Accurate monitoring of photoplethysmographic (PPG) signals is essential for deriving physiological metrics such as heart rate, pulse rate variability, and blood pressure. However, unreliable signal segments can severely distort these estimations. This study presents a signal quality assessment (SQA) framework to classify PPG recordings as high or low quality. The dataset, collected from 50 volunteers using the Polar Verity Sense device, contained 7,402 ten-second segments labeled by experts. Forty-eight temporal, spectral, and morphological features were extracted, and several classifiers were tested. The boosted decision tree achieved the best performance, reaching 98

Pulse-rate variability (PRV) offers a compelling alternative for estimating blood pressure (BP). However, ongoing debates persist regarding its suitability for BP monitoring and resemblance to HRV, with room for improvement in existing PRV studies We recruited and classified five-minute electrocardiography (ECG) and photoplethysmography (PPG) recordings from 202 patients acquired from the MIMIC-II database into three categories: normotensive (NT), prehypertensive (PHT) and hypertensive (HT). We conducted PRV and pulse rate asymmetry (PRA) analyses using various time-domain, frequency- domain and non-linear indices. We performed three database splits (1. NT/PHT/HT, 2. NT/non-NT and 3. HT/non-HT) to test our models. To validate our findings, we compared them to heart rate variability (HRV) using Bland-Altman (BA) analysis and correlation. We employed multi-class (MCC) and single- class classification with a 10-fold cross-validation approach, reserving a 20

Hypertension (HT) can lead to severe health complications. Therefore, early HT detection is of paramount importance. Photoplethysmography (PPG) stands as a promising solution for blood pressure (BP) monitoring, showing satisfactory but not optimal results. Simultaneously, deep learning (DL) has been hitherto able to improve performance in medical research. The present study aims to harness these powerful tools to assist continuous BP monitoring and detect early signs of HT. 171 five-minute PPG recordings from the MIMIC database were acquired and classified into three categories according to their BP: normotensive (NT), prehypertensive (PHT) and HT. Signals were extracted with a sampling frequency of 125 Hz, resampled to 250 Hz. Each recording was segmented into 2.5-s epochs. Signals were converted to 299 × 299 recurrence plot (RP) images, using m = 3 and τ = 6 ms. Models were trained with the Inception-ResNet-v2 network, using an 80 − 10[

Health-tracking from photoplethysmography (PPG) signals is significantly hindered by motion artifacts (MAs). Although many algorithms exist to detect MAs, the corrupted signal often remains unexploited. This work introduces a novel method able to reconstruct noisy PPGs and facilitate uninterrupted health monitoring. The algorithm starts with spectral-based MA detection, followed by signal reconstruction by using the morphological and heart-rate variability information from the clean segments adjacent to noise. The algorithm was tested on (a) 30 noisy PPGs of a maximum 20 s noise duration and (b) 28 originally clean PPGs, after noise addition (2–120 s) (1) with and (2) without cancellation of the corresponding clean segment. Sampling frequency was 250 Hz after resampling. Noise detection was evaluated by means of accuracy, sensitivity, and specificity. For the evaluation of signal reconstruction, the heart-rate (HR) was compared via Pearson correlation (PC) and absolute error (a) between ECGs and reconstructed PPGs and (b) between original and reconstructed PPGs. Bland-Altman (BA) analysis for the differences in HR estimation on original and reconstructed segments of (b) was also performed. Noise detection accuracy was 90.91% for (a) and 99.38–100% for (b). For the PPG reconstruction, HR showed 99.31% correlation in (a) and >90% for all noise lengths in (b). Mean absolute error was 1.59 bpm for (a) and 1.26–1.82 bpm for (b). BA analysis indicated that, in most cases, 90% or more of the recordings fall within the confidence interval, regardless of the noise length. Optimal performance is achieved even for signals of noise up to 2 min, allowing for the utilization and further analysis of recordings that would otherwise be discarded. Thereby, the algorithm can be implemented in monitoring devices, assisting in uninterrupted health-tracking.

