USP20 deubiquitinates and stabilizes the reticulophagy receptor RETREG1/FAM134B to drive reticulophagy.

The endoplasmic reticulum (ER) serves as a hub for various cellular processes, and maintaining ER homeostasis is essential for cell function. Reticulophagy is a selective process that removes impaired ER subdomains through autophagy-mediatedlysosomal degradation. While the involvement of ubiquitination in autophagy regulation is well-established, its role in reticulophagy remains unclear. In this study, we screened deubiquitinating enzymes (DUBs) involved in reticulophagy and identified USP20 (ubiquitin specific peptidase 20) as a key regulator of reticulophagy under starvation conditions. USP20 specifically cleaves K48- and K63-linked ubiquitin chains on the reticulophagy receptor RETREG1/FAM134B (reticulophagy regulator 1), thereby stabilizing the substrate and promoting reticulophagy. Remarkably, despite lacking a transmembrane domain, USP20 is recruited to the ER through its interaction with VAPs (VAMP associated proteins). VAPs facilitate the recruitment of early autophagy proteins, including WIPI2 (WD repeat domain, phosphoinositide interacting 2), to specific ER subdomains, where USP20 and RETREG1 are enriched. The recruitment of WIPI2 and other proteins in this process plays a crucial role in facilitating RETREG1-mediated reticulophagy in response to nutrient deprivation. These findings highlight the critical role of USP20 in maintaining ER homeostasis by deubiquitinating and stabilizing RETREG1 at distinct ER subdomains, where USP20 further recruits VAPs and promotes efficient reticulophagy.Abbreviations: ACTB actin beta; ADRB2 adrenoceptor beta 2; AMFR/gp78 autocrine motility factor receptor; ATG autophagy related; ATL3 atlastin GTPase 3; BafA1 bafilomycin A1; BECN1 beclin 1; CALCOCO1 calcium binding and coiled-coil domain 1; CCPG1 cell cycle progression 1; DAPI 4',6-diamidino-2-phenylindole; DTT dithiothreitol; DUB deubiquitinating enzyme; EBSS Earle's Balanced Salt Solution; FFAT two phenylalanines (FF) in an acidic tract; GABARAP GABA type A receptor-associated protein; GFP green fluorescent protein; HMGCR 3-hydroxy-3-methylglutaryl-CoA reductase; IL1B interleukin 1 beta; LIR LC3-interacting region; MAP1LC3/LC3 microtubule associated protein 1 light chain 3; PIK3C3/Vps34 phosphatidylinositol 3-kinase catalytic subunit type 3; RB1CC1/FIP200 RB1 inducible coiled-coil 1; RETREG1/FAM134B reticulophagy regulator 1; RFP red fluorescent protein; RHD reticulon homology domain; RIPK1 receptor interacting serine/threonine kinase 1; RTN3L reticulon 3 long isoform; SEC61B SEC61 translocon subunit beta; SEC62 SEC62 homolog, preprotein translocation factor; SIM super-resolution structured illumination microscopy; SNAI2 snail family transcriptional repressor 2; SQSTM1/p62 sequestosome 1; STING1/MITA stimulator of interferon response cGAMP interactor 1; STX17 syntaxin 17; TEX264 testis expressed 264, ER-phagy receptor; TNF tumor necrosis factor; UB ubiquitin; ULK1 unc-51 like autophagy activating kinase 1; USP20 ubiquitin specific peptidase 20; USP33 ubiquitin specific peptidase 33; VAMP8 vesicle associated membrane protein 8; VAPs VAMP associated proteins; VMP1 vacuole membrane protein 1; WIPI2 WD repeat domain, phosphoinositide interacting 2; ZFYVE1/DFCP1 zinc finger FYVE-type containing 1.

Intelligent assessment of atrial fibrillation gradation based on sinus rhythm electrocardiogram and baseline information.

