LUNet: deep learning for the segmentation of arterioles and venules in high resolution fundus images.

Objective.This study aims to automate the segmentation of retinal arterioles and venules (A/V) from digital fundus images (DFI), as changes in the spatial distribution of retinal microvasculature are indicative of cardiovascular diseases, positioning the eyes as windows to cardiovascular health.Approach.We utilized active learning to create a new DFI dataset with 240 crowd-sourced manual A/V segmentations performed by 15 medical students and reviewed by an ophthalmologist. We then developed LUNet, a novel deep learning architecture optimized for high-resolution A/V segmentation. The LUNet model features a double dilated convolutional block to widen the receptive field and reduce parameter count, alongside a high-resolution tail to refine segmentation details. A custom loss function was designed to prioritize the continuity of blood vessel segmentation.Main Results.LUNet significantly outperformed three benchmark A/V segmentation algorithms both on a local test set and on four external test sets that simulated variations in ethnicity, comorbidities and annotators.Significance.The release of the new datasets and the LUNet model (www.aimlab-technion.com/lirot-ai) provides a valuable resource for the advancement of retinal microvasculature analysis. The improvements in A/V segmentation accuracy highlight LUNet's potential as a robust tool for diagnosing and understanding cardiovascular diseases through retinal imaging.

pyPPG: a Python toolbox for comprehensive photoplethysmography signal analysis.

Objective.Photoplethysmography is a non-invasive optical technique that measures changes in blood volume within tissues. It is commonly and being increasingly used for a variety of research and clinical applications to assess vascular dynamics and physiological parameters. Yet, contrary to heart rate variability measures, a field which has seen the development of stable standards and advanced toolboxes and software, no such standards and limited open tools exist for continuous photoplethysmogram (PPG) analysis. Consequently, the primary objective of this research was to identify, standardize, implement and validate key digital PPG biomarkers.Approach.This work describes the creation of a standard Python toolbox, denotedpyPPG, for long-term continuous PPG time-series analysis and demonstrates the detection and computation of a high number of fiducial points and digital biomarkers using a standard fingerbased transmission pulse oximeter.Main results.The improved PPG peak detector had an F1-score of 88.19% for the state-of-the-art benchmark when evaluated on 2054 adult polysomnography recordings totaling over 91 million reference beats. The algorithm outperformed the open-source original Matlab implementation by ∼5% when benchmarked on a subset of 100 randomly selected MESA recordings. More than 3000 fiducial points were manually annotated by two annotators in order to validate the fiducial points detector. The detector consistently demonstrated high performance, with a mean absolute error of less than 10 ms for all fiducial points.Significance.Based on these fiducial points,pyPPGengineered a set of 74 PPG biomarkers. Studying PPG time-series variability usingpyPPGcan enhance our understanding of the manifestations and etiology of diseases. This toolbox can also be used for biomarker engineering in training data-driven models.pyPPGis available onhttps://physiozoo.com/.

Leuven-Haifa High-Resolution Fundus Image Dataset for Retinal Blood Vessel Segmentation and Glaucoma Diagnosis.

The Leuven-Haifa dataset contains 240 disc-centered fundus images of 224 unique patients (75 patients with normal tension glaucoma, 63 patients with high tension glaucoma, 30 patients with other eye diseases and 56 healthy controls) from the University Hospitals of Leuven. The arterioles and venules of these images were both annotated by master students in medicine and corrected by a senior annotator. All senior segmentation corrections are provided as well as the junior segmentations of the test set. An open-source toolbox for the parametrization of segmentations was developed. Diagnosis, age, sex, vascular parameters as well as a quality score are provided as metadata. Potential reuse is envisioned as the development or external validation of blood vessels segmentation algorithms or study of the vasculature in glaucoma and the development of glaucoma diagnosis algorithms. The dataset is available on the KU Leuven Research Data Repository (RDR).

A Review on Atrial Fibrillation Detection From Ambulatory ECG.

