SHDB-AF: a Japanese Holter ECG database of atrial fibrillation

Abstract

Saitama Heart Database Atrial Fibrillation (SHDB-AF) is a novel open-sourced Holter ECG database from Japan, containing data from 122 unique subjects with paroxysmal atrial fibrillation. Among the 128 recordings, 98 contain raw ECG data with rhythm annotations at the beat level, manually performed by a cardiology fellow. The remaining recordings consist only of ECG traces without annotations. The dataset was collected as part of a study evaluating the generalization performance of a deep learning atrial fibrillation event detection model across different distribution shifts.

Background

Atrial fibrillation (AF) is a prevalent atrial arrhythmia that reduces quality of life and leads to complications such as embolic stroke and heart failure. Recent progress in machine learning and deep learning (DL) has demonstrated the potential to improve diagnostic accuracy significantly. Ensuring that DL models are robust and applicable across diverse factors such as ethnicity, age, and sex is crucial.

Despite the availability of several ECG databases to the research community, such as MITDB [1, 2], AFDB [2, 3], LTAFDB [2, 4], IRIDIA-AF [5], Icentia11k [2, 6, 7], and CPSC2021 [2, 8], none of them incorporate a sample from the Japanese population.

The SHDB-AF dataset was collected in previous work to evaluate the generalization performance of a DL algorithm for AF event detection from beat-to-beat time intervals, termed ArNet2 [9, 10], under different distribution shifts. A second DL model named RawECGNet [11] was developed and benchmarked against ArNet2. In contrast to ArNet2, RawECGNet was developed based on the raw, single-lead ECG signal. Regarding F1 score, ArNet2 achieved an F1 score of 0.92 for SHDB using beat-to-beat time intervals and 0.93 for RawECGNet using the raw ECG signal. Further results are presented in associated papers [10, 11].

Methods

As previously mentioned, data provided in this project was initially collected for the purpose of evaluating generalization performance of a DL algorithm for AF detection denoted ArNet2, [9, 10]. Inclusion criteria and preprocessing steps applied to SHDB-AF, are described bellow.

Data collection

This database includes ECG recordings of adult subjects who underwent Holter monitoring as ordered by their treating physician from May 2019 to May 2023. Holters were recorded using Fukuda Holter monitor and digitized at 125Hz with two leads recorded, modified CC5 and NASA leads. Each recording lasts approximately 24 hours. While no reference beat annotations were extracted, each recording includes a diagnosis based on the free-text medical report prepared following the patient's examination.

Data preparation

Inclusion criteria

A total of 145 Holter recordings were collected, of which 128 were included in the current database. Seventeen recordings were excluded due to duplicates, missing clinical information, or low signal quality. Of the 128 included recordings, 98 were associated with 93 unique subjects and were annotated by certified cardiologists.

Following the methodology outlined in Biton et al. [10], approximately 100 recordings from each database were re-annotated by a cardiology fellow (MA). These subsets were stratified based on age, sex, and AF diagnosis. Specifically, 80 of these recordings came from subjects diagnosed with AF, as per the diagnosis recorded for each individual recording. Further details on the inclusion criteria for these selected recordings are provided in Biton et al. [10].

The remaining 30 recordings were included to complete the dataset, though rhythm annotations are not available for these recordings.

Preprocessing

Subsequently, all recordings underwent filtering using a zero-phase second-order infinite impulse response bandpass filter with a passband of [0.67 - 100] Hz [12] to eliminate baseline wander and high-frequency noise. Following this, the recordings were resampled to 200 Hz using an anti-aliasing filter. Beat annotations were then identified using the epltd implementation of the Pan and Tompkins algorithm [13].

Annotation protocol

A total of 98 recordings underwent manual beat-level annotation. To ensure patient confidentiality under HIPAA/GDPR regulations, cardiologists accessed a secure server remotely for the annotation process. Specifically, the PhysioZoo software [14, 15] was used to annotate the recordings. An annotation protocol was established, focusing on categorizing supraventricular arrhythmias including (1) AF, (2) Atrial Flutter, (3) Atrial tachycardia, and (4) other supraventricular tachycardias such as Wolf-Parkinson-White and intranodal tachycardias. Normal sinus rhythm and other rhythms were not annotated. MA spent an estimated average of 45 minutes per 24-hour Holter recording for annotation. A detailed description of the re-annotation protocol can be found in Biton et al. [10].

