Systems and methods for enhanced post-ablation atrial fibrillation monitoring

Abstract

A method may obtain a sample electrocardiogram (ECG) from a mobile ECG device, wherein the sample ECG is associated with a patient. A method may provide at least a portion of the sample ECG to a trained machine learning model. A method may determine an atrial fibrillation (AF) recurrence risk. A method may, based on the AF recurrence risk meeting or exceeded a predetermined risk threshold, transmitting a recommendation to increase ECG monitoring of the patient associated with the sample ECG.

Description

Docket No. 2025-005-02SYSTEMS AND METHODS FOR ENHANCED POST-ABLATION ATRIAL FIBRILLATION MONITORINGCROSS-REFERENCE TO RELATED APPLICATIONS

[0001] The present application claims priority to and the benefit of United States Provisional Patent Application 63 / 777,550 filed March 25, 2025 and titled SYSTEMS AND METHODS FOR ENHANCED POST-ABLATION ATRIAL FIBRILLATION MONITORING, the disclosure of which is incorporated herein by reference in its entirety.BACKGROUND

[0002] Atrial fibrillation (AF) is the most common arrhythmia, and it significantly increases the risk of ischemic stroke and congestive heart failure development. Over 59 million cases of AF or atrial flutter exist worldwide, with individuals aged 45 or older facing a lifetime risk of approximately one in three to five. Catheter ablation is a cornerstone treatment for AF. Typically, the success of catheter ablation is measured by the absence of AF recurrence during the follow-up period, with a recurrence rate generally ranging from 20% to 50% over one to five years. Common conventional monitoring strategies for AF patients include clinical 12-lead electrocardiogram (ECG) tracing, 24 to 72-hour Holter-monitoring, implantable loop recorder (ILR). However, a higher proportion of post-ablation patients with recurring AF post-ablation tend to experience paroxysmal AF, and many remain asymptomatic, making conventional monitoring strategies potentially inadequate as they might fail to detect episodic occurrences.SUMMARY

[0003] In some aspects, the techniques described herein relate to a method of monitoring atrial fibrillation, the method including: obtaining a sample electrocardiogram (ECG) from a mobile ECG device, wherein the sample ECG is associated with a patient; providing at least a portion of the sample ECG to a trained machine learning model; determining an atrial fibrillation (AF) recurrence risk; and based on the AF recurrence risk meeting or exceeded a predetermined risk threshold, transmitting a recommendation to increase ECG monitoring of the patient associated with the sample ECG.

[0004] This summary is provided to introduce a selection of concepts that are further described below in the detailed description. This summary is not intended to identify keyDocket No. 2025-005-02or essential features of the claimed subject matter, nor is it intended to be used as an aid in limiting the scope of the claimed subject matter.

[0005] Additional features and aspects of embodiments of the disclosure will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by the practice of such embodiments. The features and aspects of such embodiments may be realized and obtained by means of the instruments and combinations particularly pointed out in the appended claims. These and other features will become more fully apparent from the following description and appended claims or may be learned by the practice of such embodiments as set forth hereinafter.BRIEF DESCRIPTION OF THE DRAWINGS

[0006] In order to describe the manner in which the above-recited and other features of the disclosure can be obtained, a more particular description will be rendered by reference to specific embodiments thereof which are illustrated in the appended drawings. For better understanding, the like elements have been designated by like reference numbers throughout the various accompanying figures. While some of the drawings may be schematic or exaggerated representations of concepts, non-schematic drawings should be considered as being to scale for some embodiments of the present disclosure, but not to scale for other embodiments contemplated herein. Understanding that the drawings depict some example embodiments, the embodiments will be described and explained with additional specificity and detail through the use of the accompanying drawings in which:

[0007] FIG. 1 is system diagram of a system of atrial fibrillation monitoring, according to at least some embodiments of the present disclosure.

[0008] FIG. 2 is a flowchart illustrating a method of preparing input dataset, according to at least some embodiments of the present disclosure.

[0009] FIG. 3 is a flowchart illustrating a method of atrial fibrillation monitoring, according to at least some embodiments of the present disclosure.

[0010] FIG. 4 is a flowchart illustrating another method of atrial fibrillation monitoring, according to at least some embodiments of the present disclosure.