Atrial fibrillation (AF) is a prevalent clinical arrhythmia disease and is an important cause of stroke, heart failure, and sudden death. Due to the insidious onset and no obvious clinical symptoms of AF, the status of AF diagnosis and treatment is not optimal. Early AF screening or detection is essential. Internet of Things (IoT) and artificial intelligence (AI) technologies have driven the development of wearable electrocardiograph (ECG) devices used for health monitoring, which are an effective means of AF detection. The main challenges of AF analysis using ambulatory ECG include ECG signal quality assessment to select available ECG, the robust and accurate detection of QRS complex waves to monitor heart rate, and AF identification under the interference of abnormal ECG rhythm. Through ambulatory ECG measurement and intelligent detection technology, the probability of postoperative recurrence of AF can be reduced, and personalized treatment and management of patients with AF can be realized. This work describes the status of AF monitoring technology in terms of devices, algorithms, clinical applications, and future directions.

Using beat-to-beat heart signals for age-independent biometric verification.

Use of non-stationary physiological signals for biometric verification, reduces the ability to forge. Such signals should be simple to acquire with inexpensive equipment. The beat-to-beat information embedded within the time intervals between consecutive heart beats is a non-stationary physiological signal; its potential for biometric verification has not been studied. This work introduces a biometric verification method termed "CompaRR". Heartbeat was extracted from longitudinal recordings from 30 mice ranging from 6 to 24 months of age (equivalent to ~ 20-75 human years). Fifty heartbeats, which is close to resting human heartbeats in a minute, were sufficient for the verification task, achieving a minimal equal error rate of 0.21. When trained on 6-month-old mice and tested on unseen mice up to 18-months of age (equivalent to ~ 50 human years), no significant change in the verification performance was noted. Finally, when the model was trained on data from drug-treated mice, verification was still possible.

End-to-end risk prediction of atrial fibrillation from the 12-Lead ECG by deep neural networks.

BACKGROUND: Atrial fibrillation (AF) is one of the most common cardiac arrhythmias that affects millions of people each year worldwide and it is closely linked to increased risk of cardiovas- cular diseases such as stroke and heart failure. Machine learning methods have shown promising results in evaluating the risk of developing atrial fibrillation from the electrocardiogram. We aim to develop and evaluate one such algorithm on a large CODE dataset collected in Brazil. METHODS: We used the CODE cohort to develop and test a model for AF risk prediction for individual patients from the raw ECG recordings without the use of additional digital biomarkers. The cohort is a collection of ECG recordings and annotations by the Telehealth Network of Minas Gerais, in Brazil. A convolutional neural network based on a residual network architecture was implemented to produce class probabilities for the classification of AF. The probabilities were used to develop a Cox proportional hazards model and a Kaplan-Meier model to carry out survival analysis. Hence, our model is able to perform risk prediction for the development of AF in patients without the condition. RESULTS: The deep neural network model identified patients without indication of AF in the presented ECG but who will develop AF in the future with an AUC score of 0.845. From our survival model, we obtain that patients in the high-risk group (i.e. with the probability of a future AF case being >0.7) are 50% more likely to develop AF within 40 weeks, while patients belonging to the minimal-risk group (i.e. with the probability of a future AF case being less than or equal to 0.1) have >85% chance of remaining AF free up until after seven years. CONCLUSION: We developed and validated a model for AF risk prediction. If applied in clinical practice, the model possesses the potential of providing valuable and useful information in decision- making and patient management processes.

Deep learning for obstructive sleep apnea diagnosis based on single channel oximetry.

Obstructive sleep apnea (OSA) is a serious medical condition with a high prevalence, although diagnosis remains a challenge. Existing home sleep tests may provide acceptable diagnosis performance but have shown several limitations. In this retrospective study, we used 12,923 polysomnography recordings from six independent databases to develop and evaluate a deep learning model, called OxiNet, for the estimation of the apnea-hypopnea index from the oximetry signal. We evaluated OxiNet performance across ethnicity, age, sex, and comorbidity. OxiNet missed 0.2% of all test set moderate-to-severe OSA patients against 21% for the best benchmark.

Robust peak detection for photoplethysmography signal analysis.

Efficient and accurate evaluation of long-term photoplethysmography (PPG) recordings is essential for both clinical assessments and consumer products. In 2021, the top opensource peak detectors were benchmarked on the Multi-Ethnic Study of Atherosclerosis (MESA) database consisting of polysomnography (PSG) recordings and continuous sleep PPG data, where the Automatic Beat Detector (Aboy) had the best accuracy. This work presents Aboy++, an improved version of the original Aboy beat detector. The algorithm was evaluated on 100 adult PPG recordings from the MESA database, which contains more than 4.25 million reference beats. Aboy++ achieved an F1-score of 85.5%, compared to 80.99% for the original Aboy peak detector. On average, Aboy++ processed a 1 hour-long recording in less than 2 seconds. This is compared to 115 seconds (i.e., over 57-times longer) for the open-source implementation of the original Aboy peak detector. This study demonstrated the importance of developing robust algorithms like Aboy++ to improve PPG data analysis and clinical outcomes. Overall, Aboy++ is a reliable tool for evaluating long-term wearable PPG measurements in clinical and consumer contexts. The open-source algorithm is available on the physiozoo.com website (on publication of this proceeding).

