While various QRS detection and classification methods were developed in the past, the Holter ECG data acquired during daily activities by wearable devices represent new challenges such as increased noise and artefacts due to patient movements. Here, we present a deep-learning model to detect and classify QRS complexes in single-lead Holter ECG. We introduce a novel approach, delivering QRS detection and classification in one inference step. We used a private dataset (12,111 Holter ECG recordings, length of 30 s) for training, validation, and testing the method. Twelve public databases were used to further test method performance. We built a software tool to rapidly annotate QRS complexes in a private dataset, and we annotated 619,681 QRS complexes. The standardised and down-sampled ECG signal forms a 30-s long input for the deep-learning model. The model consists of five ResNet blocks and a gated recurrent unit layer. The model's output is a 30-s long 4-channel probability vector (no-QRS, normal QRS, premature ventricular contraction, premature atrial contraction). Output probabilities are post-processed to receive predicted QRS annotation marks. For the QRS detection task, the proposed method achieved the F1 score of 0.99 on the private test set. An overall mean F1 cross-database score through twelve external public databases was 0.96 ± 0.06. In terms of QRS classification, the presented method showed micro and macro F1 scores of 0.96 and 0.74 on the private test set, respectively. Cross-database results using four external public datasets showed micro and macro F1 scores of 0.95 ± 0.03 and 0.73 ± 0.06, respectively. Presented results showed that QRS detection and classification could be reliably computed in one inference step. The cross-database tests showed higher overall QRS detection performance than any of compared methods.
- MeSH
- algoritmy MeSH
- artefakty MeSH
- elektrokardiografie ambulantní metody MeSH
- elektrokardiografie metody MeSH
- komorové extrasystoly * MeSH
- lidé MeSH
- nositelná elektronika * MeSH
- počítačové zpracování signálu MeSH
- Check Tag
- lidé MeSH
- Publikační typ
- časopisecké články MeSH
- práce podpořená grantem MeSH
During the lockdown of universities and the COVID-Pandemic most students were restricted to their homes. Novel and instigating teaching methods were required to improve the learning experience and so recent implementations of the annual PhysioNet/Computing in Cardiology (CinC) Challenges posed as a reference. For over 20 years, the challenges have proven repeatedly to be of immense educational value, besides leading to technological advances for specific problems. In this paper, we report results from the class 'Artificial Intelligence in Medicine Challenge', which was implemented as an online project seminar at Technical University Darmstadt, Germany, and which was heavily inspired by the PhysioNet/CinC Challenge 2017 'AF Classification from a Short Single Lead ECG Recording'. Atrial fibrillation is a common cardiac disease and often remains undetected. Therefore, we selected the two most promising models of the course and give an insight into the Transformer-based DualNet architecture as well as into the CNN-LSTM-based model and finally a detailed analysis for both. In particular, we show the model performance results of our internal scoring process for all submitted models and the near state-of-the-art model performance for the two named models on the official 2017 challenge test set. Several teams were able to achieve F1scores above/close to 90% on a hidden test-set of Holter recordings. We highlight themes commonly observed among participants, and report the results from the self-assessed student evaluation. Finally, the self-assessment of the students reported a notable increase in machine learning knowledge.
