Robustness of Deep Learning models in electrocardiogram noise detection and classification.

BACKGROUND AND OBJECTIVE: Automatic electrocardiogram (ECG) signal analysis for heart disease detection has gained significant attention due to busy lifestyles. However, ECG signals are susceptible to noise, which adversely affects the performance of ECG signal analysers. Traditional blind filtering methods use predefined noise frequency and filter order, but they alter ECG biomarkers. Several Deep Learning-based ECG noise detection and classification methods exist, but no study compares recurrent neural network (RNN) and convolutional neural network (CNN) architectures and their complexity. METHODS: This paper introduces a knowledge-based ECG filtering system using Deep Learning to classify ECG noise types and compare popular computer vision model architectures in a practical Internet of Medical Things (IoMT) framework. Experimental results demonstrate that the CNN-based ECG noise classifier outperforms the RNN-based model in terms of performance and training time. RESULTS: The study shows that AlexNet, visual geometry group (VGG), and residual network (ResNet) achieved over 70% accuracy, specificity, sensitivity, and F1 score across six datasets. VGG and ResNet performances were comparable, but VGG was more complex than ResNet, with only a 4.57% less F1 score. CONCLUSIONS: This paper introduces a Deep Learning (DL) based ECG noise classifier for a knowledge-driven ECG filtering system, offering selective filtering to reduce signal distortion. Evaluation of various CNN and RNN-based models reveals VGG and Resnet outperform. Further, the VGG model is superior in terms of performance. But Resnet performs comparably to VGG with less model complexity.

Impact of synthetic noise signature and physiologic ECG signal on designing ML-based ECG noise detection framework.

Automatic signal analysis using artificial intelligence is getting popular in digital healthcare, such as ECG rhythm analysis, where ECG signals are collected from traditional ECG machines or wearable ECG sensors. However, the risk of using an automated system for ECG analysis when noise is present can lead to incorrect diagnosis or treatment decisions. A noise detector is crucial to minimise the risk of incorrect diagnosis. Machine learning (ML) models are used in ECG noise detection before clinical decision-making systems to mitigate false alarms. However, it is essential to prove the generalisation capability of the ML model in different situations. ML models performance is 50% lesser when the model is trained with synthetic and tested with physiologic ECG datasets compared to trained and tested with physiologic ECG datasets. This suggests that the ML model must be trained with physiologic ECG datasets rather than synthetic ones or add more various types of noise in synthetic ECG datasets that can mimic physiologic ECG.Clinical relevance- ML model trained with synthetic noisy ECG can increase the 50% misclassification rate in ECG noise detection compared to training with physiologic ECG datasets. The wrong classification of noise-free and noisy ECG will lead to misdiagnosis regarding the patient's condition, which could be a cause of death.

Domain Agnostic Post-Processing for QRS Detection Using Recurrent Neural Network.

Deep-learning-based QRS-detection algorithms often require essential post-processing to refine the output prediction-stream for R-peak localisation. The post-processing involves basic signal-processing tasks including the removal of random noise in the model's prediction stream using a basic Salt and Pepper filter, as well as, tasks that use domain-specific thresholds, including a minimum QRS size, and a minimum or maximum R-R distance. These thresholds were found to vary among QRS-detection studies and empirically determined for the target dataset, which may have implications if the target dataset differs such as the drop of performance in unknown test datasets. Moreover, these studies, in general, fail to identify the relative strengths of deep-learning models and the post-processing to weigh them appropriately. This study identifies the domain-specific post-processing, as found in the QRS-detection literature, as three steps based on the required domain knowledge. It was found that the use of minimal domain-specific post-processing is often sufficient for most of the cases and the use of additional domain-specific refinement ensures superior performance, however, it makes the process biased towards the training data and lacks generalisability. As a remedy, a domain-agnostic automated post-processing is introduced where a separate recurrent neural network (RNN)-based model learns required post-processing from the output generated from a QRS-segmenting deep learning model, which is, to the best of our knowledge, the first of its kind. The RNN-based post-processing shows superiority over the domain-specific post-processing for most of the cases (with shallow variants of the QRS-segmenting model and datasets like TWADB) and lags behind for others but with a small margin ( ≤ 2%). The consistency of the RNN-based post-processor is an important characteristic which can be utilised in designing a stable and domain agnostic QRS detector.

Interpretability and Optimisation of Convolutional Neural Networks Based on Sinc-Convolution.