BACKGROUND: Atrial fibrillation (AF) is a progressive arrhythmia that significantly affects a patient's quality of life. The 4S-AF scheme is clinically recommended for AF management; however, the evaluation process is complex and time-consuming. This renders its promotion in primary medical institutions challenging. This retrospective study aimed to simplify the evaluation process and present an objective assessment model for AF gradation. METHODS: In total, 189 12-lead electrocardiogram (ECG) recordings from 64 patients were included in this study. The data were annotated into two groups (mild and severe) according to the 4S-AF scheme. Using a preprocessed ECG during the sinus rhythm (SR), we obtained a synthesized vectorcardiogram (VCG). Subsequently, various features were calculated from both signals, and age, sex, and medical history were included as baseline characteristics. Different machine learning models, including support vector machines, random forests (RF), and logistic regression, were finally tested with a combination of feature selection techniques. RESULTS: The proposed method demonstrated excellent performance in the classification of AF gradation. With an optimized feature set of VCG and baseline features, the RF model achieved accuracy, sensitivity, and specificity of 83.02 %, 80.56 %, and 88.24 %, respectively, under the inter-patient paradigm. CONCLUSION: Our results demonstrate the value of physiological signals in AF gradation evaluation, and VCG signals were effective in identifying mild and severe AF. Considering its low computational complexity and high assessment performance, the proposed model is expected to serve as a useful prognostic tool for clinical AF management.

A multi-module algorithm for heartbeat classification based on unsupervised learning and adaptive feature transfer.

The scarcity of annotated data is a common issue in the realm of heartbeat classification based on deep learning. Transfer learning (TL) has emerged as an effective strategy for addressing this issue. However, current TL techniques in this realm overlook the probability distribution differences between the source domain (SD) and target domain (TD) databases. The motivation of this paper is to address the challenge of labeled data scarcity at the model level while exploring an effective method to eliminate domain discrepancy between SD and TD databases, especially when SD and TD are derived from inconsistent tasks. This study proposes a multi-module heartbeat classification algorithm. Initially, unsupervised feature extractors are designed to extract rich features from unlabeled SD and TD data. Subsequently, a novel adaptive transfer method is proposed to effectively eliminate domain discrepancy between features of SD for pre-training (PTF-SD) and features of TD for fine-tuning (FTF-TD). Finally, the adapted PTF-SD is employed to pre-train a designed classifier, and FTF-TD is used for classifier fine-tuning, with the objective of evaluating the algorithm's performance on the TD task. In our experiments, MNIST-DB serves as the SD database for handwritten digit image classification task, MIT-DB as the TD database for heartbeat classification task. The overall accuracy of classifying heartbeats into normal heartbeats, supraventricular ectopic beats (SVEBs), and ventricular ectopic beats (VEBs) reaches 96.7 %. Specifically, the sensitivity (Sen), positive predictive value (PPV), and F1 score for SVEBs are 0.802, 0.701, and 0.748, respectively. For VEBs, Sen, PPV, and F1 score are 0.976, 0.840, and 0.903, respectively. The results indicate that the proposed multi-module algorithm effectively addresses the challenge labeled data scarcity in heartbeat classification through unsupervised learning and adaptive feature transfer methods.

A Lightweight Hybrid Model Using Multiscale Markov Transition Field for Real-Time Quality Assessment of Photoplethysmography Signals.

OBJECTIVE: The proliferation of wearable devices has escalated the standards for photoplethysmography (PPG) signal quality. This study introduces a lightweight model to address the imperative need for precise, real-time evaluation of PPG signal quality, followed by its deployment and validation utilizing our integrated upper computer and hardware system. METHODS: Multiscale Markov Transition Fields (MMTF) are employed to enrich the morphological information of the signals, serving as the input for our proposed hybrid model (HM). HM undergoes initial pre-training utilizing the MIMIC-III and UCI databases, followed by fine-tuning the Queensland dataset. Knowledge distillation (KD) then transfers the large-parameter model's knowledge to the lightweight hybrid model (LHM). LHM is subsequently deployed on the upper computer for real-time signal quality assessment. RESULTS: HM achieves impressive accuracies of 99.1% and 96.0% for binary and ternary classification, surpassing current state-of-the-art methods. LHM, with only 0.2 M parameters (0.44% of HM), maintains high accuracy despite a 2.6% drop. It achieves an inference speed of 0.023 s per image, meeting real-time display requirements. Furthermore, LHM attains a 97.7% accuracy on a self-created database. HM outperforms current methods in PPG signal quality accuracy, demonstrating the effectiveness of our approach. Additionally, LHM substantially reduces parameter count while maintaining high accuracy, enhancing efficiency and practicality for real-time applications. CONCLUSION: The proposed methodology demonstrates the capability to achieve high-precision and real-time assessment of PPG signal quality, and its practical validation has been successfully conducted during deployment. SIGNIFICANCE: This study contributes a convenient and accurate solution for the real-time evaluation of PPG signals, offering extensive application potential.