- MeSH
- algoritmy MeSH
- COVID-19 * diagnóza MeSH
- elektrokardiografie metody MeSH
- fibrilace síní * diagnóza MeSH
- kontrola infekčních nemocí MeSH
- lidé MeSH
- strojové učení MeSH
- umělá inteligence MeSH
- Check Tag
- lidé MeSH
- Publikační typ
- časopisecké články MeSH
INTRODUCTION: Atrial fibrillation (AF), with a prevalence of 2%, is the most common cardiac arrhythmia. Catheter ablation (CA) has been documented to be superior to treatment by antiarrhythmic drugs (AADs) in terms of sinus rhythm maintenance. However, in obese patients, substantial weight loss was also associated with AF reduction. So far, no study has compared the modern non-invasive (AADs combined with risk factor modification (RFM)) approach with modern invasive (CA) treatment. The aim of the trial is to compare the efficacy of modern invasive (CA) and non-invasive (AADs with risk factor management) treatment of AF. METHODS AND ANALYSIS: The trial will be a prospective, multicentre, randomised non-inferiority trial. Patients with symptomatic AF and a body mass index >30 will be enrolled and randomised to the CA or RFM arm (RFM+AAD) in a 1:1 ratio. In the CA arm, pulmonary vein isolation (in combination with additional lesion sets in non-paroxysmal patients) will be performed. For patients in the RFM+AAD arm, the aim will be a 10% weight loss over 6-12 months, increased physical fitness and a reduction in alcohol consumption. The primary endpoint will be an episode of AF or regular atrial tachycardia lasting >30 s. The secondary endpoints include AF burden, clinical endpoints associated with AF reoccurrence, changes in the quality of life assessed using dedicated questionnaires, changes in cardiorespiratory fitness and metabolic endpoints. An AF freedom of 65% in the RFM+AAD and of 60% in the CA is expected; therefore, 202 patients will be enrolled to achieve the non-inferiority with 80% power, 5% one-sided alpha and a non-inferiority margin of 12%. ETHICS AND DISSEMINATION: The PRAGUE-25 trial will determine if modern non-invasive AF treatment strategies are non-inferior to CA. The study was approved by the Ethics Committee of the University Hospital Kralovske Vinohrady. Results of the study will be disseminated on scientific conferences and in peer-reviewed scientific journals. After the end of follow-up, data will be available upon request to principal investigator. TRIAL REGISTRATION NUMBER: ClinicalTrials.gov Registry (NCT04011800).
- MeSH
- antiarytmika terapeutické užití MeSH
- fibrilace síní * farmakoterapie chirurgie MeSH
- hmotnostní úbytek MeSH
- katetrizační ablace * metody MeSH
- kvalita života MeSH
- lidé MeSH
- multicentrické studie jako téma MeSH
- prospektivní studie MeSH
- randomizované kontrolované studie jako téma MeSH
- recidiva MeSH
- rizikové faktory MeSH
- výsledek terapie MeSH
- Check Tag
- lidé MeSH
- Publikační typ
- časopisecké články MeSH
- práce podpořená grantem MeSH
- protokol klinické studie MeSH
Telemedicínu lze definovat jako zdravotnickou službu, která zejména v oblasti diagnostiky využívá technologie současného vzdáleného přenosu velkého objemu dat od velkého množství pacientů. Tato data jsou následně centrálně zpracována a poskytována velkému množství zdravotnických subjektů, které si telemedicínskou službu po své pacienty zadávají na národní i mezinárodní úrovni. V arytmologii je telemedicína využívána zejména při dlouhodobém monitorování EKG v diagnostice arytmií a ke kontrole léčby pomocí externích záznamníků, chytrých hodinek a implantabilních přístrojů. Zpracování obrovského objemu telemedicínských dat stále více využívá umělou inteligenci.
Telemedicine can be defined as a health care service that, specifically in the field of diagnostics, employs remote transfer of a large volume of data from a large number of subjects at the same time. This data is subsequently processed on a central basis and returned to a large number of health care providers by whom the service was ordered on national or international level. In arrhythmology, telemedicine is used particularly in long-term ECG monitoring to diagnose arrhythmias and check out treatment outcome via external recorders, smart watch, and implantable devices. To facilitate analysis of large telemedicine data volume, artificial intelligence is being increasingly exploited.
- MeSH
- ambulantní monitorování metody přístrojové vybavení trendy MeSH
- defibrilátory implantabilní trendy MeSH
- elektrokardiografie ambulantní metody přístrojové vybavení trendy MeSH
- lidé MeSH
- mobilní aplikace MeSH
- srdeční arytmie * diagnóza prevence a kontrola MeSH
- telemedicína * metody přístrojové vybavení trendy MeSH
- umělá inteligence MeSH
- Check Tag
- lidé MeSH
- Publikační typ
- práce podpořená grantem MeSH
Dokumentace EKG křivky je základní diagnostickou metodou ke stanovení poruch srdečního rytmu. Používá se především k odhalení arytmií, které se nepodařilo zachytit pomocí standardního 12svodového EKG záznamu nebo 24–48hodinové holterovské monitorace. Rok od roku se možnosti EKG monitorace zlepšují, záznamníky EKG se zmenšují a doba, po kterou je možné EKG sledovat, se prodlužuje. V současnosti můžeme většinou pracovat s kontinuálním EKG záznamem. Velké možnosti do denní praxe přinášejí i tzv. „chytré hodinky“ či různé fitness náramky, které začínají být součástí sledování pacientů se známou arytmií či podezřením na ni. Nejdůležitějším kritériem pro výběr vhodného typu monitorace však zůstává především četnost a typ obtíží, které má konkrétní pacient.