BACKGROUND:Controlled donation after circulatory determination of death (cDCD) seems an effective way to mitigate the critical shortage of available organs for transplant worldwide. As a recently developed procedure for organ retrieval, some questions remain unsolved such as the uncertainty regarding the effect of functional warm ischemia time (FWIT) on organs´ viability.METHODS:We developed a multicenter prospective cohort study collecting all data from evaluated organs during cDCD from 2017 to 2020. All the procedures related to cDCD were performed with normothermic regional perfusion. The analysis included organ retrieval as endpoint and FWIT as exposure of interest. The effect of FWIT on the likelihood for organ retrieval was evaluated with Relative distribution analysis.RESULTS:A total amount of 507 organs´ related information was analyzed from 95 organ donors. Median donor age was 62 years, and 63% of donors were male. Stroke was the most common diagnosis before withdrawal of life-sustaining therapy (61%), followed by anoxic encephalopathy (21%). This analysis showed that length of FWIT was inversely associated with organ retrieval rates for liver, kidneys, and pancreas. No statistically significant association was found for lungs.CONCLUSIONS:Results showed an inverse association between functional warm ischemia time (FWIT) and retrieval rate. We also have postulated optimal FWIT's thresholds for organ retrieval. FWIT for liver retrieval remained between 6 and less than 11 min and in case of kidneys and pancreas, the optimal FWIT for retrieval was 6 to 12 min. These results could be valuable to improve organ utilization and for future analysis.

Despite the strong association between hypertension (HT) and heart-rate variability (HRV), the ability of HRV to detect HT cases under the coexistence of several pathologies remains unknown. The present study aims to optimize the HT detection using machine- learning (ML) techniques, in 202 5-minute ECG recordings from the MIMIC database. Recordings were classified by blood pressure (BP) into normotensive (NT), prehypertensive (PHT) and hypertensive (HT) with a cut-off BP of < 120/80 mm Hg, < 139/89 mm Hg and 2: 140/90 mm Hg, respectively. Time-, frequency-domain and Poincaré HRV features were explored. Multi-class (MC), between-class comparison (BC) and 1-vs-all analysis was performed. Single- (SF) and multifeature (MF) classification with 10-fold cross-validation and a 20% test set was also performed. Statistically significant differences (p < 0.022) were found in MC and BC for most HRV features. Differences were more prominent in HT group (p < 0.0003). The best MF accuracy for MC was 95% using 6 features. HT and NT detection accuracy was 87.5% and 95%, using 6 and 4 features, respectively, for 1-vs-all analysis. The proposed models can be easily implemented and achieve high classification accuracy. The results suggest the use of HRV to detect HT or HT-prone patients in diseased population.

Hypertension, a primary risk factor for various cardiovascular diseases, is a global health concern. Early identification and effective management of hypertensive individuals are vital for reducing associated health risks. This study explores the potential of deep learning (DL) techniques, specifically GoogLeNet, ResNet-18, and ResNet-50, for discriminating between normotensive (NTS) and hypertensive (HTS) individuals using photoplethysmographic (PPG) recordings. The research assesses the impact of calibration at different time intervals between measurements, considering intervals less than 1 h, 1-6 h, 6-24 h, and over 24 h. Results indicate that calibration is most effective when measurements are closely spaced, with an accuracy exceeding 90% in all the DL strategies tested. For calibration intervals below 1 h, ResNet-18 achieved the highest accuracy (93.32%), sensitivity (84.09%), specificity (97.30%), and F1-score (88.36%). As the time interval between calibration and test measurements increased, classification performance gradually declined. For intervals exceeding 6 h, accuracy dropped below 81% but with all models maintaining accuracy above 71% even for intervals above 24 h. This study provides valuable insights into the feasibility of using DL for hypertension risk assessment, particularly through PPG recordings. It demonstrates that closely spaced calibration measurements can lead to highly accurate classification, emphasizing the potential for real-time applications. These findings may pave the way for advanced, non-invasive, and continuous blood pressure monitoring methods that are both efficient and reliable.