[0011] FIG. 5 is a flowchart illustrating yet another method of atrial fibrillation monitoring, according to at least some embodiments of the present disclosure.Docket No. 2025-005-02DETAILED DESCRIPTION

[0012] The present disclosure relates generally to methods and systems for atrial fibrillation (AF) monitoring. More particularly, methods and systems for providing enhanced post-ablation treatment (AF) monitoring. In some embodiments, systems and methods according to the present disclosure obtain post-ablation treatment electrocardiogram (ECG) strips from a mobile ECG device and classify a detected sinus rhythm as a benign or severe sinus rhythm. In some embodiments, systems and methods according to the present disclosure can determine increased risk of AF recurrence in patients that have previously received catheter ablation (CA) treatment. In some embodiments, the method includes classifying obtained ECG samples into a plurality of categories, where those classified as exhibiting Sinus Rhythm are further evaluated for an AF recurrence risk. The method further includes identifying an ECG sample with an AF recurrence risk that meets or exceeds a predetermined threshold value, and alerting either the patient, a medical provider, or both. The patient may be further monitored by increasing monitoring using the same ECG device or a different ECG device based on the determined heightened risk of AF recurrence.

[0013] FIG. 1 illustrated an embodiment of a system 100 used to perform at least some methods according to the present disclosure. The system 100 includes a mobile ECG device 102. In some embodiments, the mobile ECG device 102 is a single-lead device. In some embodiments, the mobile ECG device 102 is a dedicated electronic device. In some embodiments, the mobile ECG device 102 is in data communication with a local computing device 104. In some examples, the local computing device 104 is a smartphone, a smartwatch, a wearable electronic device, a laptop computer, and a tablet computer. In some embodiments, the mobile ECG device 102 is integrated with the local computing device 104. For example, the mobile ECG device 102 may integrated into a smartwatch or other wearable electronic device including an electrical sensor. In some embodiments, the mobile ECG device 102 is integrated into a local computing device 104, and the integrated computing device / mobile ECG device 102 is further in data communication with a second local computing device 104. For example, a smartwatch including the mobile ECG device 102 may be in data communication with a local smartphone.

[0014] In some embodiments, the data communication between the mobile ECG device 102 and the local computing device 104 is a wired communication, such as a universal serial bus (USB) connection or Ethernet connection. In some embodiments, the dataDocket No. 2025-005-02communication between the mobile ECG device 102 and the local computing device 104 is a wireless communication, such as Bluetooth, BLE, WiFi, etc.

[0015] In some embodiments, the mobile ECG device 102 and / or the local computing device 104 (integrated or in local data communication) includes a processor 106 and a hardware storage device 108. The hardware storage device 108 has instructions stored thereon that, when executed by the processor 106, cause the mobile ECG device 102 and / or the local computing device 104 to perform at least a portion of an embodiment of a method described herein. In some embodiments, the mobile ECG device 102 and / or the local computing device 104 performs an entire method described herein. In some embodiments, at least a portion of the method is performed at a remote computing device 110 that is not located locally to the mobile ECG device 102 and / or the local computing device 104. For example, the remote computing device 110 may be a server computer or other remote computing device that is in data communication with the mobile ECG device 102 and / or the local computing device 104 via a network 112.

[0016] FIG. 2 is a flowchart illustrating a method 214 of training a machine learning (ML) model. All training ECG data was collected from the DECAAF-II Trial. The DECAAF-II is a prospective, randomized, multicenter trial conducted across 44 sites on three continents, comparing PVI alone with PVI plus MRI-guided fibrosis ablation. A total of 843 patients were randomized, with 422 assigned to the PVI arm and 421 to the PVI plus MRI-guided fibrosis ablation arm. Rhythm monitoring was achieved through daily handheld single-lead smartphone-based electrocardiography (ECG) recordings using the ECG Check Device (Cardiac Designs Inc.), complemented by Holter monitoring and 12-lead ECGs obtained during routine clinical care. A total of 161,518 (191 strips per patient on average) singlelead ECG strips were recorded daily after the blanking period from 843 patients and were analyzed by trained experts that identified up to 18 labeled rhythm types. The ECG strips collected from the first day after the blanking period post ablation to 541 days after ablation. Each strip is recorded at 30 seconds with a sampling rate of 200 Hz. The primary end point of the study was the first confirmed recurrence of atrial arrhythmia (including AF, atrial flutter, or atrial tachycardia) lasting for at least 30 seconds after the 90-day blanking period, demonstrated by at least 2 consecutive 1-lead smartphone ECG device tracings, 1 positive reading on a clinical 12-lead ECG tracing, ambulatory monitor, or if the patient underwent repeat ablation treatment.

