential influence and driving factors, but also estimate the active cost, and tell the management and reasonable allocation of medical resources. In this paper, a method of Bayesian network combined with regression analysis is proposed to predict the cost of treatment based on the patient's electronic medical record when the amount of data is small. Firstly, a set of text-based medical record data conversion method is established, and in the clustering method, the missing value interpolation is carried out by weighted method according to the distance, which completes the data preparation and processing for the realization of data prediction. Then, aiming at the problem of low prediction accuracy of traditional regression model, this paper establishes a prediction model combined with local weight regression method after Bayesian network interpretation and classification of patients' treatment process. Finally, the model is verified with the medical record data provided by the hospital, and the results show that the model has higher prediction accuracy.

Description: Background The liver is an important organ that undertakes the metabolic function of the human body. Liver cancer has become one of the cancers with the highest mortality. In clinic, it is an important work to extract the liver region accurately before the diagnosis and treatment of liver lesions. However, manual liver segmentation is a time-consuming and boring process. Not only that, but the segmentation results usually varies from person to person due to different work experience. In order to assist in clinical automatic liver segmentation, this paper proposes a U-shaped network with multi-scale attention mechanism for liver organ segmentation in CT images, which is called MSA-UNet. Our method makes a new design of U-Net encoder, decoder, skip connection, and context transition structure. These structures greatly enhance the feature extraction ability of encoder and the efficiency of decoder to recover spatial location information. We have designed many experiments on publicly available datasets to show the effectiveness of MSA-UNet. Compared with some other advanced segmentation methods, MSA-UNet finally achieved the best segmentation effect, reaching 98.00% dice similarity coefficient (DSC) and 96.08% intersection over union (IOU).