A "two-step classification" machine learning method for non-invasive localization of premature ventricular contraction origins based on 12-lead ECG.

BACKGROUND: Premature ventricular contraction (PVC) is a type of cardiac arrhythmia that originates from ectopic pacemaker in the ventricles. The localization of the origin of PVC is essential for successful catheter ablation. However, most studies on non-invasive PVC localization focus on elaborate localization in specific regions of the ventricle. This study aims to propose a machine learning algorithm based on 12-lead electrocardiogram (ECG) data that can improve the accuracy of PVC localization in the whole ventricle. METHODS: We collected 12-lead ECG data from 249 patients with spontaneous or pacing-induced PVCs. The ventricle was divided into 11 segments. In this paper, we propose a machine learning method consisting of two consecutive classification steps. In the first classification step, each PVC beat was labeled to one of the 11 ventricular segments using six features, including a newly proposed morphological feature called "Peak_index." Four machine learning methods were tested for comparative multi-classification performance and the best classifier result was kept to the next step. In the second classification step, a binary classifier was trained using a smaller combination of features to further differentiate segments that are easily confused. RESULTS: The Peak_index as a proposed new classification feature combined with other features is suitable for whole ventricle classification by machine learning methods. The test accuracy of the first classification reached 75.87%. It is shown that a second classification for confusable categories can improve the classification results. After the second classification, the test accuracy reached 76.84%, and when a sample classified into adjacent segments was considered correct, the test "rank accuracy" was improved to 93.49%. The binary classification corrected 10% of the confused samples. CONCLUSION: This paper proposes a "two-step classification" method to localize the origin of PVC beats into the 11 regions of the ventricle using non-invasive 12-lead ECG. It is expected to be a promising technique to be used in clinical settings to help guide ablation procedures.

Time-resolved proteomic profiling reveals compositional and functional transitions across the stress granule life cycle.

Stress granules (SGs) are dynamic, membrane-less organelles. With their formation and disassembly processes characterized, it remains elusive how compositional transitions are coordinated during prolonged stress to meet changing functional needs. Here, using time-resolved proteomic profiling of the acute to prolonged heat-shock SG life cycle, we identify dynamic SG proteins, further segregated into early and late proteins. Comparison of different groups of SG proteins suggests that their biochemical properties help coordinate SG compositional and functional transitions. In particular, early proteins, with high phase-separation-propensity, drive the rapid formation of the initial SG platform, while late proteins are subsequently recruited as discrete modules to further functionalize SGs. This model, supported by immunoblotting and immunofluorescence imaging, provides a conceptual framework for the compositional transitions throughout the acute to prolonged SG life cycle. Additionally, an early SG constituent, non-muscle myosin II, is shown to promote SG formation by increasing SG fusion, underscoring the strength of this dataset in revealing the complexity of SG regulation.

Semi-Supervised Learning for Low-Cost Personalized Obstructive Sleep Apnea Detection Using Unsupervised Deep Learning and Single-Lead Electrocardiogram.

OBJECTIVE: Obstructive sleep apnea (OSA) is a common sleep-related breathing disorder that can lead to a wide range of health issues if left untreated. This study aims to address the lack of research on personalized models for single-lead electrocardiogram (ECG)-based OSA detection, by proposing an automatic semi-supervised algorithm for automated low-cost personalization fine-tuning. METHODS: We utilize a convolutional neural network (CNN)-based auto-encoder (AE) with a modified training objective to detect anomalous region of OSA. An indicator based on model outputs is utilized as a benchmark measure to assign pseudo-labels with confidence to each sample. Finally, we perform validation of the semi-supervised algorithm on the same database and cross-database scenarios. RESULTS: By introducing semi-supervised personalization, the accuracy, AUC, and mean absolute error (MAE) of the general model (GM) of 35 subjects from the same database are improved from 86.3%, 0.915, and 5.178 to 90.3%, 0.948, and 2.593. Simultaneously, in the validation of 25 subjects from a cross-database, the accuracy, AUC, and MAE of the GM are enhanced from 75.6%, 0.800, and 9.149 to 84.3%, 0.881, and 3.509. CONCLUSION: The improved version of AE demonstrates excellent adaptability in identifying abnormal features in OSA, employing a data-driven approach to assign pseudo-labels for unknown data automatically. Additionally, leveraging the pseudo-labels through a semi-supervised fine-tuning strategy provides a solution to overcome the limitation of clinical annotations, facilitating low-cost implementation of personalized models. SIGNIFICANCE: The semi-supervised approach proposed in this article provides a high-performance and annotation-free solution for personalized adjustment of automatic OSA detection.