Atrial fibrillation (AF) is a prevalent cardiac arrhythmia often treated concomitantly with other cardiac interventions through the Cox-Maze procedure. This highly invasive intervention is still linked to a long-term recurrence rate of approximately 35% in permanent AF patients. The aim of this study is to preoperatively predict long-term AF recurrence post-surgery through the analysis of atrial activity (AA) organization from non-invasive electrocardiographic (ECG) recordings. A dataset comprising ECGs from 53 patients with permanent AF who had undergone Cox-Maze concomitant surgery was analyzed. The AA was extracted from the lead V1 of these recordings and then characterized using novel predictors, such as the mean and standard deviation of the relative wavelet energy (RWEm and RWEs) across different scales, and an entropy-based metric that computes the stationary wavelet entropy variability (SWEnV). The individual predictors exhibited limited predictive capabilities to anticipate the outcome of the procedure, with the SWEnV yielding a classification accuracy (Acc) of 68.07%. However, the assessment of the RWEs for the seventh scale (RWEs7), which encompassed frequencies associated with the AA, stood out as the most promising individual predictor, with sensitivity (Se) and specificity (Sp) values of 80.83% and 67.09%, respectively, and an Acc of almost 75%. Diverse multivariate decision tree-based models were constructed for prediction, giving priority to simplicity in the interpretation of the forecasting methodology. In fact, the combination of the SWEnV and RWEs7 consistently outperformed the individual predictors and excelled in predicting post-surgery outcomes one year after the Cox-Maze procedure, with Se, Sp, and Acc values of approximately 80%, thus surpassing the results of previous studies based on anatomical predictors associated with atrial function or clinical data. These findings emphasize the crucial role of preoperative patient-specific ECG signal analysis in tailoring post-surgical care, enhancing clinical decision making, and improving long-term clinical outcomes.

Early detection of high blood pressure (BP) is of paramount relevance because hypertension is the main risk factor for many cardiovascular diseases. This work evaluates the need of per-subject calibration for discrimination between normotensive (NTS) and hypertensive (HTS) subjects. 668 electrocardiographic (ECG), photoplethysmographic (PPG) and BP recordings from 51 subjects were analyzed. After signal preprocessing and feature selection, 17 discriminatory features were obtained to train machine learning based classifiers. Previous persubject calibration relevance was evaluated by sequential validation, using both close and distant in time calibration measurements varying from less than 1 h to more than 24h with respect to test measurements. The k-nearest neighbors classifier provided an accuracy for new subjects before calibration of 56.79%. The inclusion of just one calibration measurement into the model improved classification accuracy by 30%, reaching gradually more than 97%. Classification accuracy decreased with distance to calibration, but remained well above 83% even days after the last calibration. Thus, discrimination of NTS and HTS subjects can be significantly improved combining PPG and ECG recordings with previous per-subject calibration and, therefore, could be used for the detection of hypertension implementing these techniques in wearable devices.

Local activation waves (LAWs) detection in complex fractionated atrial electrograms (CFAEs) during catheter ablation (CA) of atrial fibrillation (AF), the commonest cardiac arrhythmia, is a complicated task due to their extreme variability and heterogeneity in amplitude and morphology. There are few published works on reliable LAWs detectors, which are efficient for regular or low fractionated bipolar electrograms (EGMs) but lack satisfactory results when CFAEs are analyzed. The aim of the present work is the development of a novel optimized method for LAWs detection in CFAEs in order to assist cardiac mapping and catheter ablation (CA) guidance. The database consists of 119 bipolar EGMs classified by AF types according to Wells' classification. The proposed method introduces an alternative Botteron's preprocessing technique targeting the slow and small-ampitude activations. The lower band-pass filter cut-off frequency is modified to 20 Hz, and a hyperbolic tangent function is applied over CFAEs. Detection is firstly performed through an amplitude-based threshold and an escalating cycle-length (CL) analysis. Activation time is calculated at each LAW's barycenter. Analysis is applied in five-second overlapping segments. LAWs were manually annotated by two experts and compared with algorithm-annotated LAWs. AF types I and II showed 100% accuracy and sensitivity. AF type III showed 92.77% accuracy and 95.30% sensitivity. The results of this study highlight the efficiency of the developed method in precisely detecting LAWs in CFAEs. Hence, it could be implemented on real-time mapping devices and used during CA, providing robust detection results regardless of the fractionation degree of the analyzed recordings.