Interpretability often seeks domain-specific facts, which is understandable to human, from deep-learning (DL) or other machine-learning (ML) models of black-box nature. This is particularly important to establish transparency in ML model's inner-working and decision-making, so that a certain level of trust is achieved when a model is deployed in a sensitive and mission-critical context, such as health-care. Model-level transparency can be achieved when its components are transparent and are capable of explaining reason of a decision, for a given input, which can be linked to domain-knowledge. This article used convolutional neural network (CNN), with sinc-convolution as its constrained first-layer, to explore if such a model's decision-making can be explained, for a given task, by observing the sinc-convolution's sinc-kernels. These kernels work like band-pass filters, having only two parameters per kernel - lower and upper cutoff frequencies, and optimised through back-propagation. The optimised frequency-bands of sinc-kernels may provide domain-specific insights for a given task. For a given input instance, the effects of sinc-kernels was visualised by means of explanation vector, which may help to identify comparatively significant frequency-bands, that may provide domain-specific interpretation, for the given task. In addition, a CNN model was further optimised by considering the identified subset of prominent sinc frequency-bands as the constrained first-layer, which yielded comparable or better performance, as compared to its all sinc-bands counterpart, as well as, a classical CNN. A minimal CNN structure, achieved through such an optimisation process, may help design task-specific interpretable models. To the best of our knowledge, the idea of sinc-convolution layer's task-specific significant sinc-kernel-based network optimisation is the first of its kind. Additionally, the idea of explanation-vector-based joint time-frequency representation to analyse time-series signals is rare in the literature. The above concept was validated for two tasks, ECG beat-classification (five-class classification task), and R-peak localisation (sample-wise segmentation task).

Performance of a Convolutional Neural Network Derived From PPG Signal in Classifying Sleep Stages.

Automatic sleep stage classification is vital for evaluating the quality of sleep. Conventionally, sleep is monitored using multiple physiological sensors that are uncomfortable for long-term monitoring and require expert intervention. In this study, we propose an automatic technique for multi-stage sleep classification using photoplethysmographic (PPG) signal. We have proposed a convolutional neural network (CNN) that learns directly from the PPG signal and classifies multiple sleep stages. We developed models for two- (Wake-Sleep), three- (Wake-NREM-REM) and four- (Wake-Light sleep-Deep sleep-REM) stages of sleep classification. Our proposed approach shows an average classification accuracy of 94.4%, 94.2%, and 92.9% for two, three, and four stages, respectively. Experimental results show that the proposed CNN model outperforms existing state-of-the-art models (classical and deep learning) in the literature.

Robustness of electrocardiogram signal quality indices.

Electrocardiogram (ECG) signal quality indices (SQIs) are essential for improving diagnostic accuracy and reliability of ECG analysis systems. In various practical applications, the ECG signals are corrupted by different types of noise. These corrupted ECG signals often provide insufficient and incorrect information regarding a patient's health. To solve this problem, signal quality measurements should be made before an ECG signal is used for decision-making. This paper investigates the robustness of existing popular statistical signal quality indices (SSQIs): relative power of QRS complex (SQIp), skewness (SQIskew), signal-to-noise ratio (SQIsnr), higher order statistics SQI (SQIhos) and peakedness of kurtosis (SQIkur). We analysed the robustness of these SSQIs against different window sizes across diverse datasets. Results showed that the performance of SSQIs considerably fluctuates against varying datasets, whereas the impact of varying window sizes was minimal. This fluctuation occurred due to the use of a static threshold value for classifying noise-free ECG signals from the raw ECG signals. Another drawback of these SSQIs is the bias towards noise-free ECG signals, that limits their usefulness in clinical settings. In summary, the fixed threshold-based SSQIs cannot be used as a robust noise detection system. In order to solve this fixed threshold problem, other techniques can be developed using adaptive thresholds and machine-learning mechanisms.

Precision psychiatry with immunological and cognitive biomarkers: a multi-domain prediction for the diagnosis of bipolar disorder or schizophrenia using machine learning.

