Search

Accurate delineation of key waveforms in an ECG is a critical step in extracting relevant features to support the diagnosis and treatment of heart conditions. Although deep learning based methods using segmentation models to locate P, QRS, and T waves have shown promising results, their ability to handle arrhythmias has not been studied in any detail. In this paper we investigate the effect of arrhythmias on delineation quality and develop strategies to improve performance in such cases. We introduce a U-Net-like segmentation model for ECG delineation with a particular focus on diverse arrhythmias. This is followed by a post-processing algorithm which removes noise and automatically determines the boundaries of P, QRS, and T waves. Our model has been trained on a diverse dataset and evaluated against the LUDB and QTDB datasets to show strong performance, with F1-scores exceeding 99% for QRS and T waves, and over 97% for P waves in the LUDB dataset. Furthermore, we assess various models across a wide array of arrhythmias and observe that models with a strong performance on standard benchmarks may still perform poorly on arrhythmias that are underrepresented in these benchmarks, such as tachycardias. We propose solutions to address this discrepancy. [ABSTRACT FROM AUTHOR]

Malaria rapid diagnostic test (mRDT) kit is one of the techniques for diagnosing malaria. Due to its inherent advantages over the microscopy technique, several brands of the kit have flooded malaria endemic countries, without prior in-country evaluation. Two of such mRDT kits are Oscar (India) and Standard Q (Korea Republic). In this study, the performance of Oscar and Standard Q mRDT kits were compared to First Response (India) and CareStart (USA) mRDTs, which have been evaluated and deployed for use approved by the Ministry of Health (MOH). In this comparative study, whole blood samples were collected from patients suspected of malaria. Plasmodium falciparum was detected in each sample using nested polymerase chain reaction (nPCR), microscopy and the four mRDTs. The sensitivities, specificities, accuracies, positive and negative predictive values and accuracies of the mRDTs were determined using nPCR as a reference technique. Kappa statistic was used to determine the level of agreement among the techniques. Two hundred (200) blood samples were analyzed in this study. The overall detection rates of P. falciparum by microscopy, First Response, CareStart, Oscar-PfHRP2, Standard Q mRDT kits and nPCR were 31.5%, 34.5%, 33.5%, 32%, 31% and 43% (x2 = 6.1, p = 0.046), respectively. The accuracies of CareStart and First Response were comparable (90.5% vs. 89.5%). Further, comparing their sensitivities, Oscar-PfHRP2 was 74.4% (95% confidence interval (CI): 63.9–83.2) while that of Standard Q was 72.1% (95% CI: 61.4–81.2), with comparable accuracies (Oscar-PfHRP2–89% and Standard Q -88%). Apart from First Response that was 98.3% specific, the others were 100% specific. Kappa test revealed perfect diagnostic agreement (κ = 0.90–0.98) among the four mRDTs. That notwithstanding, Oscar-PfHRP2 agreed better with CareStart (κ = 0.94) and First Response (κ = 0.92) compared to the agreement between Standard Q and, CareStart (κ = 0.92) and First Response (κ = 0.90). Taken together, the diagnostic performance of the four mRDT kits were statistically similar. That notwithstanding, new mRDT kits should be evaluated prior to deployment for use. [ABSTRACT FROM AUTHOR]

We provide new algorithms for two tasks relating to heterogeneous tabular datasets: clustering, and synthetic data generation. Tabular datasets typically consist of heterogeneous data types (numerical, ordinal, categorical) in columns, but may also have hidden cluster structure in their rows: for example, they may be drawn from heterogeneous (geographical, socioeconomic, methodological) sources, such that the outcome variable they describe (such as the presence of a disease) may depend not only on the other variables but on the cluster context. Moreover, sharing of biomedical data is often hindered by patient confidentiality laws, and there is current interest in algorithms to generate synthetic tabular data from real data, for example via deep learning. We demonstrate a novel EM-based clustering algorithm, MMM ("Madras Mixture Model"), that outperforms standard algorithms in determining clusters in synthetic heterogeneous data, and recovers structure in real data. Based on this, we demonstrate a synthetic tabular data generation algorithm, MMMsynth, that pre-clusters the input data, and generates cluster-wise synthetic data assuming cluster-specific data distributions for the input columns. We benchmark this algorithm by testing the performance of standard ML algorithms when they are trained on synthetic data and tested on real published datasets. Our synthetic data generation algorithm outperforms other literature tabular-data generators, and approaches the performance of training purely with real data. [ABSTRACT FROM AUTHOR]