Exploring the Applicability of Transfer Learning and Feature Engineering in Epilepsy Prediction Using Hybrid Transformer Model.

OBJECTIVE: Epilepsy prediction algorithms offer patients with drug-resistant epilepsy a way to reduce unintended harm from sudden seizures. The purpose of this study is to investigate the applicability of transfer learning (TL) technique and model inputs for different deep learning (DL) model structures, which may provide a reference for researchers to design algorithms. Moreover, we also attempt to provide a novel and precise Transformer-based algorithm. METHODS: Two classical feature engineering methods and the proposed method which consists of various EEG rhythms are explored, then a hybrid Transformer model is designed to evaluate the advantages over pure convolutional neural networks (CNN)-based models. Finally, the performances of two model structures are analyzed utilizing patient-independent approach and two TL strategies. RESULTS: We tested our method on the CHB-MIT scalp EEG database, the results showed that our feature engineering method gains a significant improvement in model performance and is more suitable for Transformer-based model. In addition, the performance improvement of Transformer-based model utilizing fine-tuning strategies is more robust than that of pure CNN-based model, and our model achieved an optimal sensitivity of 91.7% with false positive rate (FPR) of 0.00/h. CONCLUSION: Our epilepsy prediction method achieves excellent performance and demonstrates its advantage over pure CNN-based structure in TL. Moreover, we find that the information contained in the gamma ( γ ) rhythm is helpful for epilepsy prediction. SIGNIFICANCE: We propose a precise hybrid Transformer model for epilepsy prediction. The applicability of TL and model inputs is also explored for customizing personalized models in clinical application scenarios.

Value of baseline characteristics in the risk prediction of atrial fibrillation.

Introduction: Atrial fibrillation (AF) is prone to heart failure and stroke. Early management can effectively reduce the stroke rate and mortality. Current clinical guidelines screen high-risk individuals based solely on age, while this study aims to explore the possibility of other AF risk predictors. Methods: A total of 18,738 elderly people (aged over 60 years old) in Chinese communities were enrolled in this study. The baseline characteristics were mainly based on the diagnosis results of electrocardiogram (ECG) machine during follow up, accompanied by some auxiliary physical examination basic data. After the analysis of both independent and combined baseline characteristics, AF risk predictors were obtained and prioritized according to the results. Independent characteristics were studied from three aspects: Chi-square test, Mann-Whitney U test and Cox univariate regression analysis. Combined characteristics were studied from two aspects: machine learning models and Cox multivariate regression analysis, and the former was combined with recursive feature elimination method and voting decision. Results: The resulted optimal combination of risk predictors included age, atrial premature beats, atrial flutter, left ventricular hypertrophy, hypertension and heart disease. Conclusion: Patients diagnosed by short-time ECG machines with the occurrence of the above events had a higher probability of AF episodes, who are suggested to be included in the focus of long-term ECG monitoring or increased screening density. The incidence of risk predictors in different age ranges of AF patients suggests differences in age-specific patient management. This can help improve the detection rate of AF, standardize the management of patients, and slow down the progression of AF.

Stress granule homeostasis is modulated by TRIM21-mediated ubiquitination of G3BP1 and autophagy-dependent elimination of stress granules.

