Sample Entropy Computation on Signals with Missing Values

Sample entropy embeds time series into m-dimensional spaces and estimates entropy based on the distances between points in these spaces. However, when samples can be considered as missing or invalid, defining distance in the embedding space becomes problematic. Preprocessing techniques, such as deletion or interpolation, can be employed as a solution, producing time series without missing or invalid values. While deletion ignores missing values, interpolation replaces them using approximations based on neighboring points. This paper proposes a novel approach for the computation of sample entropy when values are considered as missing or invalid. The proposed algorithm accommodates points in the m-dimensional space and handles them there. A theoretical and experimental comparison of the proposed algorithm with deletion and interpolation demonstrates several advantages over these other two approaches. Notably, the deviation of the expected sample entropy value for the proposed methodology consistently proves to be lowest one.

How Do Young Soccer Players Train? A 5-Year Analysis of Weekly Training Load and its Variability Between Age Groups in an Elite Youth Academy

ABSTRACT

Connolly, DR, Stolp, S, Gualtieri, A, Ferrari Bravo, D, Sassi, R, Rampinini, E, and Coutts, AJ. How do young soccer players train? A 5-year analysis of weekly training load and its variability between age groups in an elite youth academy. J Strength Cond Res 38(8): e423-e429, 2024-The aim of this study was to quantify the session rating of perceived exertion (sRPE), duration, and training load accrued across typical training weeks undertaken by youth soccer players. Differences between starters, nonstarters, and variations in training load variables were also investigated. Data were collected from 230 elite youth players in 4 age groups (U15, U16, U17, and U19) during 5 competitive seasons. Mixed models were used to describe variation between age groups and compare starters with nonstarters, with season as a fixed covariate effect. Week-to-week variation in training load was expressed as the percentage coefficient of variation. The main findings may be used to highlight a significant effect of age and playing status on training intensity, duration, and internal training load. Weekly training load increased progressively from the U15 to U17, with significant differences between each age group (p < 0.03). Lower mean weekly perceived intensity (sRPE) was noted in U15 when compared with the older age groups (4.2 vs. 4.6-4.9 arbitrary unit for U16 to U19, p < 0.001). Low weekly training load variation was observed across the different phases of the season in each age group, with the preseason exhibiting the greatest variance (3.6-6.2%). Differences in the training load are likely more attributable to changes in training duration rather than sRPE. Control of session duration seems to play an important role when aiming to control load in the academy environment, and practitioners should closely monitor the differences in duration and load being recorded between starters and nonstarters.

Surprising concentrations of hydrogen and non-geological methane and carbon dioxide in the soil

Due to its potential use as a carbon-free energy resource with minimal environmental and climate impacts, natural hydrogen (H2) produced by subsurface geochemical processes is today the target of intensive research. In H2 exploration practices, bacteria are thought to swiftly consume H2 and, therefore, small near-surface concentrations of H2, even orders of 102 ppmv in soils, are considered a signal of active migration of geological gas, potentially revealing underground resources. Here, we document an extraordinary case of a widespread occurrence of H2 (up to 1 vol%), together with elevated concentrations of CH4 and CO2 (up to 51 and 27 vol%, respectively), in aerated meadow soils along Italian Alps valleys. Based on current literature, this finding would be classified as a discovery of pervasive and massive geological H2 seepage. Nevertheless, an ensemble of gas geochemical and soil microbiological analyses, including bulk and clumped CH4 isotopes, radiocarbon of CH4 and CO2, and DNA and mcrA gene quantitative polymerase chain reaction analyses, revealed that H2 was only coupled to modern microbial gas. The H2-CO2-CH4-H2S association, wet soil proximity, and the absence of other geogenic gases in soils and springs suggest that H2 derives from near-surface fermentation, rather than geological degassing. H2 concentrations up to 1 vol% in soils are not conclusive evidence of deep gas seepage. This study provides a new reference for the potential of microbial H2, CH4 and CO2 in soils, to be considered in H2 exploration guidelines and soil carbon and greenhouse-gas cycle research.