[0017] In some embodiments, the training instances 216 of the training dataset are cleaned by applying a bandpass filter 218 to remove baseline wandering. For example, the bandpassDocket No. 2025-005-02filter may have a filter frequency range of 1-4 Hz, three orders, and a 2-decibel ripple. In some embodiments, a min-max normalization 220 is performed after bandpass filtering 218. The training instances 216 may be de-noised 222 before being provided to the model 224.

[0018] In some embodiments, the model 224 is trained in two stages. The first stage differentiates between two types of sinus rhythms - “Benign Sinus” and “Severe Sinus” -using a deep convolutional neural network model. ECG strips were categorized using the same grouping approach, with the “Benign Sinus” group consisting of strips from patients without the primary outcome throughout the study and the “Severe Sinus” group including strips from patients with the primary outcome and those within ±3 days of any AF ECG strips collected from the same patients. To reduce variability and remove noise from the beginning and end of the recording, each ECG strip was truncated to a portion of the ECG training instance in the middle 20 seconds, standardizing the length of each training instance to 4,000 data points. Due to the significant imbalance between the two groups, a randomly selected subset of Benign Sinus ECG strips was extracted to match the number of ECG strips in the Severe Sinus group at a 1:1 ratio. Before model training, a pre-trained deep convolutional neural network model was developed using unused denoised ECGs for the same classification task. The final classification was performed using denoised ECG signals as input to the pre-trained model.

[0019] The second stage focused on extracting clinically meaningful ECG features to delineate sinus rhythm variations using statistical analyses. Denoised ECG strips were processed with the following features were extracted: Heart Rate, Heart Rate Variability (Root Mean Square of Successive Differences, RMSSD), R-peak prominence, R-peak to R-peak duration, P-wave duration, and T-wave duration. For each ECG strip, the mean and standard deviation of all features were calculated. ECG strips classified as sinus rhythm after 90 days post-ablation were included, while those with other or multiple diagnoses were excluded. To ensure comparability, all features from individual ECG strips were averaged at the patient level for statistical analysis.

[0020] FIG. 3 is a flowchart illustrating a method 326 of monitoring AF using the model 224 or a similar model, according to at least some embodiments of the present disclosure. The method 326 includes obtaining a sample ECG from a mobile ECG device, wherein the sample ECG is associated with a patient, at 328. In some embodiments, the sample ECG has a duration of no less than 30 seconds. In some embodiments, the sample ECG has a duration greater than 30 seconds to allow a truncated portion to be selected from the sampleDocket No. 2025-005-02ECG with the truncated portion of the sample ECG has a duration no less than 30 seconds. In some embodiments, the truncated portion of the sample ECG has a duration no less than 20 seconds. In some embodiments, the truncated portion of the sample ECG has a duration no less than 15 seconds. In some embodiments, the sample ECG is one of a plurality of sample ECGs collected by the mobile ECG device from the patient. For example, the patient may record a plurality of sample ECG strips daily, semi-daily, bi-daily, weekly, etc.

[0021] As described herein, the mobile ECG device mobile ECG device is, in some embodiments, a single-lead device. In some embodiments, the mobile ECG device is a dedicated electronic device. In some embodiments, the mobile ECG device is in data communication with a local computing device. In some examples, the local computing device is a smartphone, a smartwatch, a wearable electronic device, a laptop computer, and a tablet computer. In some embodiments, the mobile ECG device is integrated with the local computing device. For example, the mobile ECG device may integrate into a smartwatch or other wearable electronic device including an electrical sensor. In some embodiments, the mobile ECG device is integrated into a local computing device, and the integrated computing device / mobile ECG device is further in data communication with a second local computing device. For example, a smartwatch including the mobile ECG device may be in data communication with a local smartphone.

[0022] In some embodiments, the data communication between the mobile ECG device and the local computing device is a wired communication, such as a universal serial bus (USB) connection or Ethernet connection. In some embodiments, the data communication between the mobile ECG device and the local computing device is a wireless communication, such as Bluetooth, BLE, WiFi, etc.

[0023] In some embodiments, the method 326 includes providing at least a portion of the sample ECG to a trained ML model at 330. In some embodiments, the portion has a duration less than the full sample ECG duration. For example, the portion may be selected from the middle of the full sample ECG duration. In some embodiments, the portion has a duration at least half that of the full sample ECG duration. In some embodiments, the portion has the start of the full sample ECG duration excluded. In some embodiments, the portion has the end of the full sample ECG duration excluded. In some embodiments, the portion has a lower maximum amplitude range. For example, the sample ECG may be passed through a bandpass filter to reduce the amplitude range. In other examples, the amplitude of the sample ECG and / or portion of the sample ECG may be normalized according to the minimum and maximum values.Docket No. 2025-005-02

[0024] The portion of the sample ECG is subsequently provided to a trained ML model, such as that described in relation to FIG. 2. For example, the trained ML model may be trained on Benign and Severe Sinus Rhythm, and the trained ML model may differentiate between Sinus Rhythms associated with AF ECG strips based on the training instances provided thereto during training.