ECG recording represents an essential method for the diagnosis of heart rhythm disturbances. Long-term monitoring helps to identify arrhythmias that have not been detected by means of standard 12-lead ECG or 24-48 hour ECG Holter. With time, ECG monitoring facilities improve, the ECG recorders are becoming smaller, and the recording time is prolonging. At present, continuous ECG recording is generally available. Smart watches and fitness bracelets further expand monitoring options in patients with known or suspected arrhythmia. Individual type and frequency of symptoms remain the most significant criterion for the selection of suitable ECG recorder and recording time.
- Klíčová slova
- dlouhodobá EKG monitorace,
- MeSH
- elektrokardiografie * metody MeSH
- lidé MeSH
- telemedicína * metody MeSH
- Check Tag
- lidé MeSH
- Publikační typ
- přehledy MeSH
Telemedicína je jedním z prudce se rozvíjejících aspektů současné medicíny; v kardiologii se uplatňuje zejména v oblasti arytmií a srdečního selhání. Dlouhodobé neinvazivní transtelefonní monitorování EKG je v Česku a dalších zemích EU poskytováno už od roku 2008 Mezinárodním centrem pro telemedicínu MDT (Medical Data Transfer), které ročně provede asi 45 tisíc vyšetření s průměrnou délkou individuálního sledování 14 ± 9 dnů. Moderní dálkové monitorování implantabilních přístrojů (kardiostimulátorů a kardioverterů-defibrilátorů) poskytují všichni jejich přední výrobci s působením v Česku (Abbott, Medtronic, Biotronik, Boston Scientific). Tímto způsobem lze automaticky sledovat technické parametry přístrojů a zároveň získávat informace o detekovaných arytmiích. V nastavení serveru je možné zadávat automatické zpracování varování a alarmů, o nichž je ošetřující lékař informován e-mailem, faxem nebo SMS. Všechna odvětví telemedicíny pracují s obrovskými objemy dat, která je obtížné zpracovat bez aplikace umělé inteligence, jejíž podstata a metody jsou v článku popsány. Mezinárodní telemedicínské centrum MDT ve spolupráci s Ústavem přístrojové techniky AV ČR vyvinulo a používá jeden z jejích modelů pro analýzu dlouhodobého EKG záznamu.
Telemedicine is one of rapidly developing discipline in current medicine, including cardiology, and specifically cardiac arrhythmias and heart failure. Long-term transtelephonic ECG monitoring has been provided since 2008 in the Czech Republic and other members of EU by International Center for Telemedicine MDT (Medical Data Transfer), with approximately 45 000 monitored individuals per year and mean individual monitoring duration of 14 ± 9 days. Current home-monitoring of implantable devices is offered by all leading companies with distribution in the Czech Republic (Abbott, Medtronic, Biotronik, Boston Scientific). It enables to follow device parameters as well as get information on detected arrhythmias, and further, through the server set up to automatically process warnings and alarms and to inform the responsible physician via e-mail, fax, or SMS. All branches of telemedicine work with a huge amount of data difficult to process without the means of artificial intelligence, whose principles and methods are discussed in the article. International Center for Telemedicine in collaboration with the Institute of Scientific Instruments of the Czech Academy of Sciences have developed and employs one such a model for the long-term ECG analysis.
- MeSH
- elektrokardiografie * klasifikace přístrojové vybavení trendy MeSH
- lidé MeSH
- Check Tag
- lidé MeSH
AIMS: Optimal ECG monitoring in detecting recurrences of atrial fibrillation (AF) or atrial tachycardia (AT) after catheter ablation has not been well established. The purpose of this prospective study was to compare the utility of daily ECG monitoring with episodic card recorder (ECR) vs. periodic monitoring with episodic loop recorder (ELR) for the detection of post-blanking AF/AT recurrences during early (Months 4-6) and late (Months 7-12) periods after catheter ablation for paroxysmal AF. METHODS: The study included 105 consecutive patients, who received ECR for 12 months and were instructed to send at least 2 random ECG recordings daily with extra-recordings during symptoms. The patients were simultaneously monitored for one week with ELR at the end of each period (Months 6 and 12). RESULTS: Thirty-one and 12 patients with AF/AT recurrence were identified by means of ECR and ELR, respectively. In patients with complete and valid data, ELR technology was inferior to ECR by detecting AF/AT in 5 (31%) of 16 and 5 (26%) of 19 patients with arrhythmia identified by ECR in the early and late period, respectively. Overall, ELR had a sensitivity of 8/23 (35%) for detecting AF/AT recurrence. There was no single patient with AF/AT recurrence on ELR that would not be known from ECR monitoring. Only 2 patients with arrhythmia recurrence were completely asymptomatic throughout the study period. CONCLUSION: Daily ECG monitoring with ECR was better than periodic monitoring with ELR in detecting AF/AT recurrences during the follow-up periods. Entirely asymptomatic patients with AF/AT recurrences were rare.