The 2023 wearable photoplethysmography roadmap.

Photoplethysmography is a key sensing technology which is used in wearable devices such as smartwatches and fitness trackers. Currently, photoplethysmography sensors are used to monitor physiological parameters including heart rate and heart rhythm, and to track activities like sleep and exercise. Yet, wearable photoplethysmography has potential to provide much more information on health and wellbeing, which could inform clinical decision making. This Roadmap outlines directions for research and development to realise the full potential of wearable photoplethysmography. Experts discuss key topics within the areas of sensor design, signal processing, clinical applications, and research directions. Their perspectives provide valuable guidance to researchers developing wearable photoplethysmography technology.

FundusQ-Net: A regression quality assessment deep learning algorithm for fundus images quality grading.

OBJECTIVE: Ophthalmological pathologies such as glaucoma, diabetic retinopathy and age-related macular degeneration are major causes of blindness and vision impairment. There is a need for novel decision support tools that can simplify and speed up the diagnosis of these pathologies. A key step in this process is to automatically estimate the quality of the fundus images to make sure these are interpretable by a human operator or a machine learning model. We present a novel fundus image quality scale and deep learning (DL) model that can estimate fundus image quality relative to this new scale. METHODS: A total of 1245 images were graded for quality by two ophthalmologists within the range 1-10, with a resolution of 0.5. A DL regression model was trained for fundus image quality assessment. The architecture used was Inception-V3. The model was developed using a total of 89,947 images from 6 databases, of which 1245 were labeled by the specialists and the remaining 88,702 images were used for pre-training and semi-supervised learning. The final DL model was evaluated on an internal test set (n=209) as well as an external test set (n=194). RESULTS: The final DL model, denoted FundusQ-Net, achieved a mean absolute error of 0.61 (0.54-0.68) on the internal test set. When evaluated as a binary classification model on the public DRIMDB database as an external test set the model obtained an accuracy of 99%. SIGNIFICANCE: the proposed algorithm provides a new robust tool for automated quality grading of fundus images.

On Merging Feature Engineering and Deep Learning for Diagnosis, Risk Prediction and Age Estimation Based on the 12-Lead ECG.

OBJECTIVE: Over the past few years, deep learning (DL) has been used extensively in research for 12-lead electrocardiogram (ECG) analysis. However, it is unclear whether the explicit or implicit claims made on DL superiority to the more classical feature engineering (FE) approaches, based on domain knowledge, hold. In addition, it remains unclear whether combining DL with FE may improve performance over a single modality. METHODS: To address these research gaps and in-line with recent major experiments, we revisited three tasks: cardiac arrhythmia diagnosis (multiclass-multilabel classification), atrial fibrillation risk prediction (binary classification), and age estimation (regression). We used an overall dataset of 2.3M 12-lead ECG recordings to train the following models for each task: i) a random forest taking FE as input; ii) an end-to-end DL model; and iii) a merged model of FE+DL. RESULTS: FE yielded comparable results to DL while necessitating significantly less data for the two classification tasks. DL outperformed FE for the regression task. For all tasks, merging FE with DL did not improve performance over DL alone. These findings were confirmed on the additional PTB-XL dataset. CONCLUSION: We found that for traditional 12-lead ECG based diagnosis tasks, DL did not yield a meaningful improvement over FE, while it improved significantly the nontraditional regression task. We also found that combining FE with DL did not improve over DL alone, which suggests that the FE was redundant with the features learned by DL. SIGNIFICANCE: Our findings provides important recommendations on 12-lead ECG based machine learning strategy and data regime to choose for a given task. When looking at maximizing performance as the end goal, if the task is nontraditional and a large dataset is available then DL is preferable. If the task is a classical one and/or a small dataset is available then a FE approach may be the better choice.

Positional sleep apnea phenotyping using machine learning and digital oximetry biomarkers.