This study introduces the global-local least-squares support vector machine (GLocal-LS-SVM), a novel machine learning algorithm that combines the strengths of localised and global learning. GLocal-LS-SVM addresses the challenges associated with decentralised data sources, large datasets, and input-space-related issues. The algorithm is a double-layer learning approach that employs multiple local LS-SVM models in the first layer and one global LS-SVM model in the second layer. The key idea behind GLocal-LS-SVM is to extract the most informative data points, known as support vectors, from each local region in the input space. Local LS-SVM models are developed for each region to identify the most contributing data points with the highest support values. The local support vectors are then merged at the final layer to form a reduced training set used to train the global model. We evaluated the performance of GLocal-LS-SVM using both synthetic and real-world datasets. Our results demonstrate that GLocal-LS-SVM achieves comparable or superior classification performance compared to standard LS-SVM and state-of-the-art models. In addition, our experiments show that GLocal-LS-SVM outperforms standard LS-SVM in terms of computational efficiency. For instance, on a training dataset of 9, 000 instances, the average training time for GLocal-LS-SVM was only 2% of the time required to train the LS-SVM model while maintaining classification performance. In summary, the GLocal-LS-SVM algorithm offers a promising solution to address the challenges associated with decentralised data sources and large datasets while maintaining high classification performance. Furthermore, its computational efficiency makes it a valuable tool for practical applications in various domains. [ABSTRACT FROM AUTHOR]

Chronic Kidney disease (CKD) is an important yet under-recognized contributor to morbidity and mortality globally. Machine-learning (ML) based decision support tools have been developed across many aspects of CKD care. Notably, algorithms developed in the prediction and diagnosis of CKD development and progression may help to facilitate early disease prevention, assist with early planning of renal replacement therapy, and offer potential clinical and economic benefits to patients and health systems. Clinical implementation can be affected by the uncertainty surrounding the methodological rigor and performance of ML-based models. This systematic review aims to evaluate the application of prognostic and diagnostic ML tools in CKD development and progression. The protocol has been prepared using the Preferred Items for Systematic Review and Meta-analysis Protocols (PRISMA-P) guidelines. The systematic review protocol for CKD prediction and diagnosis have been registered with the International Prospective Register of Systematic Reviews (PROSPERO) (CRD42022356704, CRD42022372378). A systematic search will be undertaken of PubMed, Embase, the Cochrane Central Register of Controlled Trials (CENTRAL), the Web of Science, and the IEEE Xplore digital library. Studies in which ML has been applied to predict and diagnose CKD development and progression will be included. The primary outcome will be the comparison of the performance of ML-based models with non-ML-based models. Secondary analysis will consist of model use cases, model construct, and model reporting quality. This systematic review will offer valuable insight into the performance and reporting quality of ML-based models in CKD diagnosis and prediction. This will inform clinicians and technical specialists of the current development of ML in CKD care, as well as direct future model development and standardization. [ABSTRACT FROM AUTHOR]