[0025] In some embodiments, the method 326 further includes determining an atrial fibrillation (AF) recurrence risk at 332. In some embodiments, determining the AF recurrence risk includes receiving an output from the trained ML model. The trained model may output a numerical correlation between the sample ECG or portion thereof and one or more AF recurrence risk factors. For example, the sample ECG or portion thereof may be used and input to compared to the embedded risk factor(s) in the trained ML model. The AF recurrence risk may be output as a numerical value, such as between 0 and 1, and the AF recurrence risk may be compared to a predetermined threshold value.

[0026] Based on the AF recurrence risk meeting or exceeded a predetermined risk threshold, the method 326 further includes transmitting a recommendation to increase ECG monitoring of the patient associated with the sample ECG at 334. In some embodiments, the recommendation is transmitted to the patient associated with the sample ECG. For example, the recommendation may be transmitted to a phone number or email address associated with the patient. In other examples, the recommendation may be displayed to the patient on a display of the mobile ECG device and / or a display of a local computing device in data communication with the mobile ECG device.

[0027] In some embodiments, the recommendation is transmitted to a medical provider of the patient associated with the sample ECG. For example, the recommendation may be transmitted to a phone number or email address associated with the medical provider.

[0028] The recommendation may include instructions or be correlated to instructions for the increased ECG monitoring. For example, the increased ECG monitoring may include increasing a frequency of future sample ECG collection, such as increasing from weekly to daily or from daily to twice daily. In other examples, the increased ECG monitoring may include collecting future sample ECGs with an increased sample duration. In other examples, the increased ECG monitoring may include collecting future sample ECGs with a different ECG device, such as an ECG device with a greater quantity of leads.

[0029] In some embodiments, a method of monitoring AF using the model 224 or a similar model, according to at least some embodiments of the present disclosure, further includes classifying the sample ECG prior to providing the sample ECG or portion of the sampleDocket No. 2025-005-02ECG to the trained ML model. FIG. 4 is a flowchart illustrating an embodiment of a method 426 of monitoring AF with classification.

[0030] The method 426 includes obtaining a sample ECG from a mobile ECG device, wherein the sample ECG is associated with a patient, at 428, similar to as described in relation to FIG. 3. In some embodiments, the sample ECG has a duration of no less than 30 seconds. In some embodiments, the sample ECG has a duration greater than 30 seconds to allow a truncated portion to be selected from the sample ECG with the truncated portion of the sample ECG has a duration no less than 30 seconds. In some embodiments, the truncated portion of the sample ECG has a duration no less than 20 seconds. In some embodiments, the truncated portion of the sample ECG has a duration no less than 15 seconds. In some embodiments, the sample ECG is one of a plurality of sample ECGs collected by the mobile ECG device from the patient. For example, the patient may record a plurality of sample ECG strips daily, semi-daily, bi-daily, weekly, etc.

[0031] As described herein, the mobile ECG device mobile ECG device is, in some embodiments, a single-lead device. In some embodiments, the mobile ECG device is a dedicated electronic device. In some embodiments, the mobile ECG device is in data communication with a local computing device. In some examples, the local computing device is a smartphone, a smartwatch, a wearable electronic device, a laptop computer, and a tablet computer. In some embodiments, the mobile ECG device is integrated with the local computing device. For example, the mobile ECG device may integrate into a smartwatch or other wearable electronic device including an electrical sensor. In some embodiments, the mobile ECG device is integrated into a local computing device, and the integrated computing device / mobile ECG device is further in data communication with a second local computing device. For example, a smartwatch including the mobile ECG device may be in data communication with a local smartphone.

[0032] In some embodiments, the data communication between the mobile ECG device and the local computing device is a wired communication, such as a universal serial bus (USB) connection or Ethernet connection. In some embodiments, the data communication between the mobile ECG device and the local computing device is a wireless communication, such as Bluetooth, BLE, WiFi, etc.