Precision psychiatry is attracting increasing attention lately as a recognized priority. One of the goals of precision psychiatry is to develop tools capable of aiding a clinically informed psychiatric diagnosis objectively. Cognitive, inflammatory and immunological factors are altered in both bipolar disorder (BD) and schizophrenia (SZ), however, most of these alterations do not respect diagnostic boundaries from a phenomenological perspective and possess great variability in different individuals with the same phenotypic diagnosis and, consequently, none so far has proven to have the ability of reliably aiding in the differential diagnosis of BD and SZ. We developed a probabilistic multi-domain data integration model consisting of immune and inflammatory biomarkers in peripheral blood and cognitive biomarkers using machine learning to predict diagnosis of BD and SZ. A total of 416 participants, being 323, 372, and 279 subjects for blood, cognition and combined biomarkers analysis, respectively. Our multi-domain model performances for the BD vs. control (sensitivity 80% and specificity 71%) and for the SZ vs. control (sensitivity 84% and specificity 81%) pairs were high in general, however, our multi-domain model had only moderate performance for the differential diagnosis of BD and SZ (sensitivity 71% and specificity 73%). In conclusion, our results show that the diagnosis of BD and of SZ, and that the differential diagnosis of BD and SZ can be predicted with possible clinical utility by a computational machine learning algorithm employing blood and cognitive biomarkers, and that their integration in a multi-domain outperforms algorithms based in only one domain. Independent studies are needed to validate these findings.

Identifying smoker subgroups with high versus low smoking cessation attempt probability: A decision tree analysis approach.

BACKGROUND: Regression-based research has successfully identified independent predictors of smoking cessation, both its initiation and maintenance. However, it is unclear how these various independent predictors interact with each other and conjointly influence smoking behaviour. As a proof-of-concept, this study used decision tree analysis (DTA) to identify the characteristics of smoker subgroups with high versus low smoking cessation initiation probability based on the conjoint effects of four predictor variables, and determine any variations by socio-economic status (SES). METHODS: Data come from the Australian arm of the ITC project, a longitudinal cohort study of adult smokers followed up approximately annually. Reported wanting to quit smoking, worries about smoking negative health impact, quitting self-efficacy and quit intentions assessed in 2005 were used as predictors and reported quit attempts at the 2006 follow-up survey were used as the outcome for the initial model calibration and validation analyses (n = 1475), and further cross-validated using the 2012-2013 data (n = 787). RESULTS: DTA revealed that while all four predictor variables conjointly contributed to the identification of subgroups with high versus low smoking cessation initiation probability, quit intention was the most important predictor common across all SES strata. The relative importance of the other predictors showed differences by SES. CONCLUSIONS: Modifiable characteristics of smoker subgroups associated with making a quit attempt and any variations by SES can be successfully identified using a decision tree analysis approach, to provide insights as to who might benefit from targeted intervention, thus, underscoring the value of this approach to complement the conventional regression-based approach.

Multi-lag tone-entropy in neonatal stress.

Heart rate variability (HRV) has been analysed using linear and nonlinear methods. In the framework of a controlled neonatal stress model, we applied tone-entropy (T-E) analysis at multiple lags to understand the influence of external stressors on healthy term neonates. Forty term neonates were included in the study. HRV was analysed using multi-lag T-E at two resting and two stress phases (heel stimulation and a heel stick blood drawing phase). Higher mean entropy values and lower mean tone values when stressed showed a reduction in randomness with increased sympathetic and reduced parasympathetic activity. A ROC analysis was used to estimate the diagnostic performances of tone and entropy and combining both features. Comparing the resting and simulation phase separately, the performance of tone outperformed entropy, but combining the two in a quadratic linear regression model, neonates in resting as compared to stress phases could be distinguished with high accuracy. This raises the possibility that when applied across short time segments, multi-lag T-E becomes an additional tool for more objective assessment of neonatal stress.

Exploring reasoning mechanisms in ward rounds: a critical realist multiple case study.

BACKGROUND: Ward rounds are an important and ubiquitous element of hospital care with a history extending well over a century. Although originally intended as a means of educating medical trainees and junior doctors, over time they have become focused on supporting clinical practice. Surprisingly, given their ubiquity and importance, they are under-researched and inadequately understood. This study aims to contribute knowledge in human reasoning within medical teams, meeting a pressing need for research concerning the reasoning occurring in rounds. METHODS: The research reported here aimed to improve the understanding of ward round reasoning by conducting a critical realist case study exploring the collaborative group reasoning mechanisms in the ward rounds of two hospitals in Victoria, Australia. The data collection involved observing rounds, interviewing medical practitioners and holding focus group meetings. RESULTS: Nine group reasoning mechanisms concerning sharing, agreeing and recording information in the categories of information accumulation, sense-making and decision-making were identified, together forming a program theory of ward round reasoning. In addition, themes spanning across mechanisms were identified, further explaining ward round reasoning and suggesting avenues for future exploration. Themes included the use of various criteria, tensions involving mechanisms, time factors, medical roles and hierarchies. CONCLUSIONS: This paper contributes to the literature by representing rounds in a manner that strengthens understanding of the form of the group reasoning occurring within, thus supporting theory-based evaluation strategies, redesigned practices and training enhancements.