Rapidly developing technology for large scale neural recordings has allowed researchers to measure the activity of hundreds to thousands of neurons at single cell resolution in vivo. Neural decoding analyses are a widely used tool used for investigating what information is represented in this complex, high-dimensional neural population activity. Most population decoding methods assume that correlated activity between neurons has been estimated accurately. In practice, this requires large amounts of data, both across observations and across neurons. Unfortunately, most experiments are fundamentally constrained by practical variables that limit the number of times the neural population can be observed under a single stimulus and/or behavior condition. Therefore, new analytical tools are required to study neural population coding while taking into account these limitations. Here, we present a simple and interpretable method for dimensionality reduction that allows neural decoding metrics to be calculated reliably, even when experimental trial numbers are limited. We illustrate the method using simulations and compare its performance to standard approaches for dimensionality reduction and decoding by applying it to single-unit electrophysiological data collected from auditory cortex. [ABSTRACT FROM AUTHOR]

Introduction: Implementation of evidence-based care for heavy drinking and depression remains low in global health systems. We tested the impact of providing community support, training, and clinical packages of varied intensity on depression screening and management for heavy drinking patients in Latin American primary healthcare. Materials and methods: Quasi-experimental study involving 58 primary healthcare units in Colombia, Mexico and Peru randomized to receive: (1) usual care (control); (2) training using a brief clinical package; (3) community support plus training using a brief clinical package; (4) community support plus training using a standard clinical package. Outcomes were proportion of: (1) heavy drinking patients screened for depression; (2) screen-positive patients receiving appropriate support; (3) all consulting patients screened for depression, irrespective of drinking status. Results: 550/615 identified heavy drinkers were screened for depression (89.4%). 147/230 patients screening positive for depression received appropriate support (64%). Amongst identified heavy drinkers, adjusting for country, sex, age and provider profession, provision of community support and training had no impact on depression activity rates. Intensity of clinical package also did not affect delivery rates, with comparable performance for brief and standard versions. However, amongst all consulting patients, training providers resulted in significantly higher rates of alcohol measurement and in turn higher depression screening rates; 2.7 times higher compared to those not trained. Conclusions: Training using a brief clinical package increased depression screening rates in Latin American primary healthcare. It is not possible to determine the effectiveness of community support on depression activity rates due to the impact of COVID-19. [ABSTRACT FROM AUTHOR]

Introduction: Training Basic Life Support saves lives. However, current BLS training approaches are time-consuming and costly. Alternative cost-efficient and effective training methods are highly needed. The present study evaluated whether a video-feedback supported peer-guided Basic Life Support training approach achieves similar practical performance as a standard instructor-guided training in laypersons. Methods: In a randomized controlled non-inferiority trial, 288 first-year medical students were randomized to two study arms with different Basic Life Support training methods: 1) Standard Instructor Feedback (SIF) or 2) a Peer Video Feedback (PVF). Outcome parameters were objective data for Basic Life Support performance (compression depth and rate) from a resuscitation manikin with recording software as well as overall Basic Life Support performance and subjective confidence. Non-inferiority margins (Δ) for these outcome parameters and sample size calculation were based on previous studies with Standard Instructor Feedback. Two-sided 95% confidence intervals were employed to determine significance of non-inferiority. Results: Results confirmed non-inferiority of Peer Video Feedback to Standard Instructor Feedback for compression depth (proportion difference PVF–SIF = 2.9%; 95% CI: -8.2% to 14.1%; Δ = -19%), overall Basic Life Support performance (proportion difference PVF–SIF = 6.7%; 95% CI: 0.0% to 14.3%; Δ = -27%) and subjective confidence for CPR performance (proportion difference PVF–SIF = -0.01; 95% CI: -0.18–0.17; Δ = -0.5) and emergency situations (proportion difference PVF–SIF = -0.02; 95% CI: -0.21–0.18; Δ = -0.5). Results for compression rate were inconclusive. Discussion: Peer Video Feedback achieves comparable results as standard instructor-based training methods. It is an easy-to-apply and cost-efficient alternative to standard Basic Life Support training methods. To improve performance with respect to compression rate, additional implementation of a metronome is recommended. [ABSTRACT FROM AUTHOR]