Bicomb QT interval prediction

Abstract

Embodiments of the present invention provide a mobile electrocardiogram (ECG) sensor comprising an electrode assembly including an electrode, wherein the electrode assembly senses a cardiac related signal when in contact with a user's body and generates an electrical signal representative of the sensed cardiac related signal. The ECG sensor further comprises a processing device operably coupled to the electrode assembly, the processing device providing the sensed cardiac related signal to a machine learning module trained to predict a 12-lead QT interval (QTc) value from a mobile ECG sensor comprising less than 12 leads. The ECG sensor further comprises a housing housing the electrode assembly and the processing device.

Description

[0001] Cross-references to related applications

[0002] This application claims the benefit of U.S. Provisional Application No. 63 / 044,882, entitled “TWO-LEAD QT INTERVAL PREDICTION”, filed on June 26, 2020, the contents of which are incorporated herein by reference in their entirety. Background Technology

[0003] It is estimated that more than 23 million people will die from cardiovascular disease each year by 2030. Cardiovascular disease is prevalent in populations of both first- and third-world countries, regardless of socioeconomic status. Monitoring cardiovascular function can aid in the treatment and prevention of cardiovascular disease. For example, Holter monitors or other flow electrocardiogram (FEC) devices can be used to monitor patients with A-fib (or other types of arrhythmias) for extended periods to manage the condition. Such devices can, for example, continuously monitor the electrical activity of the cardiovascular system for at least 24 hours. This monitoring can be crucial in the detection of conditions such as acute coronary syndrome (ACS).

[0004] In mammals, the heart generates and conducts electrical currents that signal and initiate coordinated contractions of the heart. In humans, these electrical signals are generated by a part of the heart called the SA segment. After being generated by the SA segment, the current travels through the myocardium in a predictable manner within a healthy heart.

[0005] Generally, an electrocardiogram (ECG) is a graphical representation of the heart's electrical conduction over time as it is projected onto the body's surface. ECGs are typically displayed on a graph with x and y axes. Typically, the x-axis of an ECG shows time, while the y-axis shows the potential (millivolts) of the current conducted through the heart during normal cardiac function. Attached Figure Description

[0006] The novel features of the invention are specifically set forth in the appended claims. The features and advantages of the invention will be better understood by referring to the following detailed description, which illustrates illustrative embodiments using the principles of the invention, and the accompanying drawings:

[0007] Figure 1 This is an illustration of a prior art electrocardiograph according to some embodiments of the present invention, which has 10 electrodes positioned on the patient's body for acquiring a prior art 12-lead electrocardiogram.

[0008] Figure 2 This is an illustration of the chest according to some embodiments of the present invention, showing an example of electrode placement on the chest for acquiring a prior art 12-lead electrocardiogram.

[0009] Figure 3 An example lead I is shown according to some embodiments of the invention, which is labeled to show the PQRST wave generated by a 12-lead electrocardiograph.

[0010] Figure 4 An example 12-lead electrocardiogram in a conventional format is shown according to some embodiments of the present invention.

[0011] Figure 5A An exemplary ECG sensing device is shown according to some embodiments of the present invention.

[0012] Figure 5B According to some embodiments of the present invention Figure 5A Hardware block diagram of the ECG sensing device.

[0013] Figure 5C Examples of the use of [some embodiments of the invention] are illustrated. Figure 5A The ECG sensing device is used to measure the user's limb leads.

[0014] Figure 5D and Figure 5E Examples of the use of [some embodiments of the invention] are illustrated. Figure 5A The ECG sensing device measures the user's precordial leads.

[0015] Figure 6 This is a flowchart of a method for predicting QT intervals according to some embodiments of the present invention.

[0016] Figure 7 This is a block diagram of an example computing device that can perform one or more of the operations described herein, according to some embodiments of the present invention. Detailed Implementation

[0017] It should be understood that the invention, in its application, is not limited to the details of the construction, experiments, exemplary data, and / or arrangement of components set forth in the following description. Embodiments of the invention can have other embodiments or can be practiced or performed in various ways. Furthermore, it should be understood that the terminology used herein is for descriptive purposes and should not be considered limiting.

[0018] An electrocardiogram (ECG) provides numerous ECG waveforms representing the electrical activity of a person's heart. An ECG monitoring device may include an electrode assembly for recording these ECG waveforms of a patient's heart (also referred to herein as "ECG acquisition"). The electrode assembly may be placed on the patient's skin at multiple sites, and the electrical signals (ECG waveforms) recorded between the electrode pairs within the electrode assembly can be referred to as leads. Different numbers of leads can be used to acquire an ECG, and various numbers and combinations of electrodes can be used to form a variety of leads. Example numbers of leads used for ECG acquisition are 3 leads, 5 leads, and 12 leads.