Early warning of atrial fibrillation using deep learning

Atrial fibrillation (AF), the most prevalent cardiac rhythm disorder, significantly increases hospitalization and health risks. Reverting from AF to sinus rhythm (SR) often requires intensive interventions. This study presents a deep-learning model capable of predicting the transition from SR to AF on average 30.8 min before the onset appears, with an accuracy of 83% and an F1 score of 85% on the test data. This performance was obtained from R-to-R interval signals, which can be accessible from wearable technology. Our model, entitled Warning of Atrial Fibrillation (WARN), consists of a deep convolutional neural network trained and validated on 24-h Holter electrocardiogram data from 280 patients, with 70 additional patients used for testing and further evaluation on 33 patients from two external centers. The low computational cost of WARN makes it ideal for integration into wearable technology, allowing for continuous heart monitoring and early AF detection, which can potentially reduce emergency interventions and improve patient outcomes.

Nutritional counselling and risk factors for obesity: an observational study in toddlers

BACKGROUND

Nutrition exerts a fundamental role in the prevention of obesity (OB). The aim of this study was to assess the extent to which well recognized risk factors for early OB can be associated to overweight (OW) or OB under a standardized nutritional approach and surveillance in toddlers.

METHODS

The eligible population was represented by 676 toddlers aged 24-36 months, assigned to 18 primary care pediatricians trained on nutritional issues who shared a standardized nutritional approach. Six-hundred-twenty-nine children (333 boys), mean age 27.8 ± 4.2 months were effectively included in this observational study. Parents received nutritional advice with particular emphasis to proteins and sugar composition supported by leaflets and reinforced at each visit. Body mass index was assessed at the age of 24-36 months. The following individual and family risk factors were considered: gestational age, birth weight, eutocic/caesarean delivery, milk feeding history, household smoking or antibiotics exposure, parents' weight, height and educational level. Prevalence of OW/OB was compared to a group of 742 toddlers (373 boys) under usual care.

RESULTS

Under a standardized nutritional counselling, 28.1% toddlers were classified as OW/OB compared to 36.9% toddlers under usual care (p = 0.005). In unadjusted models, parental OW/OB was significantly associated to OW/OB in toddlers (p < 0.01), while high birth weight did not reach statistical significance (p = 0.07). In adjusted models, including all the explanatory variables studied, only paternal OW/OB vs. normal weight was significantly associated to OW/OB in toddlers (OR 2.035, 95% confidence interval 1.206-3.436). No protective effect of exclusive breast feeding during the first 6 months of age was demonstrated.

CONCLUSIONS

Toddlers under a standardized nutrition counselling focused to limit protein and simple sugars, showed lower prevalence of OW/OB compared to usual care. Healthy promotion activities should take into account the influence of paternal BMI on the offspring adiposity.

Imaging in osteogenesis imperfecta: Where we are and where we are going

Osteogenesis imperfecta (OI) is a rare phenotypically and genetically heterogeneous group of inherited skeletal dysplasias. The hallmark features of OI include bone fragility and susceptibility to fractures, bone deformity, and diminished growth, along with a plethora of associated secondary features (both skeletal and extraskeletal). The diagnosis of OI is currently made on clinical grounds and may be confirmed by genetic testing. However, imaging remains pivotal in the evaluation of this disease. The aim of this article is to review the current role played by the various radiologic techniques in the diagnosis and monitoring of OI in the postnatal setting as well as to discuss recent advances and future perspectives in OI imaging. Conventional Radiography and Dual-energy X-ray Absorptiometry (DXA) are currently the two most used imaging modalities in OI. The cardinal radiographic features of OI include generalized osteopenia/osteoporosis, bone deformities, and fractures. DXA is currently the most available technique to assess Bone Mineral Density (BMD), specifically areal BMD (aBMD). However, DXA has important limitations and cannot fully characterize bone fragility in OI based on aBMD. Novel DXA-derived parameters, such as Trabecular Bone Score (TBS), may provide further insight into skeletal changes induced by OI, but evidence is still limited. Techniques like Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) can be useful as problem-solvers or in specific settings, including the evaluation of cranio-cervical abnormalities. Recent evidence supports the use of High-Resolution peripheral Quantitative Computed Tomography (HR-pQCT) as a promising tool to improve the characterization of bone fragility in OI. However, HR-pQCT remains a primarily research technique at present. Quantitative Computed Tomography (QCT) is an alternative to DXA for the determination of BMD at central sites, with distinct advantages but considerably higher radiation exposure. Quantitative Ultrasound (QUS) is a portable, inexpensive, and radiation-free modality that may complement DXA evaluation, providing information on bone quality. However, evidence of usefulness of QUS in OI is poor. Radiofrequency Echographic Multi Spectrometry (REMS) is an emerging non-ionizing imaging method that holds promise for the diagnosis of low BMD and for the prediction of fracture risk, but so far only one published study has investigated its role in OI. To conclude, several different radiologic techniques have proven to be effective in the diagnosis and monitoring of OI, each with their own specificities and peculiarities. Clinicians should be aware of the strategic role of the various modalities in the different phases of the patient care process. In this scenario, the development of international guidelines including recommendations on the role of imaging in the diagnosis and monitoring of OI, accompanied by continuous active research in the field, could significantly improve the standardization of patient care.