De-identification

Patient identifiers were anonymized following HIPAA guidelines and local regulations. Each participant was assigned a random 3-digit unique identifier ranging from 000 to 143. Numbers with fewer than three digits were padded with zeros to maintain consistent length. Dates directly associated with participants were randomly shifted by at least one year.

Data Description

ECG data

All ECG recordings are stored in the WFDB format. Each ECG file (.dat suffix) contains two channels: 'ECG1', representing modified CC5 leads, and 'ECG2', representing NASA leads.

Rhythm annotations

Rhythm annotations are provided at the beat level in separate annotation files (.atr). These files categorize rhythms into five types:

1. (AFIB (atrial fibrillation);
2. (AFL (atrial flutter);
3. (AT (atrial tachycardia);
4. (PAT (Other supraventricular tachycardias such as Wolf-Parkinson-White) and (NOD (intranodal tachycardias); and
5. (N (other, such as NSR, that were not labeled).

In the annotation files, rhythm annotations are listed under the 'aux_note' parameter. Each rhythm mark within the annotation files indicates the start of a rhythm interval. The appearance of a subsequent rhythm mark marks the beginning of a new interval.

The shortest interval lasts 2.5 seconds, while the longest extends to 24 hours, with a median duration of 47.5 seconds (Q2-Q3 range: 17.0 to 270.25 seconds).

Below are the counts for each annotated beat by rhythm and the overall intervals.

R-peak annotations are provided in files with the .qrs suffix.

Clinical data

The AdditionalData.csv file provides comprehensive clinical and demographic information for each subject, extracted from their Electronic Medical Records (EMR). The following table summarizes key characteristics of the dataset, offering an overview of patient demographics, medical history, and AF-related details. 

The dataset includes the following variables, providing further context on patient conditions and Holter recordings:

• <Subject_ID>: The unique identifier for the subject.
• <Data_ID>: A three-digit identifier assigned to each Holter recording.
• <Annotated>: Indicates whether the Holter recording includes rhythm annotations.
• <Height>: The patient's height as recorded in the EMR, in meters.
• <Weight>: The patient's weight as recorded in the EMR, in kg.
• <BMI>: The patient's Body Mass Index (BMI), calculated based on the height and weight in the EMR.
• <Date_Holter>: The date on which the Holter recording was made.
• <Holter_start_time>: Represents the relative time of day for each recording, formatted as hh:mm AM/PM, indicating the start time of the recording within a 24-hour period.
• <Holter_recording_length>: Indicates the total duration of each recording, presented in the format hh:mm:ss.
• <Indication_Holter>: The reason for the Holter recording, as documented in the medical report.
• <Age_at_Holter>: The subject's age in years at the time of the Holter recording.
• <Sex>: The subject's gender, categorized as either "male" or "female."
• <AF_Type>: The final diagnosis of the subject based on the medical report following the Holter examination.
• <Previously_Documented_AFL>: Indicates whether atrial flutter was documented prior to the Holter recording.
• <Previous_AF_Ablation>: Indicates whether the subject had undergone catheter ablation for atrial fibrillation prior to the Holter examination.
• <PPM_on_Holter>/<PPM_after_Holter>: Indicates whether the subject had a pacemaker implanted during or after the Holter recording.
• <PPM_Indication>: The reason for pacemaker implantation, if applicable.
• <PPM_Date>: The date on which the pacemaker was implanted, if applicable.
• <Date_of_First_Diagnosis_of_AF_AFL>: The date when the subject was first diagnosed with atrial fibrillation or atrial flutter.
• <AF_Duration_Months>: The duration of atrial fibrillation in months, from the first diagnosis of atrial fibrillation/atrial flutter until the Holter recording.
• <Antiarrhythmic_Drug_nonBB>/<BB>/<Anticoagulation>: Lists any permanent medications taken by the subject prior to the Holter recording, including Beta-Blockers (BB), non-BB antiarrhythmic drugs, and anticoagulants.
• <Date_of_1st_AF_Ablation>/<Ablation1_PVI>/<Ablation1_CTI>/<Ablation1_Others>/<Date_Redo_AF_Ablation>/<Redo_Detail>: Information regarding any atrial fibrillation ablation procedures, including the date of the first procedure, the type of ablation (PVI, CTI, or others), and details on any redo ablations.
• <Echo_Date>/<Echo_LAD>/<Echo_LVEF>/<Echo_LV_Asynergy>: Details about any echocardiogram performed on the subject, including the date of the echo, left atrial diameter (LAD), left ventricular ejection fraction (LVEF), and left ventricular asynergy.
• <Moderate_or_Severe_MR>/<Moderate_or_Severe_TR>/<Moderate_or_Severe_AS>/<Moderate_or_Severe_AR>: Indication of moderate and severe heart valve conditions present in the subject, including Mitral Regurgitation (MR), Tricuspid Regurgitation (TR), Aortic Stenosis (AS), and Aortic Regurgitation (AR).
• <CHF>/<HTN>/<DM>/<Vascular_Diseases>: Indications of relevant comorbidities, including Congestive Heart Failure (CHF), Hypertension (HTN), Diabetes Mellitus (DM), and Vascular Disease (Vascular_Diseases).
• <Stroke>: Indicates if the subject had a stroke prior to the Holter recording.
• <Comments>: Miscellaneous textual information not covered in the other columns.