[0033] The method 426 includes classifying the sample ECG as one of Normal Sinus Rhythm, AFib, Bradycardia, and Tachycardia at 436. Based on identifying the sample ECG as exhibiting Normal Sinus Rhythm, the method 426 includes providing at least a portion of the sample ECG to a trained ML model at 430 to determine the presence of Benign sinusDocket No. 2025-005-02rhythm and / or severe sinus rhythm. In some embodiments, the portion has a duration less than the full sample ECG duration. For example, the portion may be selected from the middle of the full sample ECG duration. In some embodiments, the portion has a duration at least half that of the full sample ECG duration. In some embodiments, the portion has the start of the full sample ECG duration excluded. In some embodiments, the portion has the end of the full sample ECG duration excluded. In some embodiments, the portion has a lower maximum amplitude range. For example, the sample ECG may be passed through a bandpass filter to reduce the amplitude range. In other examples, the amplitude of the sample ECG and / or portion of the sample ECG may be normalized according to the minimum and maximum values.

[0034] The portion of the sample ECG is subsequently provided to a trained ML model, such as that described in relation to FIG. 2. For example, the trained ML model may be trained on Benign and Severe Sinus Rhythm, and the trained ML model may differentiate between Benign and Severe Sinus Rhythms associated with AF ECG strips based on the training instances provided thereto during training.

[0035] In some embodiments, the method 426 further includes determining an AF recurrence risk at 432 based on the provided portion of the sample ECG. In some embodiments, determining the AF recurrence risk includes reclassifying the sample ECG as a Severe Sinus Rhythm or Benign Sinus Rhythm based on the output of the trained ML model. In some embodiments, determining the AF recurrence risk includes receiving an output from the trained ML model. The trained model may output a numerical correlation between the sample ECG or portion thereof and one or more AF recurrence risk factors. For example, the sample ECG or portion thereof may be vectorized and compared to the embedded risk factor(s) in the trained ML model. The AF recurrence risk may be output as a numerical value, such as between 0 and 1, and the AF recurrence risk may be compared to a predetermined threshold value.

[0036] Based on the AF recurrence risk meeting or exceeded a predetermined risk threshold, the method 426 further includes transmitting a recommendation to increase ECG monitoring of the patient associated with the sample ECG at 434. In some embodiments, the recommendation is transmitted to the patient associated with the sample ECG. For example, the recommendation may be transmitted to a phone number or email address associated with the patient. In other examples, the recommendation may be displayed to the patient on a display of the mobile ECG device and / or a display of a local computing device in data communication with the mobile ECG device.Docket No. 2025-005-02

[0037] In some embodiments, the recommendation is transmitted to a medical provider of the patient associated with the sample ECG. For example, the recommendation may be transmitted to a phone number or email address associated with the medical provider.

[0038] The recommendation may include instructions or be correlated to instructions for the increased ECG monitoring. For example, the increased ECG monitoring may include increasing a frequency of future sample ECG collection, such as increasing from weekly to daily or from daily to twice daily. In other examples, the increased ECG monitoring may include collecting future sample ECGs with an increased sample duration. In other examples, the increased ECG monitoring may include collecting future sample ECGs with a different ECG device, such as an ECG device with a greater quantity of leads.

[0039] FIG. 5 is a flowchart illustrating another embodiment of a method 526 of monitoring AF with classification. The method 526 includes obtaining a sample ECG from a mobile ECG device, wherein the sample ECG is associated with a patient, at 528, similar to as described in relation to FIG. 3. In some embodiments, the sample ECG has a duration of no less than 30 seconds. In some embodiments, the sample ECG has a duration greater than 30 seconds to allow a truncated portion to be selected from the sample ECG with the truncated portion of the sample ECG has a duration no less than 30 seconds. In some embodiments, the truncated portion of the sample ECG has a duration no less than 20 seconds. In some embodiments, the truncated portion of the sample ECG has a duration no less than 15 seconds. In some embodiments, the sample ECG is one of a plurality of sample ECGs collected by the mobile ECG device from the patient. For example, the patient may record a plurality of sample ECG strips daily, semi-daily, bi-daily, weekly, etc.

[0040] As described herein, the mobile ECG device mobile ECG device is, in some embodiments, a single-lead device. In some embodiments, the mobile ECG device is a dedicated electronic device. In some embodiments, the mobile ECG device is in data communication with a local computing device. In some examples, the local computing device is a smartphone, a smartwatch, a wearable electronic device, a laptop computer, and a tablet computer. In some embodiments, the mobile ECG device is integrated with the local computing device. For example, the mobile ECG device may integrate into a smartwatch or other wearable electronic device including an electrical sensor. In some embodiments, the mobile ECG device is integrated into a local computing device, and the integrated computing device / mobile ECG device is further in data communication with a second local computing device. For example, a smartwatch including the mobile ECG device may be in data communication with a local smartphone.Docket No. 2025-005-02

[0041] In some embodiments, the data communication between the mobile ECG device and the local computing device is a wired communication, such as a universal serial bus (USB) connection or Ethernet connection. In some embodiments, the data communication between the mobile ECG device and the local computing device is a wireless communication, such as Bluetooth, BLE, WiFi, etc.