Maternal prenatal depression is associated with dysregulation over the first five years of life moderated by child polygenic risk for comorbid psychiatric problems

Dysregulation is a combination of emotion, behavior, and attention problems associated with lifelong psychiatric comorbidity. There is evidence for the stability of dysregulation from childhood to adulthood, which would be more fully characterized by determining the likely stability from infancy to childhood. Early origins of dysregulation can further be validated and contextualized in association with environmental and biological factors, such as prenatal stress and polygenic risk scores (PRS) for overlapping child psychiatric problems. We aimed to determine trajectories of dysregulation from 3 months to 5 years (N = 582) in association with maternal prenatal depression moderated by multiple child PRS (N = 232 pairs with available PRS data) in a prenatal cohort. Mothers reported depression symptoms at 24-26 weeks' gestation and child dysregulation at 3, 6, 18, 36, 48, and 60 months. The PRS were for major depressive disorder, attention deficit hyperactivity disorder, cross disorder, and childhood psychiatric problems. Covariates were biological sex, maternal education, and postnatal depression. Analyses included latent classes and regression. Two dysregulation trajectories emerged: persistently low dysregulation (94%), and increasingly high dysregulation (6%). Stable dysregulation emerged at 18 months. High dysregulation was associated with maternal prenatal depression, moderated by PRS for child comorbid psychiatric problems. Males were at greater risk of high dysregulation.

A Systematic Survey of Data Augmentation of ECG Signals for AI Applications

AI techniques have recently been put under the spotlight for analyzing electrocardiograms (ECGs). However, the performance of AI-based models relies on the accumulation of large-scale labeled datasets, which is challenging. To increase the performance of AI-based models, data augmentation (DA) strategies have been developed recently. The study presented a comprehensive systematic literature review of DA for ECG signals. We conducted a systematic search and categorized the selected documents by AI application, number of leads involved, DA method, classifier, performance improvements after DA, and datasets employed. With such information, this study provided a better understanding of the potential of ECG augmentation in enhancing the performance of AI-based ECG applications. This study adhered to the rigorous PRISMA guidelines for systematic reviews. To ensure comprehensive coverage, publications between 2013 and 2023 were searched across multiple databases, including IEEE Explore, PubMed, and Web of Science. The records were meticulously reviewed to determine their relevance to the study's objective, and those that met the inclusion criteria were selected for further analysis. Consequently, 119 papers were deemed relevant for further review. Overall, this study shed light on the potential of DA to advance the field of ECG diagnosis and monitoring.

Recommender system for ablation lines to treat complex atrial tachycardia

BACKGROUND AND OBJECTIVE

Planning the optimal ablation strategy for the treatment of complex atrial tachycardia (CAT) is a time consuming task and is error-prone. Recently, directed network mapping, a technology based on graph theory, proved to efficiently identify CAT based solely on data of clinical interventions. Briefly, a directed network was used to model the atrial electrical propagation and reentrant activities were identified by looking for closed-loop paths in the network. In this study, we propose a recommender system, built as an optimization problem, able to suggest the optimal ablation strategy for the treatment of CAT.

METHODS

The optimization problem modeled the optimal ablation strategy as that one interrupting all reentrant mechanisms while minimizing the ablated atrial surface. The problem was designed on top of directed network mapping. Considering the exponential complexity of finding the optimal solution of the problem, we introduced a heuristic algorithm with polynomial complexity. The proposed algorithm was applied to the data of i) 6 simulated scenarios including both left and right atrial flutter; and ii) 10 subjects that underwent a clinical routine.