[0019] Figure 1 This is an illustration of the 10 electrodes of a standard ECG sensing device placed on a patient to obtain a standard 12-lead ECG. The electrode placed on the right arm is generally called RA. The electrode placed on the left arm is called LA. The RA and LA electrodes are placed in the same locations on the left and right arms, preferably, but not necessarily, near the wrist. The leg electrodes may be referred to as RL for the right leg and LL for the left leg. The RL and LL electrodes are placed in the same locations on the left and right legs, preferably, but not necessarily, near the ankle.

[0020] Figure 2 This example illustrates the placement of six electrodes on the chest, labeled V1, V2, V3, V4, V5, and V6. V1 is placed in the fourth intercostal space, for example, between ribs 4 and 5, just to the right of the sternum. V2 is placed in the fourth intercostal space, for example, between ribs 4 and 5, just to the left of the sternum. V3 is placed in the fifth intercostal space, midway between electrodes V2 and V4. V4 is placed in the fifth intercostal space between ribs 5 and 6, on the left midclavicular line. V5 is placed horizontally and level with V4 on the left anterior axillary line. V6 is placed horizontally and level with V4 and V5 on the left middle axillary line.

[0021] The electrocardiograph then calculates and outputs waveforms in three limb leads. Limb leads I, II, and III are bipolar leads with one positive and one negative terminal. Lead I is the voltage between the left arm (LA) and the right arm (RA), for example, I = LA - RA. Lead II is the voltage between the left leg (LL) and the right arm (RA), for example, II = LL - RA. Lead III is the voltage between the left leg (LL) and the left arm (LA), for example, III = LL - LA. Leads I, II, and III are collectively referred to as "limb leads".

[0022] Unipolar leads also have two poles; however, the negative pole is a complex pole composed of signals from multiple other electrodes. In a standard electrocardiograph used to obtain a 12-lead ECG, all leads except the limb leads are unipolar (aVR, aVL, aVF, V1, V2, V3, V4, V5, and V6). Enhanced limb leads (aVR, aVL, and aVF) observe the heart from different angles (or vectors) and are determined based on the potential difference between one of RA, LA, and LL and a complex composed of two of RA, LA, and LL. Therefore, three electrodes located at RA, LA, and LL will simultaneously sense aVR, aVL, and aVF based on the above relationship. That is, while leads I, II, and III each require input from only two electrodes, aVR, aVL, and aVF may require input from three electrodes located at RA, LA, and LL.

[0023] For example, the enhanced right vector (aVR) positions the positive electrode on the right arm, while the negative electrode is a combination of the left arm and left leg electrodes, which "enhances" the signal strength of the positive electrode on the right arm. Therefore, the enhanced right vector (aVR) equals RA - (LA + LL) / 2 or -(I + II) / 2. The enhanced left vector (aVL) equals LA - (RA + LL) / 2 or (I - II) / 2. The enhanced left vector (aVF) equals LL - (RA + LA) / 2 or (II - I) / 2.

[0024] In one embodiment, six electrodes on the patient's chest are close enough to the heart that they do not require enhancement. A central terminal, called Wilson's (often labeled CT), is used. W V W The composite electrode (or WCT) is used as the negative terminal. Wilson's central terminal is generated by connecting the electrodes RA, LA, and LL together through a simple resistive network to give an average potential on the body that approximates the potential at infinity (i.e., zero). Wilson's central terminal WCT is calculated as (RA + LA + LL) / 3.

[0025] ECG waveforms (each corresponding to a lead in the ECG) recorded by an ECG monitoring device can include data corresponding to the electrical activity of the human heart. A typical heartbeat can include several potential changes, which can be classified into waves and wave groups, including P waves, QRS complexes, T waves, and U waves, as known in the art. In other words, as known in the art, individual ECG waveforms can include P waves, QRS complexes, T waves, and U waves, etc. The shape and duration of these waves can be correlated with various characteristics of the human heart, such as the size of the atria (e.g., indicating atrial enlargement) and can be the primary source of human-specific heartbeat characteristics. Individual waves or wave groups (i.e., QRS complexes) are associated with different phases of cardiac depolarization and repolarization. ECG waveforms can be analyzed for various indicators useful in detecting cardiac events or states, such as arrhythmia detection and characterization (typically after standard filtering and "cleaning" of the signal). For example, such indicators may include ECG waveform amplitude and morphology (e.g., QRS complex amplitude and morphology), R-wave-ST segment and T-wave amplitude analysis, and heart rate variability (HRV).