This additional dataset helps to enrich the analysis by providing crucial context on each subject's medical background, diagnosis, and treatment history, facilitating more detailed research and insights from the Holter recording data.

Each file name corresponds uniquely to a Data_ID, specifically matching the 'Data_ID' field in the AdditionalData.csv.

The file included, example.png, provides an illustration of an ECG example.

Usage Notes

This dataset has been partially used in the publications "Generalizable and robust deep learning algorithm for atrial fibrillation diagnosis across geography, ages, and sexes" and "RawECGNet: Deep Learning Generalization for Atrial Fibrillation Detection From the Raw ECG". It serves as a benchmark for developing and validating machine learning and deep learning models for AF detection and classification. The combination of raw ECG signals with detailed patient metadata enables research into AI-driven AF detection and risk stratification, allowing models to assess patient-specific risk factors based on comorbidities, medication use, and clinical history. Additionally, the dataset supports feature extraction and biomarker discovery by identifying ECG-based indicators associated with AF episodes, contributing to improved risk assessment and early diagnosis. It is also valuable for benchmarking AF detection algorithms, studying signal processing techniques for noise reduction and waveform analysis, and evaluating model generalization across diverse patient populations.

While the dataset provides a rich set of ECG and clinical data, it remains relatively small, with 128 ECG traces, and is derived from a single institution. Additionally, the dataset does not include systematic longitudinal follow-ups with repeated ECGs or survival data over time. Despite these limitations, it remains a valuable resource for AI-driven AF research and predictive modeling. Further details about the dataset can be found in our associated paper [16].

Release Notes

Version 1.0.0: Initial release on August 12, 2024.

Version 1.0.1: SHDB-AF v1.0.1 was released on December 29, 2024.

This update includes the following:

• Removal of a subset of Data_IDs identified as duplicates. Future data releases will exclude the following recordings: 016 and 030.
• Addition of 30 patient recordings without rhythm annotations.
• Expansion of available data elements in AdditionalData.csv. Additional information for subjects included in SHDB-AF now includes data from the EMR and retrieved information from the day of the Holter recording.
• As recordings from multiple subjects have been identified, the following modifications and additions were made to the AdditionalData.csv file:
  • A Data_ID column has been added to provide a unique identifier for each recording, replacing the previous 'Study ID' column.
  • A Subject_ID column has been added which contains subject ID identifiers, to link the various Data_IDs to their corresponding unique subjects.
  • A Holter_start_time column has been added, and this information has been removed from the ECG .dat file.
  • A Holter_recording_length column has been added, and this information has been removed from the ECG .dat file.
  • Additionally, the following column names have been updated for clarity:
    • 'Age' has been changed to 'Age at Holter',
    • 'Dx' has been changed to 'AF_\type',
    • 'AFL' has been changed to 'Previously_Documented_AFL',
    • 'Previous ablation' has been changed to 'Previous_AF_Ablation',
    • 'Pacemaker' has been changed to 'PPM_on_Holter',
    • 'AAD' has been extended to three separate columns. It has been renamed and divided into the following categories: <Antiarrhythmic_Drug_nonBB>, <BB>, and <Anticoagulation>.

Ethics

The authors declare no ethics concerns. The present work was approved by the institutional ethics committee at Saitama Medical University International Medical Center (IRB number 2023-145). Due to retrospective and descriptive nature of the study, written informed consent was waived.

Acknowledgements

Hittman: Technion EVPR Fund: Hittman Family Fund.

Conflicts of Interest

The authors have no conflicts of interest to declare.