RESULTS

The recommender system suggested the optimal strategy in 4 out of 6 simulated scenarios. On clinical data, the recommended ablation lines were found satisfactory on 67% of the cases according to the clinician's opinion, while they were correctly located in 89%. The algorithm made use of only data collected during mapping and was able to process them nearly real-time.

CONCLUSIONS

The first recommender system for the identification of the optimal ablation lines for CAT, based solely on the data collected during the intervention, is presented. The study may open up interesting scenarios for the application of graph theory for the treatment of CAT.

Antarctic permafrost degassing in Taylor Valley by extensive soil gas investigation

Ongoing studies conducted in northern polar regions reveal that permafrost stability plays a key role in the modern carbon cycle as it potentially stores considerable quantities of greenhouse gases. Rapid and recent warming of the Arctic permafrost is resulting in significant greenhouse gas emissions, both from physical and microbial processes. The potential impact of greenhouse gas release from the Antarctic region has not, to date, been investigated. In Antarctica, the McMurdo Dry Valleys comprise 10 % of the ice-free soil surface areas in Antarctica and like the northern polar regions are also warming albeit at a slower rate. The work presented herein examines a comprehensive sample suite of soil gas (e.g., CO₂, CH₄ and He) concentrations and CO₂ flux measurements conducted in Taylor Valley during austral summer 2019/2020. Analytical results reveal the presence of significant concentrations of CO₂, CH₄ and He (up to 3.44 vol%, 18,447 ppmv and 6.49 ppmv, respectively) at the base of the active layer. When compared with the few previously obtained measurements, we observe increased CO₂ flux rates (estimated CO₂ emissions in the study area of 21.6 km² ≈ 15 tons day^-1). We suggest that the gas source is connected with the deep brines migrating from inland (potentially from beneath the Antarctic Ice Sheet) towards the coast beneath the permafrost layer. These data provide a baseline for future investigations aimed at monitoring the changing rate of greenhouse gas emissions from Antarctic permafrost, and the potential origin of gases, as the southern polar region warms.

Neuroimaging profiling identifies distinct brain maturational subtypes of youth with mood and anxiety disorders

Mood and anxiety disorders typically begin in adolescence and have overlapping clinical features but marked inter-individual variation in clinical presentation. The use of multimodal neuroimaging data may offer novel insights into the underlying brain mechanisms. We applied Heterogeneity Through Discriminative Analysis (HYDRA) to measures of regional brain morphometry, neurite density, and intracortical myelination to identify subtypes of youth, aged 9-10 years, with mood and anxiety disorders (N = 1931) compared to typically developing youth (N = 2823). We identified three subtypes that were robust to permutation testing and sample composition. Subtype 1 evidenced a pattern of imbalanced cortical-subcortical maturation compared to the typically developing group, with subcortical regions lagging behind prefrontal cortical thinning and myelination and greater cortical surface expansion globally. Subtype 2 displayed a pattern of delayed cortical maturation indicated by higher cortical thickness and lower cortical surface area expansion and myelination compared to the typically developing group. Subtype 3 showed evidence of atypical brain maturation involving globally lower cortical thickness and surface coupled with higher myelination and neural density. Subtype 1 had superior cognitive function in contrast to the other two subtypes that underperformed compared to the typically developing group. Higher levels of parental psychopathology, family conflict, and social adversity were common to all subtypes, with subtype 3 having the highest burden of adverse exposures. These analyses comprehensively characterize pre-adolescent mood and anxiety disorders, the biopsychosocial context in which they arise, and lay the foundation for the examination of the longitudinal evolution of the subtypes identified as the study sample transitions through adolescence.