Eukaryotic stress granules (SGs) are highly dynamic assemblies of untranslated mRNAs and proteins that form through liquid-liquid phase separation (LLPS) under cellular stress. SG formation and elimination process is a conserved cellular strategy to promote cell survival, although the precise regulation of this process is poorly understood. Here, we screened six E3 ubiquitin ligases present in SGs and identified TRIM21 (tripartite motif containing 21) as a central regulator of SG homeostasis that is highly enriched in SGs of cells under arsenite-induced oxidative stress. Knockdown of TRIM21 promotes SG formation whereas overexpression of TRIM21 inhibits the formation of physiological and pathological SGs associated with neurodegenerative diseases. TRIM21 catalyzes K63-linked ubiquitination of the SG core protein, G3BP1 (G3BP stress granule assembly factor 1), and G3BP1 ubiquitination can effectively inhibit LLPS, in vitro. Recent reports suggested the involvement of macroautophagy/autophagy, as a stress response pathway, in the regulation of SG homeostasis. We systematically investigated well-defined autophagy receptors and identified SQSTM1/p62 (sequestosome 1) and CALCOCO2/NDP52 (calcium binding and coiled-coil domain 2) as the primary receptors that directly interact with G3BP1 during arsenite-induced stress. Endogenous SQSTM1 and CALCOCO2 localize to the periphery of SGs under oxidative stress and mediate SG elimination, as single knockout of each receptor causes accumulation of physiological and pathological SGs. Collectively, our study broadens the understanding in the regulation of SG homeostasis by showing that TRIM21 and autophagy receptors modulate SG formation and elimination respectively, suggesting the possibility of clinical targeting of these molecules in therapeutic strategies for neurodegenerative diseases.Abbreviations: ACTB: actin beta; ALS: amyotrophic lateral sclerosis; BafA1: bafilomycin A1; BECN1: beclin 1; C9orf72: C9orf72-SMCR8 complex subunit; CALCOCO2/NDP52: calcium binding and coiled-coil domain 2; Co-IP: co-immunoprecipitation; DAPI: 4',6-diamidino-2-phenylindole; FTD: frontotemporal dementia; FUS: FUS RNA binding protein; G3BP1: G3BP stress granule assembly factor 1; GFP: green fluorescent protein; LLPS: liquid-liquid phase separation; MAP1LC3/LC3: microtubule associated protein 1 light chain 3; NBR1: NBR1 autophagy cargo receptor; NES: nuclear export signal; OPTN: optineurin; RFP: red fluorescent protein; SQSTM1/p62: sequestosome 1; SG: stress granule; TAX1BP1: Tax1 binding protein 1; TOLLIP: toll interacting protein; TRIM21: tripartite motif containing 21; TRIM56: tripartite motif containing 56; UB: ubiquitin; ULK1: unc-51 like autophagy activating kinase 1; WT: wild-type.

A 3D-CNN with temporal-attention block to predict the recurrence of atrial fibrillation based on body-surface potential mapping signals.

Catheter ablation has become an important treatment for atrial fibrillation (AF), but its recurrence rate is still high. The aim of this study was to predict AF recurrence using a three-dimensional (3D) network model based on body-surface potential mapping signals (BSPMs). BSPMs were recorded with a 128-lead vest in 14 persistent AF patients before undergoing catheter ablation (Maze-IV). The torso geometry was acquired and meshed by point cloud technology, and the BSPM was interpolated into the torso geometry by the inverse distance weighted (IDW) method to generate the isopotential map. Experiments show that the isopotential map of BSPMs can reflect the propagation of the electrical wavefronts. The 3D isopotential sequence map was established by combining the spatial-temporal information of the isopotential map; a 3D convolutional neural network (3D-CNN) model with temporal attention was established to predict AF recurrence. Our study proposes a novel attention block that focuses the characteristics of atrial activations to improve sampling accuracy. In our experiment, accuracy (ACC) in the intra-patient evaluation for predicting the recurrence of AF was 99.38%. In the inter-patient evaluation, ACC of 3D-CNN was 81.48%, and the area under the curve (AUC) was 0.88. It can be concluded that the dynamic rendering of multiple isopotential maps can not only comprehensively display the conduction of cardiac electrical activity on the body surface but also successfully predict the recurrence of AF after CA by using 3D isopotential sequence maps.

Long-term single-lead electrocardiogram monitoring to detect new-onset postoperative atrial fibrillation in patients after cardiac surgery.