Study Objectives. To examine the feasibility of using digital oximetry biomarkers (OBMs) and body position to identify positional obstructive sleep apnea (POSA) phenotypes.Methods. A multiclass extreme gradient boost (XGBoost) was implemented to classify between three POSA phenotypes, i.e., positional patients (PP), including supine-predominant OSA (spOSA), and supine-isolated OSA (siOSA), and non-positional patients (NPP). A total of 861 individuals with OSA from the multi ethnic study of atherosclerosis (MESA) dataset were included in the study. Overall, 43 OBMs were computed for supine and non-supine positions and used as input features together with demographic and clinical information (META). Feature selection, using mRMR, was implemented, and nested cross validation was used for the model's performance evaluation.Results. The best performance for the multiclass classification yielded a median weighted F1 of 0.79 with interquartile range (IQR) of 0.06. Binary classification between PP to NPP achieved weighted F1 of 0.87 (0.04).Conclusion. Using OBMs computed in PP and NPP with OSA, it is possible to distinguish between the different phenotypes of POSA. This data-driven algorithm may be embedded in portable home sleep tests.

Liquid Biopsy-Based Volatile Organic Compounds from Blood and Urine and Their Combined Data Sets for Highly Accurate Detection of Cancer.

Liquid biopsy is seen as a prospective tool for cancer screening and tracking. However, the difficulty lies in effectively sieving, isolating, and overseeing cancer biomarkers from the backdrop of multiple disrupting cells and substances. The current study reports on the ability to perform liquid biopsy without the need to physically filter and/or isolate the cancer cells per se. This has been achieved through the detection and classification of volatile organic compounds (VOCs) emitted from the cancer cells found in the headspace of blood or urine samples or a combined data set of both. Spectrometric analysis shows that blood and urine contain complementary or overlapping VOC information on kidney cancer, gastric cancer, lung cancer, and fibrogastroscopy subjects. Based on this information, a nanomaterial-based chemical sensor array in conjugation with machine learning as well as data fusion of the signals achieved was carried out on various body fluids to assess the VOC profiles of cancer. The detection of VOC patterns by either Gas Chromatography-Mass Spectrometry (GC-MS) analysis or our sensor array achieved >90% accuracy, >80% sensitivity, and >80% specificity in different binary classification tasks. The hybrid approach, namely, analyzing the VOC datasets of blood and urine together, contributes an additional discrimination ability to the improvement (>3%) of the model's accuracy. The contribution of the hybrid approach for an additional discrimination ability to the improvement of the model's accuracy is examined and reported.

Generalizable and robust deep learning algorithm for atrial fibrillation diagnosis across geography, ages and sexes.

To drive health innovation that meets the needs of all and democratize healthcare, there is a need to assess the generalization performance of deep learning (DL) algorithms across various distribution shifts to ensure that these algorithms are robust. This retrospective study is, to the best of our knowledge, an original attempt to develop and assess the generalization performance of a DL model for AF events detection from long term beat-to-beat intervals across geography, ages and sexes. The new recurrent DL model, denoted ArNet2, is developed on a large retrospective dataset of 2,147 patients totaling 51,386 h obtained from continuous electrocardiogram (ECG). The model's generalization is evaluated on manually annotated test sets from four centers (USA, Israel, Japan and China) totaling 402 patients. The model is further validated on a retrospective dataset of 1,825 consecutives Holter recordings from Israel. The model outperforms benchmark state-of-the-art models and generalized well across geography, ages and sexes. For the task of event detection ArNet2 performance was higher for female than male, higher for young adults (less than 61 years old) than other age groups and across geography. Finally, ArNet2 shows better performance for the test sets from the USA and China. The main finding explaining these variations is an impairment in performance in groups with a higher prevalence of atrial flutter (AFL). Our findings on the relative performance of ArNet2 across groups may have clinical implications on the choice of the preferred AF examination method to use relative to the group of interest.

Descriptive characteristics of continuous oximetry measurement in moderate to severe covid-19 patients.