Predicting human cardiac QT alterations and pro-arrhythmic effects of compounds with a 3D beating heart-on-chip platform

Determining the potential cardiotoxicity and pro-arrhythmic effects of drug candidates remains one of the most relevant issues in the drug development pipeline (DDP). New methods enabling to perform more representative preclinical in vitro studies by exploiting induced pluripotent stem cell-derived cardiomyocytes (iPSC-CM) are under investigation to increase the translational power of the outcomes. Here we present a pharmacological campaign conducted to evaluate the drug-induced QT alterations and arrhythmic events on uHeart, a 3D miniaturized in vitro model of human myocardium encompassing iPSC-CM and dermal fibroblasts embedded in fibrin. uHeart was mechanically trained resulting in synchronously beating cardiac microtissues in 1 week, characterized by a clear field potential (FP) signal that was recorded by means of an integrated electrical system. A drug screening protocol compliant with the new International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) guidelines was established and uHeart was employed for testing the effect of 11 compounds acting on single or multiple cardiac ion channels and well-known to elicit QT prolongation or arrhythmic events in clinics. The alterations of uHeart's electrophysiological parameters such as the beating period, the FP duration, the FP amplitude, and the detection of arrhythmic events prior and after drug administration at incremental doses were effectively analyzed through a custom-developed algorithm. Results demonstrated the ability of uHeart to successfully anticipate clinical outcome and to predict the QT prolongation with a sensitivity of 83.3%, a specificity of 100% and an accuracy of 91.6%. Cardiotoxic concentrations of drugs were notably detected in the range of the clinical highest blood drug concentration (Cmax), qualifying uHeart as a fit-to-purpose preclinical tool for cardiotoxicity studies.

Examining attachment, cortisol secretion, and cognitive neurodevelopment in preschoolers and its predictive value for telomere length at age seven

Background

Secure attachment reflects caregiver-child relationship in which the caregiver is responsive when support and comforting are needed by the child. This pattern of bond has an important buffering role in the response to stress by the reduction of the negative experience and its associated physiological response. Disruption of the physiological stress system is thought to be a central mechanism by which early care impacts children. Early life stress causes cellular and molecular changes in brain regions associated with cognitive functions that are fundamental for early learning.

Methods

The association between attachment, cortisol response before and after the Strange Situation Experiment, and neurodevelopment was examined in a sample of 107 preschoolers at age three. Also, the predictive effect of cortisol reactivity and attachment on telomere length at age seven was investigated in a followed-up sample of 77 children.

Results

Children with insecure attachment had higher cortisol secretion and poorer neurodevelopmental skills at age three. A significant cortisol change was observed across the experiment with non-significant interaction with attachment. The attachment and neurodevelopment association was not mediated by cortisol secretion. Preschoolers' attachment and cortisol did not associate nor interacted to predict telomere length at age seven.

Conclusion

These findings add evidence to the detrimental effects of insecure attachment as an aggravator of the physiological response to stress and poorer neurodevelopment during the preschool period. Although attachment and cortisol were not predictive of telomere length, intervention policies that promote secure attachment are more likely to positively echo on several health domains.

Spatial Correlation Between Myocyte's Repolarization Times and their Alternans Drives T-Wave Alternans on the ECG

OBJECTIVE

T-wave alternans (TWA) manifests as beat-to-beat fluctuations of T-wave morphology on the electrocardiogram (ECG), with physiological bases not fully understood. Using a biophysical model of the ECG, we demonstrate and give explicit relations that TWA depends on the i) spatial covariance between myocytes' repolarization time and alternans; and ii) global alternans (common to every myocyte).

METHODS

We quantified the spatial covariance and global alternans by means of two new metrics, R index and δ, respectively. They were validated on both synthetic and real signals. Computerized simulations were generated using a biophysical model linking the action potentials with the surface ECG. Then, the metrics were computed in STAFF-III dataset, containing ECGs from patients who underwent coronary angioplasty with prolonged balloon inflations, and the time courses of the metrics were analyzed together with TWA measured on the surface ECG.

RESULTS

The metrics properly estimated the spatial covariance and global alternans in the synthetic data. In the STAFF-III dataset, the R index progressively increased from baseline to the fourth minute of inflation (median ∆R=0.81 ms; p 0.05), whereas δ was mostly unaltered during the intervention ( δ=0 ms).

CONCLUSION

We reported, for the first time, that TWA is significantly driven by the myocyte's spatial covariance between their repolarization times and alternans, and not by global alternans, when TWA is generated by regional ischemia.

SIGNIFICANCE

The metrics may reveal new complementary insights into the mechanisms underlying TWA.