[0042] In some embodiments, the portion has a duration less than the full sample ECG duration. For example, the portion may be selected from the middle of the full sample ECG duration. In some embodiments, the portion has a duration at least half that of the full sample ECG duration. In some embodiments, the portion has the start of the full sample ECG duration excluded. In some embodiments, the portion has the end of the full sample ECG duration excluded. In some embodiments, the portion has a lower maximum amplitude range. For example, the sample ECG may be passed through a bandpass filter to reduce the amplitude range. In other examples, the amplitude of the sample ECG and / or portion of the sample ECG may be normalized according to the minimum and maximum values.

[0043] In some embodiments, the method 526 includes denoising the sample ECG or the portion of the sample at 538. In some embodiments, denoising the sample ECG occurs before a portion is selected from the sample ECG. In some embodiments, a portion is selected from the sample ECG before the portion is denoised. Denoising the sample and / or portion includes, in some embodiments, a de-noising Al model to improve signal clarity and reduce artifacts in ECG data, ensuring high-quality waveform analysis.

[0044] The portion of the sample ECG is subsequently provided to a trained ML model, such as that described in relation to FIG. 2. For example, the trained ML model may be trained on Benign and Severe Sinus Rhythm, and the trained ML model may differentiate between Sinus Rhythms associated with AF ECG strips based on the training instances provided thereto during training.

[0045] In some embodiments, the method 526 further includes determining an atrial fibrillation (AF) recurrence risk at 532. In some embodiments, determining the AF recurrence risk includes receiving an output from the trained ML model. The trained model may output a numerical correlation between the sample ECG or portion thereof and one or more AF recurrence risk factors. For example, the sample ECG or portion thereof may be vectorized and compared to the embedded risk factor(s) in the trained ML model. The AF recurrence risk may be output as a numerical value, such as between 0 and 1, and the AF recurrence risk may be compared to a predetermined threshold value.Docket No. 2025-005-02

[0046] Based on the AF recurrence risk meeting or exceeded a predetermined risk threshold, the method 526 further includes transmitting a recommendation to increase ECG monitoring of the patient associated with the sample ECG at 534. In some embodiments, the recommendation is transmitted to the patient associated with the sample ECG. For example, the recommendation may be transmitted to a phone number or email address associated with the patient. In other examples, the recommendation may be displayed to the patient on a display of the mobile ECG device and / or a display of a local computing device in data communication with the mobile ECG device.

[0047] In some embodiments, the recommendation is transmitted to a medical provider of the patient associated with the sample ECG. For example, the recommendation may be transmitted to a phone number or email address associated with the medical provider.

[0048] The recommendation may include instructions or be correlated to instructions for the increased ECG monitoring. For example, the increased ECG monitoring may include increasing a frequency of future sample ECG collection, such as increasing from weekly to daily or from daily to twice daily. In other examples, the increased ECG monitoring may include collecting future sample ECGs with an increased sample duration. In other examples, the increased ECG monitoring may include collecting future sample ECGs with a different ECG device, such as an ECG device with a greater quantity of leads. It should be understood that portions of embodiments of methods may be combinable, such as the embodiment of a method 426 with classification of FIG. 4 with the denoising 538 described in relation to FIG. 5.

[0049] The present disclosure relates generally to methods and systems for providing enhanced post-ablation atrial fibrillation monitoring according to any of the clauses herein:

[0050] Clause 1. A method of monitoring atrial fibrillation, the method comprising: obtaining a sample electrocardiogram (ECG) from a mobile ECG device, wherein the sample ECG is associated with a patient; providing at least a portion of the sample ECG to a trained machine learning model; determining an atrial fibrillation (AF) recurrence risk; and based on the AF recurrence risk meeting or exceeded a predetermined risk threshold, transmitting a recommendation to increase ECG monitoring of the patient associated with the sample ECG.

[0051] Clause 2. The method of any preceding clause, wherein the sample ECG is obtained from a patient within one year of a catheter ablation (CA).