- MeSH
- antiarytmika MeSH
- časové faktory MeSH
- elektrokardiografie ambulantní * MeSH
- fibrilace síní diagnóza patofyziologie chirurgie MeSH
- katetrizační ablace * škodlivé účinky MeSH
- lidé středního věku MeSH
- lidé MeSH
- pooperační komplikace diagnóza patofyziologie MeSH
- prospektivní studie MeSH
- recidiva MeSH
- senioři MeSH
- výsledek terapie MeSH
- Check Tag
- lidé středního věku MeSH
- lidé MeSH
- mužské pohlaví MeSH
- senioři MeSH
- ženské pohlaví MeSH
- Publikační typ
- časopisecké články MeSH
- srovnávací studie MeSH
Left atrial (LA) volume (LAV) is used for the selection of patients with atrial fibrillation (AF) to rhythm control strategies. Calculation of LAV from the LA diameters and areas by two-dimensional (2D) echocardiography may result in significant error. Accuracy of atrial volume assessment has never been studied in patients with long-standing persistent AF (LSPAF) and significant atrial remodeling. This study investigated correlation and agreement between 2D echocardiographic (Simpson method) and electroanatomic (CARTO, Biosense Webster) left and right atrial (RA) volumes (LAV(ECHO) vs. LAV(CARTO) and RAV(ECHO) vs. RAV(CARTO)) in patients undergoing catheter ablation for LSPAF. The study enrolled 173 consecutive subjects (females: 21 %, age: 59+/-9 years). There was only modest correlation between LAV(ECHO) (92+/-31 ml) and LAV(CARTO) (178+/-37 ml) (R=0.57), and RAV(ECHO) (71+/-29 ml) and RAV(CARTO) (173+/-34 ml) (R=0.42), respectively. LAV(ECHO) and RAV(ECHO) underestimated LAV(CARTO) and RAV(CARTO) with the absolute bias (+/-1.96 standard deviation) of -85 (-148; -22) ml and -102 (-169; -35) ml, respectively, and with the relative bias of -48 (-75; -21) % and -59 (-88; -30) %, respectively (all P<0.000001 for their mutual difference). Significant confounders of this difference were not identified. In patients with LSPAF, 2D echocardiography significantly underestimated both LA and RA volumes as compared with electroanatomic reference. This disagreement was independent of clinical, echocardiographic and mapping characteristics.
- MeSH
- chirurgie s pomocí počítače metody MeSH
- dospělí MeSH
- echokardiografie metody MeSH
- fibrilace síní diagnostické zobrazování patologie chirurgie MeSH
- katetrizační ablace metody MeSH
- lidé středního věku MeSH
- lidé MeSH
- mapování potenciálů tělesného povrchu metody MeSH
- reprodukovatelnost výsledků MeSH
- retrospektivní studie MeSH
- senioři MeSH
- senzitivita a specificita MeSH
- srdeční síně diagnostické zobrazování patologie chirurgie MeSH
- velikost orgánu MeSH
- výsledek terapie MeSH
- Check Tag
- dospělí MeSH
- lidé středního věku MeSH
- lidé MeSH
- mužské pohlaví MeSH
- senioři MeSH
- ženské pohlaví MeSH
- Publikační typ
- časopisecké články MeSH
- MeSH
- kardiologie * metody zákonodárství a právo MeSH
- kardiostimulace umělá MeSH
- lidé MeSH
- mimotělní oběh metody přístrojové vybavení MeSH
- oxygenátory MeSH
- podpůrné srdeční systémy klasifikace MeSH
- srdeční elektrofyziologie klasifikace metody MeSH
- telemedicína metody MeSH
- Check Tag
- lidé MeSH
- Publikační typ
- přehledy MeSH