Non-invasive oxygen saturation (SpO2) is a central vital sign used to shape the management of COVID-19 patients. Yet, there have been no report quantitatively describing SpO2 dynamics and patterns in COVID-19 patients using continuous SpO2 recordings. We performed a retrospective observational analysis of the clinical information and 27 K hours of continuous SpO2 high-resolution (1 Hz) recordings of 367 critical and non-critical COVID-19 patients hospitalised at the Rambam Health Care Campus, Haifa, Israel. An absolute SpO2 threshold of 93% most efficiently discriminated between critical and non-critical patients, regardless of oxygen support. Oximetry-derived digital biomarker (OBMs) computed per 1 h monitoring window showed significant differences between groups, notably the cumulative time below 93% SpO2 (CT93). Patients with CT93 above 60% during the first hour of monitoring, were more likely to require oxygen support. Mechanical ventilation exhibited a strong effect on SpO2 dynamics by significantly reducing the frequency and depth of desaturations. OBMs related to periodicity and hypoxic burden were markedly affected, up to several hours before the initiation of the mechanical ventilation. In summary, OBMs, traditionally used in the field of sleep medicine research, are informative for continuous assessment of disease severity and response to respiratory support of hospitalised COVID-19 patients. In conclusion, OBMs may improve risk stratification and therapy management of critical care patients with respiratory impairment.

SleepPPG-Net: A Deep Learning Algorithm for Robust Sleep Staging From Continuous Photoplethysmography.

Sleep staging is an essential component in the diagnosis of sleep disorders and management of sleep health. Sleep is traditionally measured in a clinical setting and requires a labor-intensive labeling process. We hypothesize that it is possible to perform automated robust 4-class sleep staging using the raw photoplethysmography (PPG) time series and modern advances in deep learning (DL). We used two publicly available sleep databases that included raw PPG recordings, totalling 2,374 patients and 23,055 hours of continuous data. We developed SleepPPG-Net, a DL model for 4-class sleep staging from the raw PPG time series. SleepPPG-Net was trained end-to-end and consists of a residual convolutional network for automatic feature extraction and a temporal convolutional network to capture long-range contextual information. We benchmarked the performance of SleepPPG-Net against models based on the best-reported state-of-the-art (SOTA) algorithms. When benchmarked on a held-out test set, SleepPPG-Net obtained a median Cohen's Kappa ( κ) score of 0.75 against 0.69 for the best SOTA approach. SleepPPG-Net showed good generalization performance to an external database, obtaining a κ score of 0.74 after transfer learning. Overall, SleepPPG-Net provides new SOTA performance. In addition, performance is high enough to open the path to the development of wearables that meet the requirements for usage in clinical applications such as the diagnosis and monitoring of obstructive sleep apnea.

Characteristics of patients with positional OSA according to ethnicity and the identification of a novel phenotype-lateral positional patients: a Multi-Ethnic Study of Atherosclerosis (MESA) study.

STUDY OBJECTIVES: We investigated the characteristics of obstructive sleep apnea (OSA) positional patients' (PP) phenotypes among different ethnic groups in the Multi-Ethnic Study of Atherosclerosis (MESA) dataset. Moreover, we hypothesized the existence of a new OSA PP phenotype we coined "Lateral PP," for whom the lateral apnea-hypopnea index is at least double the supine apnea-hypopnea index. METHODS: From 2,273 adults with sleep information, we analyzed data of 1,323 participants who slept more than 4 hours and had at least 30 minutes of sleep in both the supine and the nonsupine positions. Demographics and clinical information were compared for the different PP and ethnic groups. RESULTS: 861 (65.1%) patients had OSA, and 35 (4.1%) were Lateral PP. Lateral PP patients were mainly females (62.9%), obese (median body mass index: 31.4 kg/m2), had mild-moderate OSA (94.3%), and mostly were non-Chinese American (97.1%). Among all patients with OSA, 550 (63.9%) were Supine PP and 17.7% were supine-isolated OSA. Supine PP and Lateral PP were present in 73.1% and 1.0% of Chinese Americans, 61.0% and 3.4% of Hispanics, 68.3% and 4.7% of White/Caucasian, and 56.2% and 5.2% of Black/African-American patients with OSA. CONCLUSIONS: Chinese Americans have the highest prevalence of Supine PP, whereas Black/African-American patients lean toward less Supine PP and higher Lateral PP. Lateral PP appears to be a novel OSA phenotype. However, Lateral PP was observed in a small group of patients with OSA and thus its existence should be further validated. CITATION: Ben Sason Y, Oksenberg A, Sobel JA, Behar JA. Characteristics of patients with positional OSA according to ethnicity and the identification of a novel phenotype-lateral positional patients: a Multi-Ethnic Study of Atherosclerosis (MESA) study. J Clin Sleep Med. 2023;19(3):529-538.