Coronary sinus (CS) catheterization is critical during catheter ablation (CA) of atrial fibrillation (AF). However, the association of CS electrical activity with atrial substrate modification has been barely investigated and mostly limited to analyses during AF. In sinus rhythm (SR), atrial substrate modification is principally assessed at a global level through P-wave analysis. Cross-correlating CS electrograms (EGMs) and P-waves’ features could potentiate the understanding of AF mechanisms. Five-minute surface lead II and bipolar CS recordings before, during, and after CA were acquired from 40 paroxysmal AF patients. Features related to duration, amplitude, and heart-rate variability of atrial activations were evaluated. Heart-rate adjustment (HRA) was applied. Correlations between each P-wave and CS local activation wave (LAW) feature were computed with cross-quadratic sample entropy (CQSE), Pearson correlation (PC), and linear regression (LR) with 10-fold cross-validation. The effect of CA between different ablation steps was compared with PC. Linear correlations: poor to mediocre before HRA for analysis at each P-wave/LAW (PC: max. +18.36%, p = 0.0017, LR: max. +5.33%, p = 0.0002) and comparison between two ablation steps (max. +54.07%, p = 0.0205). HRA significantly enhanced these relationships, especially in duration (P-wave/LAW: +43.82% to +69.91%, p < 0.0001 for PC and +18.97% to +47.25%, p < 0.0001 for LR, CA effect: +53.90% to +85.72%, p < 0.0210). CQSE reported negligent correlations (0.6–1.2). Direct analysis of CS features is unreliable to evaluate atrial substrate modification due to CA. HRA substantially solves this problem, potentiating correlation with P-wave features. Hence, its application is highly recommended.

Motion artifacts (MAs) is a major issue in photo-plethysmography (PPG), complicating health monitoring. While many algorithms focus on detecting MAs, little is known on how MA segments could be further utilized. This work proposes an algorithm for electrocardiograph-independent (ECG) PPGs re-construction. MAs on PPGs are detected by spectral analysis and HR on the clean segments is calculated. For MAs shorter than 20 seconds, reconstruction is performed by the characteristic pulse of the clean segments surrounding MAs, using the heart rate (HR) of the closest-to-MAs pulses and the HR and amplitude variability. Thirty eight-minute PPG and ECG recordings at 125 Hz sampling frequency of the BIDMC database were employed for validation, after upsampling to 250 Hz. HR was calculated in reconstructed PPGs and the corresponding ECGs and compared with Pearson correlation (PC). Mean absolute error (MAE) on HR estimation was also calculated. Pulse transit time (PTT) was computed for the reconstructed segment, as the difference in time between the ECGs R-peak and the peak of PPG pulses. PTT was also calculated for the clean segments before $(\text{PTT}_{b})$ and after $(\text{PTT}_{a})$ the reconstructed signal and compared via PC. Median HR in PPGs: 87.5 bpm, in ECGs: 88.01 bpm (MAE: 1.59 bpm). ECG-PPG HR correlation: 99.31% $(p$ < 0.0001). Median PTT: 142.1 ms, $\text{PTT}_{b}$ : 95.58 ms and $\text{PTT}_{a}$ : 99.08 ms. $\text{PTT-PTT} _{b}$ correlation: 81.56% (p < 0.0001). $\text{PTT-PTT}_{a}$ correlation: 78.56% $(p$ < 0.0001). $\text{PTT}_{b}{\text{-PTT}_{a}}$ correlation: 92.56% $(p$ < 0.0001). The method shows outstanding performance for HR estimation during noise and can be used for remote HR monitoring. The non-perfect correlation of the two clean segments stresses the difficulty of a high performance on PTT calculation, implying that the method could also be used for PTT estimation.