[0026] Figure 3Example lead I is shown, labeled to illustrate the P, QRS, and T waves / synapses generated by a 12-lead ECG. Typically, a normal heartbeat ECG has predictable waveforms across all 12 ECG leads. The portion of the ECG between two waves is called a segment, while the portion between more waves is called an interval. For example, the portion of the ECG between the end of the S wave (part of the QRS complex) and the beginning of the T wave is called the ST segment, and the portion of the ECG between the beginning of the Q wave (part of the QRS complex) and the end of the T wave is called the QT interval.

[0027] Figure 4 This shows an example of a traditional 12-lead electrocardiogram. Figure 4 As shown, for a standard ECG waveform trace, 12 ECG leads are displayed on the X and Y axes, respectively, where the Y-axis represents time and the X-axis represents voltage. In these traces, all 12 ECG waveforms are aligned relative to their X-axis. That is, the P, QRS, and T waveforms of all leads appear simultaneously along the X-axis of their respective traces. For example, in a conventional ECG waveform trace, if a QRS complex appears at 1 second on the X-axis of lead I, then a QRS complex will also appear at 1 second in each of the other eleven ECG waveforms (i.e., leads II, III, aVR, aVL, aVF, V1, V2, V3, V4, V5, and V6).

[0028] The standard time alignment format allows healthcare providers to more easily derive information from the twelve sensed ECG waveforms. In a conventional ECG trace, time alignment is achieved by simultaneously sensing waveforms using ten electrodes of a conventional ECG, all of which are simultaneously positioned on the skin of the individual being sensed. In other words, because all twelve ECG leads of a conventional ECG are sensed simultaneously, time alignment is achieved by simply displaying all waveforms together on the same axis.

[0029] It should be noted that sets of two or more leads can be analyzed to derive information to generate a complete 12-lead ECG. This transformation can be performed using machine learning models (e.g., neural networks, deep learning techniques, etc.). Machine learning models can be trained using 12-lead ECG data corresponding to a group of individuals. Before being input into the machine learning model, the data can be preprocessed in a manner suitable for the application to filter it. For example, before using it to train one or more machine learning models, the data can be categorized according to height, gender, weight, nationality, etc., so that one or more models are fine-tuned for specific types of individuals. In another embodiment, the machine learning model can be further trained based on the user's own ECG data to fine-tune and personalize the model, thereby further reducing any residual synthesis errors.

[0030] While a standard 12-lead electrocardiogram (ECG) provides very useful information about an individual's heart health and condition, standard ECG equipment is expensive, and the procedure is often unavailable outside of hospitals and doctors' offices. Therefore, monitoring is frequently not possible even in first-world countries, and ECG machines may not be accessible in poorer parts of the world.

[0031] Figure 5A The images show upper and lower surface views of an exemplary ECG sensing device 500 including an assembly of electrodes 502 (also referred to as an electrode assembly) according to some embodiments of the present invention. In some embodiments, one or more capacitive electrodes are used in the ECG sensing device 500, such that, for example, the capacitive electrodes sense potentials through clothing worn on a user's body. Similarly, a conductive spray or gel may be placed on a user's body, such that a typical electrode senses potentials through clothing worn on the user's body.

[0032] In one embodiment, the ECG sensing device 500 is constructed wholly or partially of stainless steel or some other suitable material. In one embodiment, the ECG device 500 includes an external coating, such as titanium nitride or other suitable coating. Advantageously, such a material can increase biocompatibility and optimize electrode characteristics.

[0033] In one embodiment, device 500, referred to herein as a mobile computing device, includes all necessary components for sensing, recording, and displaying ECG signals and performing analysis. In another embodiment, device 500 is connected via a wired or wireless connection to a separate mobile computing device (e.g., computing device 550). In this case, device 500 can sense ECG signals and transmit unmodified or modified signals to the mobile computing device for further analysis and / or display. In yet another embodiment, any combination of the two examples above is possible. For example, while ECG sensing device 500 can be considered a standalone mobile computing device capable of performing all the operations described herein, ECG sensing device 500 can still connect to and interact with a second mobile computing device for any suitable purpose (offloading processing / analysis, display, etc.).

[0034] ECG sensing device 500 may include one or more controls and / or indicators. For example, device 500 may include buttons, dials, etc., to select functions (e.g., turn ECG reading on / off, start sending ECG information, etc.). ECG sensing device 500 may also include a display showing the recorded ECGs.

[0035] The ECG sensing device 500 may include a housing 520, wherein two electrodes 502A and 502B are positioned on the upper surface of the housing 520, and a third electrode 502C is positioned on the lower surface of the housing 520, as shown below. Figure 5A As shown. Electrodes 502 may be insulated from each other via dielectric 504 or other suitable materials, enabling them to sense and record different signals. In some embodiments, electrodes 502 may consist of silver-silver chloride (or some other suitable material) electrodes. In some embodiments, ECG sensing device 500 may include electrode connectors (not shown), for example, female sockets on one end or side that allow one or more ECG electrodes to be connected to ECG sensing device 500 for use on skin with or without adhesives (e.g., conductive gel and electrodes 502).