[0052] Clause 3. The method of any preceding clause, wherein the entire method is performed on the mobile ECG device.Docket No. 2025-005-02

[0053] Clause 4. The method of any preceding clause, wherein the mobile ECG device is in data communication with a remote computing device via a network, and determining the AF recurrence risk is performed at the remote computing device.

[0054] Clause 5. The method of any preceding clause, wherein the mobile ECG device is selected from a group consisting of a smartphone, a smartwatch, a wearable electronic device, a laptop computer, and a tablet computer.

[0055] Clause 6. The method of any preceding clause, wherein the mobile ECG device includes a single-lead device in local data communication with one or more of a smartphone, a smartwatch, a wearable electronic device, a laptop computer, and a tablet computer.

[0056] Clause 7. The method of any preceding clause, wherein the sample ECG is one of a plurality of daily sample ECGs of the patient.

[0057] Clause 8. The method of any preceding clause, wherein the sample ECG is no less than 30 seconds.

[0058] Clause 9. The method of any preceding clause, wherein the portion of the sample ECG is no less than 30 seconds.

[0059] Clause 10. The method of any preceding clause, wherein the AF recurrence risk includes comparing the sample ECG to a second sample ECG obtained at a prior time.

[0060] Clause 11. The method of any preceding clause, wherein the mobile ECG device is a single-lead ECG device.

[0061] Clause 12. The method of any preceding clause, further comprising identifying a Normal Sinus Rhythm in the sample ECG prior to providing at least the portion of the sample ECG to the trained machine learning model.

[0062] Clause 13. The method of clause 12, wherein identifying the Normal Sinus Rhythm further includes classifying the sample ECG as one of a Normal Sinus Rhythm, AFib, Bradycardia, and Tachycardia.

[0063] Clause 14. The method of any preceding clause, further comprising de-noising at least the portion of the sample ECG.

[0064] Clause 15. The method of any preceding clause, wherein transmitting a recommendation to increase ECG monitoring of the patient associated with the sample ECG includes transmitting the recommendation to a medical provider.

[0065] Clause 16. The method of any preceding clause, wherein transmitting a recommendation to increase ECG monitoring of the patient associated with the sample ECG includes transmitting the recommendation to the patient.Docket No. 2025-005-02

[0066] Clause 17. The method of any preceding clause, wherein transmitting a recommendation to increase ECG monitoring of the patient associated with the sample ECG includes transmitting a recommendation to obtain future sample ECGs with a different ECG device.

[0067] Clause 18. The method of any preceding clause further comprising obtaining future sample ECGs from the patient at an increased sampling rate with the mobile ECG device.

[0068] Clause 19. A system comprising: a mobile ECG device; and a computing device in data communication with the mobile ECG device, wherein the computing device includes: a processor, and a hardware storage device in communication with the processor and having instructions stored thereon that, when executed by the processor, cause the computing device to perform the method of any preceding clause.

[0069] Clause 20. The system of clause 19, wherein the mobile ECG device is a singlelead device.

[0070] It should be understood that references to “one embodiment” or “an embodiment” of the present disclosure are not intended to be interpreted as excluding the existence of additional embodiments that also incorporate the recited features. For example, any element described in relation to an embodiment herein may be combinable with any element of any other embodiment described herein, to the extent such features are not described as being mutually exclusive. Numbers, percentages, ratios, or other values stated herein are intended to include that value, and also other values that are “about”, “substantially”, or “approximately” the stated value, as would be appreciated by one of ordinary skill in the art encompassed by embodiments of the present disclosure. A stated value should therefore be interpreted broadly enough to encompass values that are at least close enough to the stated value to perform a desired function or achieve a desired result. The stated values include at least the variation to be expected in a suitable manufacturing or production process, and may include values that are within 5%, within 1%, within 0.1%, or within 0.01% of a stated value.

[0071] The terms “approximately,” “about,” and “substantially” as used herein represent an amount close to the stated amount that is within standard manufacturing or process tolerances, or which still performs a desired function or achieves a desired result. For example, the terms “approximately,” “about,” and “substantially” may refer to an amount that is within less than 5% of, within less than 1% of, within less than 0.1% of, and within less than 0.01% of a stated amount. Further, it should be understood that any directions or reference frames in the preceding description are merely relative directions or movements.Docket No. 2025-005-02For example, any references to “up” and “down” or “above” or “below” are merely descriptive of the relative position or movement of the related elements.