Family Pediatrician and Public Health collaboration, an alliance to increase vaccination coverage: an experience with MenB vaccination in Italy

BACKGROUND

Invasive Meningococcal Disease is a severe disease mainly affecting infants and young children. Most infections are caused by serogroups A, B, C, W, X, and Y. In the last 10 years, serogroup B has been the main cause of Invasive Meningococcal Disease in Europe. Recent data resulting from an observational study conducted in Italy show a significant reduction in the number of Invasive Meningococcal Disease cases due to Neisseria meningitidis B after the introduction of vaccine 4CMenB. Thus, the Naples Team of Federation of Italian Primary Care Pediatricians and the Public Health Department started an active collaboration focused on vaccination process management (named "Progetto Via") with the aim of increasing Meningococcal B vaccination coverage.

STUDY DESIGN

Source of data is the regional platform "GE.VA.". Every Primary care Pediatrician uses daily to record vaccination activity. This platform is integrated with data entered by operators of the District/Vaccination Center.

METHODS

Time: January 2019 - December 2019. The Federation of Italian Primary Care Pediatricians/Naples organized a meeting to identify six coordinators. The pediatricians could choose to counsel in their own offices and send children to the vaccination center or to counsel and vaccinate directly in their own clinics.

RESULTS

A total of 78 pediatricians took part in the project: 46 did only counseling and 32 did both counseling and vaccination in their medical clinic. Data obtained show an overall average vaccination coverage growth of about 13% in the first 4 months of the survey, and a further growth of about 11% in the following seven months, with a total growth in the entire period of 24%. The pediatricians' counseling is essential to recover non-compliant subjects, considering both the relationship of trust with the families and the visits already scheduled as an ideal moment for vaccinations' status check.

CONCLUSIONS

The project highlights how an effective collaboration between family pediatricians and the Local Health Authority becomes valuable in getting closer to reach the Ministerial goal of 95%. Vaccination coverage increased significantly when family pediatricians supported the activity of vaccine centers in distress in many regional situations. The trust relationship, the hourly availability and the capillary network of family pediatricians' clinics were key elements for the success of this project and were also recognized by parents.

Machine Learning Using a Single-Lead ECG to Identify Patients With Atrial Fibrillation-Induced Heart Failure

Aims

Atrial fibrillation (AF) and heart failure often co-exist. Early identification of AF patients at risk for AF-induced heart failure (AF-HF) is desirable to reduce both morbidity and mortality as well as health care costs. We aimed to leverage the characteristics of beat-to-beat-patterns in AF to prospectively discriminate AF patients with and without AF-HF.

Methods

A dataset of 10,234 5-min length RR-interval time series derived from 26 AF-HF patients and 26 control patients was extracted from single-lead Holter-ECGs. A total of 14 features were extracted, and the most informative features were selected. Then, a decision tree classifier with 5-fold cross-validation was trained, validated, and tested on the dataset randomly split. The derived algorithm was then tested on 2,261 5-min segments from six AF-HF and six control patients and validated for various time segments.

Results

The algorithm based on the spectral entropy of the RR-intervals, the mean value of the relative RR-interval, and the root mean square of successive differences of the relative RR-interval yielded an accuracy of 73.5%, specificity of 91.4%, sensitivity of 64.7%, and PPV of 87.0% to correctly stratify segments to AF-HF. Considering the majority vote of the segments of each patient, 10/12 patients (83.33%) were correctly classified.

Conclusion

Beat-to-beat-analysis using a machine learning classifier identifies patients with AF-induced heart failure with clinically relevant diagnostic properties. Application of this algorithm in routine care may improve early identification of patients at risk for AF-induced cardiomyopathy and improve the yield of targeted clinical follow-up.

International perspective on integrated care models in child and adult mental health

The dearth of child and adolescent mental health services (CAMHS) is a global problem. Integrating CAMHS in primary care has been offered as a solution. We sampled integrated care perspectives from colleagues around the world. Our findings include various models of integrated care namely: the stepped care model in Australia; shared care in the United Kingdom (UK) and Spain; school-based collaborative care in Qatar, Singapore and the state of Texas in the US; collaborative care in Canada, Brazil, US, and Uruguay; coordinated care in the US; and, developing collaborative care models in low-resource settings, like Kenya and Micronesia. These findings provide insights into training initiatives necessary to build CAMHS workforce capacity using integrated care models, each with the ultimate goal of improving access to care. Despite variations and progress in implementing integrated care models internationally, common challenges exist: funding within complex healthcare systems, limited training mechanisms, and geopolitical/policy issues. Supportive healthcare policy, robust training initiatives, ongoing quality improvement and measurement of outcomes across programs would provide data-driven support for the expansion of integrated care and ensure its sustainability.