Background: Postoperative atrial fibrillation (POAF) is often associated with serious complications. In this study, we collected long-term single-lead electrocardiograms (ECGs) of patients with preoperative sinus rhythm to build statistical models and machine learning models to predict POAF. Methods: All patients with preoperative sinus rhythm who underwent cardiac surgery were enrolled and we collected long-term ECG data 24 h before surgery and 7 days after surgery by single-lead ECG. The patients were divided into a POAF group a no-POAF group. A clinical model and a clinical + ECG model were constructed. The ECG parameters were designed and support vector machine (SVM) was selected to build a machine learning model and evaluate its prediction efficiency. Results: A total of 100 patients were included. The detection rate of POAF in long-term ECG monitoring was 31% and that in conventional monitoring was 19%. We calculated 7 P-wave parameters, Pmax (167 ± 31 ms vs. 184 ± 37 ms, P = 0.018), Pstd (15 ± 7 vs. 19 ± 11, P = 0.031), and PWd (62 ± 28 ms vs. 80 ± 35 ms, P = 0.008) were significantly different. The AUC of the clinical model (sex, age, LA diameter, GFR, mechanical ventilation time) was 0.86. Clinical + ECG model (sex, age, LA diameter, GFR, mechanical ventilation time, Pmax, Pstd, PWd), AUC was 0.89. In the machine learning model, the accuracy (Ac) of the train set and test set was above 80 and 60%, respectively. Conclusion: Long-term ECG monitoring could significantly improve the detection rate of POAF. The clinical + ECG model and the machine learning model based on P-wave parameters can predict POAF.

Associations between electrocardiogram and carotid ultrasound parameters: a healthy chinese group study.

Background: Electrocardiogram (ECG) and carotid ultrasound (CUS) are important tools for the diagnosis and prediction of cardiovascular disease (CVD). This study aimed to investigate the associations between ECG and CUS parameters and explore the feasibility of assessing carotid health with ECG. Methods: This cross-sectional cohort study enrolled 319 healthy Chinese subjects. Standard 12-lead ECG parameters (including the ST-segment amplitude [STA]), CUS parameters (intima-media thickness [IMT] and blood flow resistance index [RI]), and CVD risk factors (including sex, age, and systolic blood pressure [SBP]) were collected for analysis. Participants were divided into the high-level RI group (average RI ≥ 0.76, n = 171) and the normal RI group (average RI < 0.76, n = 148). Linear and stepwise multivariable regression models were performed to explore the associations between ECG and CUS parameters. Results: Statistically significant differences in sex, age, SBP, STA and other ECG parameters were observed in the normal and the high-level RI group. The STA in lead V3 yielded stronger significant correlations (r = 0.27-0.42, p < 0.001) with RI than STA in other leads, while ECG parameters yielded weak correlations with IMT (|r| ≤ 0.20, p < 0.05). STA in lead V2 or V3, sex, age, and SBP had independent contributions (p < 0.01) to predicting RI in the stepwise multivariable models, although the models for IMT had only CVD risk factors (age, body mass index, and triglyceride) as independent variables. The prediction model for RI in the left proximal common carotid artery (CCA) had higher adjusted R2 (adjusted R2 = 0.31) than the model for RI in the left middle CCA (adjusted R2 = 0.29) and the model for RI in the right proximal CCA (adjusted R2 = 0.20). Conclusion: In a cohort of healthy Chinese individuals, the STA was associated with the RI of CCA, which indicated that ECG could be utilized to assess carotid health. The utilization of ECG might contribute to a rapid screening of carotid health with convenient operations.

A two-step method for paroxysmal atrial fibrillation event detection based on machine learning.