Machine learning to support triage of children at risk for epileptic seizures in the pediatric intensive care unit.

Objective.Epileptic seizures are relatively common in critically-ill children admitted to the pediatric intensive care unit (PICU) and thus serve as an important target for identification and treatment. Most of these seizures have no discernible clinical manifestation but still have a significant impact on morbidity and mortality. Children that are deemed at risk for seizures within the PICU are monitored using continuous-electroencephalogram (cEEG). cEEG monitoring cost is considerable and as the number of available machines is always limited, clinicians need to resort to triaging patients according to perceived risk in order to allocate resources. This research aims to develop a computer aided tool to improve seizures risk assessment in critically-ill children, using an ubiquitously recorded signal in the PICU, namely the electrocardiogram (ECG).Approach.A novel data-driven model was developed at a patient-level approach, based on features extracted from the first hour of ECG recording and the clinical data of the patient.Main results.The most predictive features were the age of the patient, the brain injury as coma etiology and the QRS area. For patients without any prior clinical data, using one hour of ECG recording, the classification performance of the random forest classifier reached an area under the receiver operating characteristic curve (AUROC) score of 0.84. When combining ECG features with the patients clinical history, the AUROC reached 0.87.Significance.Taking a real clinical scenario, we estimated that our clinical decision support triage tool can improve the positive predictive value by more than 59% over the clinical standard.

Detecting beats in the photoplethysmogram: benchmarking open-source algorithms.

The photoplethysmogram (PPG) signal is widely used in pulse oximeters and smartwatches. A fundamental step in analysing the PPG is the detection of heartbeats. Several PPG beat detection algorithms have been proposed, although it is not clear which performs best.Objective:This study aimed to: (i) develop a framework with which to design and test PPG beat detectors; (ii) assess the performance of PPG beat detectors in different use cases; and (iii) investigate how their performance is affected by patient demographics and physiology.Approach:Fifteen beat detectors were assessed against electrocardiogram-derived heartbeats using data from eight datasets. Performance was assessed using theF1score, which combines sensitivity and positive predictive value.Main results:Eight beat detectors performed well in the absence of movement withF1scores of ≥90% on hospital data and wearable data collected at rest. Their performance was poorer during exercise withF1scores of 55%-91%; poorer in neonates than adults withF1scores of 84%-96% in neonates compared to 98%-99% in adults; and poorer in atrial fibrillation (AF) withF1scores of 92%-97% in AF compared to 99%-100% in normal sinus rhythm.Significance:Two PPG beat detectors denoted 'MSPTD' and 'qppg' performed best, with complementary performance characteristics. This evidence can be used to inform the choice of PPG beat detector algorithm. The algorithms, datasets, and assessment framework are freely available.

Effect of temporal resolution on the detection of cardiac arrhythmias using HRV features and machine learning.

Objective.Arrhythmia is an abnormal cardiac rhythm that affects the pattern and rate of the heartbeat. Wearable devices with the functionality to measure and store heart rate (HR) data are growing in popularity and enable diagnosing and monitoring arrhythmia on a large scale. The typical sampling resolution of HR data available from non-medical grade wearable devices varies from seconds to several minutes depending on the device and its settings. However, the impact of sampling resolution on the performance and quality of arrhythmia detection has not yet been quantified.Approach.In this study, we investigated the detection and classification of three arrhythmias, namely atrial fibrillation, bradycardia, tachycardia, from down-sampled HR data with various temporal resolution (5-, 15-, 30- and 60 s averages) in 1 h segments extracted from an annotated Holter ECG database acquired at the University of Virginia Heart Station. For the classification task, a total of 15 common heart rate variability (HRV) features were engineered based on the HR time series of each patient. Three different types of machine learning classifiers were evaluated, namely logistic regression, support vector machine and random forest.Main results.A decrease in temporal resolution drastically impacted the detection of atrial fibrillation but did not substantially affect the detection of bradycardia and tachycardia. A HR resolution up to 15 s average demonstrated reasonable performance with a sensitivity of 0.92 and a specificity of 0.86 for a multiclass random forest classifier.Significance.HRV features extracted from low resolution long HR recordings have the potential to increase the early detection of arrhythmias in undiagnosed individuals.