References

1. Moody GB, Mark RG. The impact of the MIT-BIH arrhythmia database. IEEE engineering in medicine and biology magazine. 2001 May;20(3):45-50.
2. Goldberger AL, Amaral LA, Glass L, Hausdorff JM, Ivanov PC, Mark RG, Mietus JE, Moody GB, Peng CK, Stanley HE. PhysioBank, PhysioToolkit, and PhysioNet: components of a new research resource for complex physiologic signals. circulation. 2000 Jun 13;101(23):e215-20.
3. Moody G. A new method for detecting atrial fibrillation using RR intervals. Proc. Comput. Cardiol.. 1983;10:227-30.
4. Petrutiu S, Sahakian AV, Swiryn S. Abrupt changes in fibrillatory wave characteristics at the termination of paroxysmal atrial fibrillation in humans. Europace. 2007 Jul 1;9(7):466-70.
5. Gilon C, Grégoire JM, Mathieu M, Carlier S, Bersini H. IRIDIA-AF, a large paroxysmal atrial fibrillation long-term electrocardiogram monitoring database. Scientific data. 2023 Oct 18;10(1):714.
6. Tan, S., Ortiz-Gagné, S., Beaudoin-Gagnon, N., Fecteau, P., Courville, A., Bengio, Y., & Cohen, J. P. (2022). Icentia11k Single Lead Continuous Raw Electrocardiogram Dataset (version 1.0). PhysioNet.
7. Tan S, Androz G, Chamseddine A, Fecteau P, Courville A, Bengio Y, Cohen JP. Icentia11k: An unsupervised representation learning dataset for arrhythmia subtype discovery. arXiv preprint arXiv:1910.09570. 2019 Oct 21.
8. Wang X, Ma C, Zhang X, Gao H, Clifford GD, Liu C. Paroxysmal atrial fibrillation events detection from dynamic ecg recordings: The 4th china physiological signal challenge 2021. Proc. PhysioNet. 2021:1-83.
9. Chocron A, Oster J, Biton S, Mandel F, Elbaz M, Zeevi YY, Behar JA. Remote atrial fibrillation burden estimation using deep recurrent neural network. IEEE Transactions on Biomedical Engineering. 2020 Dec 4;68(8):2447-55.
10. Biton S, Aldhafeeri M, Marcusohn E, Tsutsui K, Szwagier T, Elias A, Oster J, Sellal JM, Suleiman M, Behar JA. Generalizable and robust deep learning algorithm for atrial fibrillation diagnosis across geography, ages and sexes. NPJ Digital Medicine. 2023 Mar 17;6(1):44.
11. Ben-Moshe N, Tsutsui K, Biton S, Zvuloni E, Sörnmo L, Behar JA. RawECGNet: Deep Learning Generalization for Atrial Fibrillation Detection From the Raw ECG. IEEE Journal of Biomedical and Health Informatics. 2024 May 24.
12. Kligfield P, Gettes LS, Bailey JJ, Childers R, Deal BJ, Hancock EW, Van Herpen G, Kors JA, Macfarlane P, Mirvis DM, Pahlm O. Recommendations for the standardization and interpretation of the electrocardiogram: part I: the electrocardiogram and its technology: a scientific statement from the American Heart Association Electrocardiography and Arrhythmias Committee, Council on Clinical Cardiology; the American College of Cardiology Foundation; and the Heart Rhythm Society endorsed by the International Society for Computerized Electrocardiology. Circulation. 2007 Mar 13;115(10):1306-24.
13. Pan J, Tompkins WJ. A real-time QRS detection algorithm. IEEE transactions on biomedical engineering. 1985 Mar(3):230-6.
14. Behar JA, Levy J, Zvuloni E, Gendelman S, Rosenberg A, Biton S, Derman R, Sobel JA, Alexandrovich A, Charlton PH, Goda MÁ. PhysioZoo: The Open Digital Physiological Biomarkers Resource. In2023 Computing in Cardiology (CinC) 2023 Oct 1 (Vol. 50, pp. 1-4). IEEE.
15. Gendelman S, Biton S, Derman R, Zvuloni E, Levy J, Lugassy S, Alexandrovich A, Behar JA. PhysioZoo ECG: Digital electrocardiography biomarkers to assess cardiac conduction. In2021 Computing in Cardiology (CinC) 2021 Sep 13 (Vol. 48, pp. 1-4). IEEE.
16. Tsutsui K, Brimer SB, Ben-Moshe N, Sellal JM, Oster J, Mori H, Ikeda Y, Arai T, Nakano S, Kato R, Behar JA. SHDB-AF: a Japanese Holter ECG database of atrial fibrillation. arXiv preprint arXiv:2406.16974. 2024 Jun 22.