Detection of epileptic seizure based on entropy analysis of short-term EEG.

Entropy measures that assess signals' complexity have drawn increasing attention recently in biomedical field, as they have shown the ability of capturing unique features that are intrinsic and physiologically meaningful. In this study, we applied entropy analysis to electroencephalogram (EEG) data to examine its performance in epilepsy detection based on short-term EEG, aiming at establishing a short-term analysis protocol with optimal seizure detection performance. Two classification problems were considered, i.e., 1) classifying interictal and ictal EEGs (epileptic group) from normal EEGs; and 2) classifying ictal from interictal EEGs. For each problem, we explored two protocols to analyze the entropy of EEG: i) using a single analytical window with different window lengths, and ii) using an average of multiple windows for each window length. Two entropy methods-fuzzy entropy (FuzzyEn) and distribution entropy (DistEn)-were used that have valid outputs for any given data lengths. We performed feature selection and trained classifiers based on a cross-validation process. The results show that performance of FuzzyEn and DistEn may complement each other and the best performance can be achieved by combining: 1) FuzzyEn of one 5-s window and the averaged DistEn of five 1-s windows for classifying normal from epileptic group (accuracy: 0.93, sensitivity: 0.91, specificity: 0.96); and 2) the averaged FuzzyEn of five 1-s windows and DistEn of one 5-s window for classifying ictal from interictal EEGs (accuracy: 0.91, sensitivity: 0.93, specificity: 0.90). Further studies are warranted to examine whether this proposed short-term analysis procedure can help track the epileptic activities in real time and provide prompt feedback for clinical practices.

Detection of ground parrot vocalisation: A multiple instance learning approach.

Ground parrot vocalisation can be considered as an audio event. Test-based diverse density multiple instance learning (TB-DD-MIL) is proposed for detecting this event in audio files recorded in the field. The proposed method is motivated by the advantages of multiple instance learning from incomplete training data. Spectral features suitable for encoding the vocal source information of the ground parrot vocalization are also investigated. The proposed method was benchmarked against a dataset collected in various environmental conditions and an audio detection evaluation scheme is proposed. The evaluation includes a study on performance of the various vocal source features and comparison with other classification techniques. Experimental results indicated that the most appropriate feature to encode ground parrot calls is the spectral bandwidth and the proposed TB-DD-MIL method outperformed other existing classification methods.

Guidelines for Developing and Reporting Machine Learning Predictive Models in Biomedical Research: A Multidisciplinary View.

BACKGROUND: As more and more researchers are turning to big data for new opportunities of biomedical discoveries, machine learning models, as the backbone of big data analysis, are mentioned more often in biomedical journals. However, owing to the inherent complexity of machine learning methods, they are prone to misuse. Because of the flexibility in specifying machine learning models, the results are often insufficiently reported in research articles, hindering reliable assessment of model validity and consistent interpretation of model outputs. OBJECTIVE: To attain a set of guidelines on the use of machine learning predictive models within clinical settings to make sure the models are correctly applied and sufficiently reported so that true discoveries can be distinguished from random coincidence. METHODS: A multidisciplinary panel of machine learning experts, clinicians, and traditional statisticians were interviewed, using an iterative process in accordance with the Delphi method. RESULTS: The process produced a set of guidelines that consists of (1) a list of reporting items to be included in a research article and (2) a set of practical sequential steps for developing predictive models. CONCLUSIONS: A set of guidelines was generated to enable correct application of machine learning models and consistent reporting of model specifications and results in biomedical research. We believe that such guidelines will accelerate the adoption of big data analysis, particularly with machine learning methods, in the biomedical research community.

Hybrid metaheuristic approaches to the expectation maximization for estimation of the hidden Markov model for signal modeling.