Detection of atrial fibrillation (AF) events is significant for early clinical diagnosis and appropriate intervention. However, in existing detection algorithms for paroxysmal AF (AFp), the location of AF starting and ending points in AFp is not concerned. To achieve an accurate identification of AFp events in the long-term dynamic electrocardiograms (ECGs), this paper proposes a two-step method based on machine learning. In the first step, based on features extracted from the calculated R-to-R intervals (RR intervals, the cycle of heart beat), the rhythm type of the ECG signal is first classified into three classes (AFp rhythm, persistent AF (AFf) rhythm, and non-atrial fibrillation (non-AF, N) rhythm) using support vector machine (SVM). In the second step, the starting and ending points for AF episodes of AFp rhythms predicted in the first step are further located based on heartbeat classification. By training a deep convolutional neural network with phased training, the segmented beats of AFp rhythms are divided into AF beats and non-AF beats to determine the beginning and end of any AF episode. The proposed two-step method is trained and tested on the 4th China Physiological Signal Challenge 2021 databases. A final score U of 1.9310 is obtained on the unpublished test set maintained by the challenge organizers, which demonstrates the advantage of the two-step method in AFp event detection. The work is useful for assessing AF burden index for AFp patients.

MGNN: A multiscale grouped convolutional neural network for efficient atrial fibrillation detection.

The reliable detection of atrial fibrillation (AF) is of great significance for monitoring disease progression and developing tailored care paths. In this work, we proposed a novel and robust method based on deep learning for the accurate detection of AF. Using RR interval sequences, a multiscale grouped convolutional neural network (MGNN) combined with self-attention was designed for automatic feature extraction, and AF and non-AF classification. An average accuracy of 97.07% was obtained in the 5-fold cross-validation. The generalization ability of the proposed MGNN was further independently tested on four other unseen datasets, and the accuracy was 92.23%, 96.86%, 94.23% and 95.91%. Moreover, comparison of the network structures indicated that the MGNN had not only better detection performance but also lower computational complexity. In conclusion, the proposed model is shown to be an efficient AF detector that has great potential for use in clinical auxiliary diagnosis and long-term home monitoring based on wearable devices.

A novel SCN9A gene variant identified in a Chinese girl with paroxysmal extreme pain disorder (PEPD): a rare case report.

BACKGROUND: Paroxysmal extreme pain disorder (PEPD) is a rare autosomal dominant hereditary disease, characterized by paroxysmal burning pain in the rectum, eyes or mandible and autonomic nervous symptoms, including skin redness and bradycardia. PEPD is a sodium channel dysfunctional disorder caused by SCN9A gene variants. It occurs mainly in Caucasians and only one case has been reported in the Chinese population. Here, we report the second PEPD case in a Chinese indivisual. CASE PRESENTATION: A 2 years and 6 months old girl initially presented with non-epileptic tonic seizures at 7 days after birth. Her clinical symptoms in order of presentation were non-epileptic tonic seizures, harlequin color change and pain. Genetic analysis showed the patient carried a heterozygous variant c.4384T>A (p.F1462I) in the SCN9A gene, which was speculated to cause PEPD symptoms. After administrating carbamazepine, the symptoms were relieved and the patient's condition improved. However, the patient's mother, who carries the same SCN9A variant as her daughter, only showed bradycardia and sinus arrest but no PEPD-related pain. CONCLUSIONS: This is the second PEPD case reported in the Chinese population. With the discovery of a novel variant in SCN9A, we expanded the genotype spectrum of PEPD. This is the first case suggesting that the clinical presentations of SCN9A-associated PEPD may show inter familial phenotypic diversity. In the future of clinical diagnosis, patients with triggered non-epileptic tonic seizures or pain and harlequin color change should be considered for PEPD and proper and prompt treatment should be given.

Biometric recognition based on scalable end-to-end convolutional neural network using photoplethysmography: A comparative study.

Photoplethysmography (PPG), as one of the most widely used physiological signals on wearable devices, with dominance for portability and accessibility, is an ideal carrier of biometric recognition for guaranteeing the security of sensitive information. However, the existing state-of-the-art methods are restricted to practical deployment since power-constrained and compute-insufficient for wearable devices. 1D convolutional neural networks (1D-CNNs) have succeeded in numerous applications on sequential signals. Still, they fall short in modeling long-range dependencies (LRD), which are extremely needed in high-security PPG-based biometric recognition. In view of these limitations, this paper conducts a comparative study of scalable end-to-end 1D-CNNs for capturing LRD and parameterizing authorized templates by enlarging the receptive fields via stacking convolution operations, non-local blocks, and attention mechanisms. Compared to a robust baseline model, seven scalable models have different impacts (-0.2%-9.9%) on the accuracy of recognition over three datasets. Experimental cases demonstrate clear-cut improvements. Scalable models achieve state-of-the-art performance with an accuracy of over 97% on VitalDB and with the best accuracy on BIDMC and PRRB datasets performing 99.5% and 99.3%, respectively. We also discuss the effects of capturing LRD in generated templates by visualizations with Gramian Angular Summation Field and Class Activation Map. This study conducts that the scalable 1D-CNNs offer a performance-excellent and complexity-feasible approach for biometric recognition using PPG.