[0072] A person having ordinary skill in the art should realize in view of the present disclosure that equivalent constructions do not depart from the spirit and scope of the present disclosure, and that various changes, substitutions, and alterations may be made to embodiments disclosed herein without departing from the spirit and scope of the present disclosure. Equivalent constructions, including functional “means-plus-function” clauses are intended to cover the structures described herein as performing the recited function, including both structural equivalents that operate in the same manner, and equivalent structures that provide the same function. It is the express intention of the applicant not to invoke means-plus-function or other functional claiming for any claim except for those in which the words ‘means for’ appear together with an associated function. Each addition, deletion, and modification to the embodiments that falls within the meaning and scope of the claims is to be embraced by the claims. The described embodiments are therefore to be considered as illustrative and not restrictive, and the scope of the disclosure is indicated by the appended claims rather than by the foregoing description.

Claims

Docket No. 2025-005-02CLAIMSWhat is claimed is:

1. A method of monitoring atrial fibrillation, the method comprising:obtaining a sample electrocardiogram (ECG) from a mobile ECG device, wherein the sample ECG is associated with a patient;providing at least a portion of the sample ECG to a trained machine learning model;determining an atrial fibrillation (AF) recurrence risk; andbased on the AF recurrence risk meeting or exceeded a predetermined risk threshold, transmitting a recommendation to increase ECG monitoring of the patient associated with the sample ECG.

2. The method of claim 1, wherein the sample ECG is obtained from a patient within one year of a catheter ablation (CA).

3. The method of claim 1, wherein the method is performed entirely on the mobile ECG device.

4. The method of claim 1, wherein the mobile ECG device is in data communication with a remote computing device via a network, and determining the AF recurrence risk is performed at the remote computing device.

5. The method of claim 1, wherein the mobile ECG device is selected from a group consisting of a smartphone, a smartwatch, a wearable electronic device, a laptop computer, and a tablet computer.

6. The method of claim 1, wherein the mobile ECG device includes a single-lead device in local data communication with one or more of a smartphone, a smartwatch, a wearable electronic device, a laptop computer, and a tablet computer.

7. The method of claim 1, wherein the sample ECG is one of a plurality of daily sample ECGs of the patient.Docket No. 2025-005-028. The method of claim 1, wherein the sample ECG is no less than 30 seconds.

9. The method of claim 1, wherein the portion of the sample ECG is no less than 30 seconds.

10. The method of claim 1, wherein the AF recurrence risk includes comparing the sample ECG to a second sample ECG obtained at a prior time.

11. The method of claim 1, wherein the mobile ECG device is a single-lead ECG device.

12. The method of claim 1, further comprising identifying a Normal Sinus Rhythm in the sample ECG prior to providing at least the portion of the sample ECG to the trained machine learning model.

13. The method of claim 12, wherein identifying the Normal Sinus Rhythm further includes classifying the sample ECG as one of a Normal Sinus Rhythm, AFib, Bradycardia, and Tachycardia.

14. The method of claim 1, further comprising de-noising at least the portion of the sample ECG.

15. The method of claim 1, wherein transmitting a recommendation to increase ECG monitoring of the patient associated with the sample ECG includes transmitting the recommendation to a medical provider.

16. The method of claim 1, wherein transmitting a recommendation to increase ECG monitoring of the patient associated with the sample ECG includes transmitting the recommendation to the patient.

17. The method of claim 1, wherein transmitting a recommendation to increase ECG monitoring of the patient associated with the sample ECG includes transmitting a recommendation to obtain future sample ECGs with a different ECG device.Docket No. 2025-005-0218. A system comprising:a mobile electrocardiogram (ECG) device; anda computing device in data communication with the mobile ECG device, wherein the computing device includes:a processor, anda hardware storage device in communication with the processor and having instructions stored thereon that, when executed by the processor, cause the computing device to:obtain a sample ECG from the mobile ECG device, wherein the sample ECG is associated with a patient;provide at least a portion of the sample ECG to a trained machine learning model;determine an atrial fibrillation (AF) recurrence risk; and based on the AF recurrence risk meeting or exceeded a predetermined risk threshold, transmit a recommendation to increase ECG monitoring of the patient associated with the sample ECG.

19. The system of claim 18, wherein the mobile ECG device is a single-lead device.

20. A non-transitory computer readable medium (CRM) having instructions stored thereon that, when executed by a processor, performs the steps of:obtaining a sample electrocardiogram (ECG) from a mobile ECG device, wherein the sample ECG is associated with a patient;providing at least a portion of the sample ECG to a trained machine learning model;determining an atrial fibrillation (AF) recurrence risk; andbased on the AF recurrence risk meeting or exceeded a predetermined risk threshold, transmitting a recommendation to increase ECG monitoring of the patient associated with the sample ECG.

Images