The expectation maximization (EM) is the standard training algorithm for hidden Markov model (HMM). However, EM faces a local convergence problem in HMM estimation. This paper attempts to overcome this problem of EM and proposes hybrid metaheuristic approaches to EM for HMM. In our earlier research, a hybrid of a constraint-based evolutionary learning approach to EM (CEL-EM) improved HMM estimation. In this paper, we propose a hybrid simulated annealing stochastic version of EM (SASEM) that combines simulated annealing (SA) with EM. The novelty of our approach is that we develop a mathematical reformulation of HMM estimation by introducing a stochastic step between the EM steps and combine SA with EM to provide better control over the acceptance of stochastic and EM steps for better HMM estimation. We also extend our earlier work and propose a second hybrid which is a combination of an EA and the proposed SASEM, (EA-SASEM). The proposed EA-SASEM uses the best constraint-based EA strategies from CEL-EM and stochastic reformulation of HMM. The complementary properties of EA and SA and stochastic reformulation of HMM of SASEM provide EA-SASEM with sufficient potential to find better estimation for HMM. To the best of our knowledge, this type of hybridization and mathematical reformulation have not been explored in the context of EM and HMM training. The proposed approaches have been evaluated through comprehensive experiments to justify their effectiveness in signal modeling using the speech corpus: TIMIT. Experimental results show that proposed approaches obtain higher recognition accuracies than the EM algorithm and CEL-EM as well.

A novel canonical dual computational approach for prion AGAAAAGA amyloid fibril molecular modeling.

Many experimental studies have shown that the prion AGAAAAGA palindrome hydrophobic region (113-120) has amyloid fibril forming properties and plays an important role in prion diseases. However, due to the unstable, noncrystalline and insoluble nature of the amyloid fibril, to date structural information on AGAAAAGA region (113-120) has been very limited. This region falls just within the N-terminal unstructured region PrP (1-123) of prion proteins. Traditional X-ray crystallography and nuclear magnetic resonance (NMR) spectroscopy experimental methods cannot be used to get its structural information. Under this background, this paper introduces a novel approach of the canonical dual theory to address the 3D atomic-resolution structure of prion AGAAAAGA amyloid fibrils. The novel and powerful canonical dual computational approach introduced in this paper is for the molecular modeling of prion AGAAAAGA amyloid fibrils, and that the optimal atomic-resolution structures of prion AGAAAAGA amyloid fibils presented in this paper are useful for the drive to find treatments for prion diseases in the field of medicinal chemistry. Overall, this paper presents an important method and provides useful information for treatments of prion diseases.

A constraint-based evolutionary learning approach to the expectation maximization for optimal estimation of the hidden Markov model for speech signal modeling.

This paper attempts to overcome the tendency of the expectation-maximization (EM) algorithm to locate a local rather than global maximum when applied to estimate the hidden Markov model (HMM) parameters in speech signal modeling. We propose a hybrid algorithm for estimation of the HMM in automatic speech recognition (ASR) using a constraint-based evolutionary algorithm (EA) and EM, the CEL-EM. The novelty of our hybrid algorithm (CEL-EM) is that it is applicable for estimation of the constraint-based models with many constraints and large numbers of parameters (which use EM) like HMM. Two constraint-based versions of the CEL-EM with different fusion strategies have been proposed using a constraint-based EA and the EM for better estimation of HMM in ASR. The first one uses a traditional constraint-handling mechanism of EA. The other version transforms a constrained optimization problem into an unconstrained problem using Lagrange multipliers. Fusion strategies for the CEL-EM use a staged-fusion approach where EM has been plugged with the EA periodically after the execution of EA for a specific period of time to maintain the global sampling capabilities of EA in the hybrid algorithm. A variable initialization approach (VIA) has been proposed using a variable segmentation to provide a better initialization for EA in the CEL-EM. Experimental results on the TIMIT speech corpus show that CEL-EM obtains higher recognition accuracies than the traditional EM algorithm as well as a top-standard EM (VIA-EM, constructed by applying the VIA to EM).

A hybrid neural learning algorithm using evolutionary learning and derivative free local search method.

In this paper we investigate a hybrid model based on the Discrete Gradient method and an evolutionary strategy for determining the weights in a feed forward artificial neural network. Also we discuss different variants for hybrid models using the Discrete Gradient method and an evolutionary strategy for determining the weights in a feed forward artificial neural network. The Discrete Gradient method has the advantage of being able to jump over many local minima and find very deep local minima. However, earlier research has shown that a good starting point for the discrete gradient method can improve the quality of the solution point. Evolutionary algorithms are best suited for global optimisation problems. Nevertheless they are cursed with longer training times and often unsuitable for real world application. For optimisation problems such as weight optimisation for ANNs in real world applications the dimensions are large and time complexity is critical. Hence the idea of a hybrid model can be a suitable option. In this paper we propose different fusion strategies for hybrid models combining the evolutionary strategy with the discrete gradient method to obtain an optimal solution much quicker. Three different fusion strategies are discussed: a linear hybrid model, an iterative hybrid model and a restricted local search hybrid model. Comparative results on a range of standard datasets are provided for different fusion hybrid models.