Hybrid machine learning to localize atrial flutter substrates using the surface 12-lead electrocardiogram

Aims

Atrial flutter (AFlut) is a common re-entrant atrial tachycardia driven by self-sustainable mechanisms that cause excitations to propagate along pathways different from sinus rhythm. Intra-cardiac electrophysiological mapping and catheter ablation are often performed without detailed prior knowledge of the mechanism perpetuating AFlut, likely prolonging the procedure time of these invasive interventions. We sought to discriminate the AFlut location [cavotricuspid isthmus-dependent (CTI), peri-mitral, and other left atrium (LA) AFlut classes] with a machine learning-based algorithm using only the non-invasive signals from the 12-lead electrocardiogram (ECG).

Methods and results

Hybrid 12-lead ECG dataset of 1769 signals was used (1424 in silico ECGs, and 345 clinical ECGs from 115 patients—three different ECG segments over time were extracted from each patient corresponding to single AFlut cycles). Seventy-seven features were extracted. A decision tree classifier with a hold-out classification approach was trained, validated, and tested on the dataset randomly split after selecting the most informative features. The clinical test set comprised 38 patients (114 clinical ECGs). The classifier yielded 76.3% accuracy on the clinical test set with a sensitivity of 89.7%, 75.0%, and 64.1% and a positive predictive value of 71.4%, 75.0%, and 86.2% for CTI, peri-mitral, and other LA class, respectively. Considering majority vote of the three segments taken from each patient, the CTI class was correctly classified at 92%.

Conclusion

Our results show that a machine learning classifier relying only on non-invasive signals can potentially identify the location of AFlut mechanisms. This method could aid in planning and tailoring patient-specific AFlut treatments.

Opening the black box: interpretability of machine learning algorithms in electrocardiography

Recent studies have suggested that cardiac abnormalities can be detected from the electrocardiogram (ECG) using deep machine learning (DL) models. However, most DL algorithms lack interpretability, since they do not provide any justification for their decisions. In this study, we designed two new frameworks to interpret the classification results of DL algorithms trained for 12-lead ECG classification. The frameworks allow us to highlight not only the ECG samples that contributed most to the classification, but also which between the P-wave, QRS complex and T-wave, hereafter simply called 'waves', were the most relevant for the diagnosis. The frameworks were designed to be compatible with any DL model, including the ones already trained. The frameworks were tested on a selected Deep Neural Network, trained on a publicly available dataset, to automatically classify 24 cardiac abnormalities from 12-lead ECG signals. Experimental results showed that the frameworks were able to detect the most relevant ECG waves contributing to the classification. Often the network relied on portions of the ECG which are also considered by cardiologists to detect the same cardiac abnormalities, but this was not always the case. In conclusion, the proposed frameworks may unveil whether the network relies on features which are clinically significant for the detection of cardiac abnormalities from 12-lead ECG signals, thus increasing the trust in the DL models. This article is part of the theme issue 'Advanced computation in cardiovascular physiology: new challenges and opportunities'.

Relationship Between Deceleration Morphology and Phase Rectified Signal Averaging-Based Parameters During Labor