Atrial substrate modification after pulmonary vein isolation (PVI) of paroxysmal atrial fibrillation (pAF) can be assessed non-invasively by analyzing P-wave duration in the electrocardiogram (ECG). However, whether right (RA) and left atrium (LA) contribute equally to this phenomenon remains unknown. The present study splits fundamental P-wave features to investigate the different RA and LA contributions to P-wave duration. Recordings of 29 pAF patients undergoing first-ever PVI were acquired before and after PVI. P-wave features were calculated: P-wave duration (PWD), duration of the first (PWDon-peak) and second (PWDpeak-off) P-wave halves, estimating RA and LA conduction, respectively. P-wave onset (PWon-R) or offset (PWoff-R) to R-peak interval, measuring combined atrial/atrioventricular and single atrioventricular conduction, respectively. Heart-rate fluctuation was corrected by scaling. Pre- and post-PVI results were compared with Mann–Whitney U-test. PWD was correlated with the remaining features. Only PWD (non-scaling: Δ=−9.84%, p=0.0085, scaling: Δ=−17.96%, p=0.0442) and PWDpeak-off (non-scaling: Δ=−22.03%, p=0.0250, scaling: Δ=−27.77%, p=0.0268) were decreased. Correlation of all features with PWD was significant before/after PVI (p<0.0001), showing the highest value between PWD and PWon-R (ρmax=0.855). PWD correlated more with PWDon-peak (ρ= 0.540–0.805) than PWDpeak-off (ρ= 0.419–0.710). PWD shortening after PVI of pAF stems mainly from the second half of the P-wave. Therefore, noninvasive estimation of LA conduction time is critical for the study of atrial substrate modification after PVI and should be addressed by splitting the P-wave in order to achieve improved estimations.

Introduction Rapid deployment aortic valve replacement has been recently introduced in clinical practice. Different studies have reported a significant reduction in surgical times with excellent hemodynamic profiles and short-term results. However, an increase in permanent pacemaker requirements compared with conventional aortic valve replacement has been described. Nevertheless, risk factors for postoperative pacemaker implantation are not well known. The aim of this study is to report our early outcomes with rapid deployment aortic valve replacement within the RADAR Registry, especially focusing on risk factors for postoperative pacemaker implantation. Methods Between April 2012 and January 2016, 164 patients undergoing isolated or combined aortic valve replacement with Edwards INTUITY Elite (Edwards Lifesciences, Irvine, CA, USA) were included in the RADAR Registry. Pre-, intra- and postoperative clinical data results and complications were recorded, especially focusing on risk factors for the development of postoperative complete or high-grade AV block requiring pacemaker implantation. Patients were followed up for up to 1 year with evaluation of clinical and echocardiographic outcomes. Results A total of 164 consecutive patients were included in this study, where 128 patients (78.05%) had an isolated aortic valve replacement (group 1) and 36 (21.95%) a concomitant procedure (group 2). The surgical approach was ministernotomy in 61 patients (37.20%) and median sternotomy in 100 patients (60.98%). Complications with valve implantation were observed in three patients. Postoperative complete or high-degree AV block requiring a permanent pacemaker implantation developed in ten patients (6.9%). Seven patients died in-hospital (4.27%). No significant differences between groups were found in terms of stroke, postoperative infection, mortality, atrial fibrillation and postoperative atrioventricular block. Seven patients presented acute renal impairment (5.51%) in group 1 versus seven patients (20%) in group 2 ( p = 0.007). In multivariate analysis, low weight and preoperative arrhythmia (atrial fibrillation, bifascicular block, left bundle branch block) emerged as risk factors for postoperative AV block requiring a pacer. In median follow-up of 1 year, seven (4.27%) patients died, and no cases of structural valve deterioration or endocarditis were observed. Significant patient-prosthesis mismatch was found in seven (4.27%) patients. Conclusion Initial experience with rapid deployment aortic valve replacement in the RADAR Registry demonstrates low rates of implantation complications and good perioperative and 1-year clinical and echocardiographic outcomes. Incidence of postoperative AV block requiring a pacer correlated with low weight and preoperative arrythmias (atrial fibrillation, bifascicular block and left bundle branch block). Avoidance of oversizing and careful consideration of implantation of this technology in patients with pre-existing arrythmias could minimize the risk for postoperative pacemaker implantation.