Comprehensive analysis of resting tremor based on acceleration signals of patients with Parkinson's disease.

BACKGROUND: Resting tremor is an essential characteristic in patients suffering from Parkinson's disease (PD). OBJECTIVE: Quantification and monitoring of tremor severity is clinically important to help achieve medication or rehabilitation guidance in daily monitoring. METHODS: Wrist-worn tri-axial accelerometers were utilized to record the long-term acceleration signals of PD patients with different tremor severities rated by Unified Parkinson's Disease Rating Scale (UPDRS). Based on the extracted features, three kinds of classifiers were used to identify different tremor severities. Statistical tests were further designed for the feature analysis. RESULTS: The support vector machine (SVM) achieved the best performance with an overall accuracy of 94.84%. Additional feature analysis indicated the validity of the proposed feature combination and revealed the importance of different features in differentiating tremor severities. CONCLUSION: The present work obtains a high-accuracy classification in tremor severity, which is expected to play a crucial role in PD treatment and symptom monitoring in real life.

Poly(I:C)-exposed zebrafish shows autism-like behaviors which are ameliorated by fabp2 gene knockout.

Introduction: Autism spectrum disorder (ASD) is a group of neurodevelopmental disorders mainly representing impaired social communication. The etiology of ASD includes genetic and environmental risk factors. Rodent models containing ASD risk gene mutations or environmental risk factors, such as exposure to maternal inflammation, show abnormal behavior. Although zebrafish conserves many important brain structures of humans and has sophisticated and fine behaviors in social interaction, it is unknown whether the social behaviors of their offspring would be impaired due to exposure to maternal inflammation. Methods: We exposed zebrafish to maternal immune activation (MIA) by injection with polyinosinic:polycytidylic acid [poly(I:C)], and screened their behaviors through social behavioral tests such as social preference and shoaling behavior tests. We compared phenotypes resulted from different ways of poly(I:C) exposure. RNA sequencing was performed to explore the differential expression genes (DEGs). Gene ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) and protein-protein interaction (PPI) network analysis was performed with the detected DEGs to find the concentrated pathways. Finally, we knocked out the fatty acid-binding protein 2 (fabp2), a key node of the concentrated PPI network, to find its rescues on the altered social behavior. Results: We reported here that MIA offspring born to mothers injected with poly(I:C) exhibited impaired social approach and social cohesion that mimicked human ASD phenotypes. Both maternal exposure and direct embryo exposure to poly(I:C) resulted in activations of the innate immune system through toll-like receptors 3 and 4. RNA-sequencing results from MIA brain tissues illustrated that the numbers of overexpressed genes were significantly more than that of underexpressed genes. GO and KEGG analyses found that MIA-induced DEGs were mainly concentrated in complement and coagulation cascade pathways. PPI network analyses suggested that villin-1 (vil1) pathway might play a key role in MIA-induced ASD. Knockout of fabp2 in F0 zebrafish rescued the social behavior deficits in MIA offspring. Conclusions: Overall, our work established an ASD model with assessable behavior phenotype in zebrafish and provided key insights into environmental risk factor in ASD etiology and the influence of fabp2 gene on ASD-like behavior.

[Design and Implementation of Software Platform for AI-ECG Algorithm Research].

A software platform for AI-ECG algorithm research is designed and implemented to better serve the research of ECG artificial intelligence classification algorithm and to solve the problem of subjects data information management. Matlab R2019b and MySQL Sever 8.0 are used to design the software platform. The software platform is divided into three modules including data management module, data receiving module and data processing module. The software platform can be used to query and set the subjects information. It has realized the functions of data receiving, signal processing and the display, analysis and storage of ECG data. The software platform is easy to operate and meets the basic needs of scientific research. It is of great significance to the research of AI-ECG algorithm.