During labor, uterine contractions trigger the response of the autonomic nervous system (ANS) of the fetus, producing sawtooth-like decelerations in the fetal heart rate (FHR) series. Under chronic hypoxia, ANS is known to regulate FHR differently with respect to healthy fetuses. In this study, we hypothesized that such different ANS regulation might also lead to a change in the FHR deceleration morphology. The hypothesis was tested in an animal model comprising nine normoxic and five chronically hypoxic fetuses that underwent a protocol of umbilical cord occlusions (UCOs). Deceleration morphologies in the fetal inter-beat time interval (FRR) series were modeled using a trapezoid with four parameters, i.e., baseline b, deceleration depth a, UCO response time τ u and recovery time τ r . Comparing normoxic and hypoxic sheep, we found a clear difference for τ u (24.8±9.4 vs. 39.8±9.7 s; p < 0.05), a (268.1±109.5 vs. 373.0±46.0 ms; p < 0.1) and Δτ = τ u - τ r (13.2±6.9 vs. 23.9±7.5 s; p < 0.05). Therefore, the animal model supported the hypothesis that hypoxic fetuses have a longer response time τ u and larger asymmetry Δτ as a response to UCOs. Assessing these morphological parameters during labor is challenging due to non-stationarity, phase desynchronization and noise. For this reason, in the second part of the study, we quantified whether acceleration capacity (AC), deceleration capacity (DC), and deceleration reserve (DR), computed through Phase-Rectified Signal Averaging (PRSA, known to be robust to noise), were correlated with the morphological parameters. DC, AC and DR were correlated with τ u , τ r and Δτ for a wide range of the PRSA parameter T (Pearson's correlation ρ > 0.8, p < 0.05). In conclusion, deceleration morphologies have been found to differ between normoxic and hypoxic sheep fetuses during UCOs. The same difference can be assessed through PRSA based parameters, further motivating future investigations on the translational potential of this methodology on human data.

Directed Network Mapping Approach to Rotor Localization in Atrial Fibrillation Simulation

Catheter ablation for atrial fibrillation (AF) is one of the most commonly performed electrophysiology procedures. Despite significant advances in our understanding of AF mechanisms in the last years, ablation outcomes remain suboptimal for many patients, particularly those with persistent or long-standing AF. A possible reason is that ablation techniques mainly focus on anatomic, rather than patient-specific functional targets for ablation. The identification of such ablation targets remains challenging. The purpose of this study is to investigate a novel approach based on directed networks, which allow the automatic detection of important arrhythmia mechanisms, that can be convenient for guiding the ablation strategy. The networks are generated by processing unipolar electrograms (EGMs) collected by the catheters positioned at the different regions of the atria. Network vertices represent the locations of the recordings and edges are determined using cross-covariance time-delay estimation method. The algorithm identifies rotational activity, spreading from vertex to vertex creating a cycle. This work is a simulation study and it uses a highly detailed computational 3D model of human atria in which sustained rotor activation of the atria was achieved. Virtual electrodes were placed on the endocardial surface, and EGMs were calculated at each of these electrodes. The propagation of the electric wave fronts in the atrial myocardium during AF is very complex, so in order to properly capture wave propagation patterns, we split EGMs into multiple short time frames. Then, a specific network for each of these time frames was generated, and the cycles repeating in consecutive networks point us to the stable rotor's location. The respective atrial voltage map served as reference. By detecting a cycle between the same 3 nodes in 19 out of 58 networks, where 10 of these networks were in consecutive time frames, a stable rotor was successfully located.

Atrial Flutter Mechanism Detection Using Directed Network Mapping

Atrial flutter (AFL) is a common atrial arrhythmia typically characterized by electrical activity propagating around specific anatomical regions. It is usually treated with catheter ablation. However, the identification of rotational activities is not straightforward, and requires an intense effort during the first phase of the electrophysiological (EP) study, i.e., the mapping phase, in which an anatomical 3D model is built and electrograms (EGMs) are recorded. In this study, we modeled the electrical propagation pattern of AFL (measured during mapping) using network theory (NT), a well-known field of research from the computer science domain. The main advantage of NT is the large number of available algorithms that can efficiently analyze the network. Using directed network mapping, we employed a cycle-finding algorithm to detect all cycles in the network, resembling the main propagation pattern of AFL. The method was tested on two subjects in sinus rhythm, six in an experimental model of in-silico simulations, and 10 subjects diagnosed with AFL who underwent a catheter ablation. The algorithm correctly detected the electrical propagation of both sinus rhythm cases and in-silico simulations. Regarding the AFL cases, arrhythmia mechanisms were either totally or partially identified in most of the cases (8 out of 10), i.e., cycles around the mitral valve, tricuspid valve and figure-of-eight reentries. The other two cases presented a poor mapping quality or a major complexity related to previous ablations, large areas of fibrotic tissue, etc. Directed network mapping represents an innovative tool that showed promising results in identifying AFL mechanisms in an automatic fashion. Further investigations are needed to assess the reliability of the method in different clinical scenarios.