Cardiovascular disease is one of the leading causes of death, with hypertension (HT) being its main risk factor. Its complications can be avoided with early treatment, but since these patients do not present any symptoms, HT is often detected at very advanced stages. This work presents a model for estimating blood pressure (BP) from electrocardiographic (ECG) and photoplethysmographic (PPG) signals, which can be easily obtained by means of wearable continuous monitoring devices. ECG, PPG and BP recordings from 86 patients were analyzed.A total of 34 standard and new features based on previous works were defined, such as pulse arrival times (PAT), and morphological characteristics of PPG signal. 37 classification models, ranging from Logistic Regression, Support Vector Machines (SVM), Nearest Neighbors, Naive Bayes or Coarse Trees were trained to compare discrimination results. The classifier that provided the highest performance when comparing normotensive patients with prehyperten-sive and hypertensive patients were Coarse Tree, providing an F1 score of 85.44% (Se of 86.27% and Sp of 77.14%). The use of PPG and ECG features has successfully discriminated between healthy and hypertensive individuals and, thus, could be used to detect HT by embedding these techniques in wearable devices.

Atrial substrate alteration due to catheter ablation (CA) of atrial fibrillation (AF) is primarily assessed from P-waves. Nonetheless, how CA affects critical structures is ignored. The aim of the current study is to investigate if CA effect on CS, the principal CA reference, is related to that observed from P-waves analysis. Five-minute lead II and bipolar CS recordings of 29 paroxysmal AF patients were obtained before, during and after CA. Duration, amplitude, area and heart-rate (HR) variability (HRV) features were calculated for P-waves and local activation waves (LAWs). Normalization mitigated the effect of HR fluctuations. Linear correlations between each P-wave and LAW were tested with linear regression (LR) and Pearson correlation (PC) and nonlinear correlations with cross-quadratic sample entropy (CQSE). Correlation between the CA effect on P-waves and LAWs was investigated with PC. Negligent statistical correlations were found by PC and LR for amplitude and area (−3.30%< corr < +18.36%, p < 0.0142). After normalization, correlation in duration increased from non-significant to significant and from low to moderate (up to 47.25%, p < 0.0001). CQSE values were from 0.6 to 1.2. The effect of CA on P-waves and LAWs duration showed a moderate/high concordance after normalization (from 50% to 89%, p < 0.0210) and a highly tuned HRV (> 90%, p < 0.0297). Apart from HRV, no significant correlations between CS LAWs and P-waves have been found. HR fluctuations mask any possible tuning and normalization should be applied prior to the analysis.

Catheter ablation (CA) is the star treatment of atrial fibrillation (AF). However, important issues regarding its procedure have only been superficially explored. While universal CA effect is assessed, the role of right (RPVI) and left pulmonary vein isolation (LPVI) is ignored. Although coronary sinus (CS) is the prevailing CA reference, how CS itself is modified by CA is unknown. This work evaluates the effect of each ablation step on the atrial substrate and CS funtion. Five-minute lead II and bipolar CS recordings of 29 patients undergoing paroxysmal AF CA were acquired before CA, after LPVI and after RPVI (end of CA). Separate lead II and CS analysis was performed. Duration, amplitude, area and slope rate were calculated for each surface and invasive activation, then signal-averaged. Dispersion, morphology variability (MV) and time-domain heart-rate variability (HRV) features were also calculated. Non-parametric tests were recruited to compare each feature among all and in pairs of different ablation steps with Bonferroni correction. Variation of each feature was calculated in percentages. In surface recordings, duration was significantly shortened after LPVI (Δ= -13%, p=0.001) and HRV showed a trend for attenuation (Δ< -25%, p< 0.069) after RPVI. In CS recordings, HRV showed an increasing trend after LPVI (Δ>+73%, p<0.04S), tending to decrease after RPVI (Δ< -33%, p<0.064). Higher dispersion in variations was observed in CS than surface recordings. LPVI causes major alterations in atrial substrate, more prominently observed from lead II analysis. Notwithstanding, HRV variations are better illustrated in CS recordings. A combined analysis of both is recommended.