[0036] Figure 5B The illustrated hardware block diagram of ECG sensing device 500 includes hardware such as processing device 505 (e.g., processor, central processing unit (CPU)), memory 510 (e.g., random access memory (RAM), storage device (e.g., hard disk drive (HDD)), solid-state drive, etc.) and other hardware devices (e.g., analog-to-digital converter (ADC), etc.). The storage device may include persistent storage capable of storing data. Persistent storage may be a local storage unit or a remote storage unit. Persistent storage may be a magnetic storage unit, optical storage unit, solid-state storage unit, electronic storage unit (main memory), or similar storage unit. Persistent storage may also be a monolithic / single device or a distributed collection of devices. In some embodiments, processing device 505 may include a dedicated ECG waveform processing and analysis chip that provides built-in lead detachment detection. ECG sensing device 500 may include an ADC (not shown) with a sufficiently high sampling frequency for accurately converting the ECG waveform measured by the set of electrodes 502 into a digital signal (e.g., a 24-bit ADC operating at 500 Hz or higher) for processing by processing device 505.

[0037] The memory 510 may include a lead synthesis software module 510A (hereinafter referred to as module 510A) and a QT prediction software module 510B (hereinafter referred to as module 510B). The processing device 205 may execute module 207A to synthesize ECG waveforms corresponding to leads not measured by the electrodes of the ECG sensing device 500, as discussed in further detail herein. The processing device 505 may execute module 510B to accurately predict the user's QT interval, as discussed in further detail herein.

[0038] The ECG sensing device 500 may also include a transceiver 508, which may implement any suitable protocol for wirelessly transmitting ECG data to one or more local and / or remote computing devices (e.g., computing device 550). For example, the transceiver 508 may include Bluetooth for transmitting ECG data to a local computing device (e.g., a user's laptop or smartphone) via Bluetooth. TM Chip. In other embodiments, transceiver 508 may include (or be coupled to) a network interface device configured to connect to a cellular data network (e.g., using GSM, GSM+EDGE, CDMA, quad-band, or other cellular protocols) or a WiFi network (e.g., 802.11 protocol) to transmit ECG data to a remote computing device (e.g., a physician's or healthcare provider's computing device) and / or a local computing device.

[0039] As discussed in further detail herein, computing device 550 may be used to provide instructions for operating ECG sensing device 500, or may correspond to a healthcare provider system to which ECG data measured by ECG sensing device 500 will be sent.

[0040] like Figure 5C As shown, in a practical example, the user holds the device with one or both hands, with each hand contacting electrodes 502A and 502B on the ECG sensing device 500, while the left leg contacts electrode 502C. The ECG sensing device 500 (optionally having a separate motion computing device) can then be used to record leads I, II, and III, from which at least three additional leads can be determined (e.g., by execution module 510A), as described further in detail herein. Specifically, leads I, II, and III can be used to determine enhanced leads aVR, aVL, and aVF. The user can be sitting, standing, or in any comfortable position.

[0041] Figure 5D and Figure 5EThe following embodiment illustrates an example where a user can also use the ECG sensing device 500 as described herein to record precordial leads V1, V2, V3, V4, V5, and V6. The user can hold the ECG sensing device 500 such that each of the user's hands contacts electrodes 502A and 502B, while a third electrode (e.g., 502C) remains against the chest, thus contacting one of the six precordial chest locations designated "CP1", "CP2", "CP3", "CP4", "CP5", and "CP6". For example, the user can begin by positioning the ECG sensing device 500 such that electrode 502C contacts CP1, and from there the user can move the ECG sensing device 500 so that it sequentially contacts each of the six electrode locations corresponding to leads V2, V3, V4, V5, and V6. In some embodiments, when a user touches electrodes 502A and 502B of the ECG sensing device 500 with their right and left hands respectively, and simultaneously rests against the third electrode (e.g., 502C) of the position holding device 500 on their chest corresponding to V1, V2, V3, V4, V5, and V6, potentials sensed at chest positions corresponding to V1, V2, V3, V4, V5, and V6, as well as potentials sensed at LA and RA, are simultaneously sensed. Lead I is equivalent to the potential difference between LA and RA. Therefore, in some embodiments, measuring the potential at a chest position corresponding to any one of V1, V2, V3, V4, V5, and V6, and the potentials at the LA and RA positions, is equivalent to the potential difference between the chest position and lead I. That is, for example, using all three electrodes of the device 500 as described above, V1 (potential at the chest position of V1) = ("CP1") - WCT (WCT = (RA + LA) + LL) / 3 or (lead I + lead II) / 3).

[0042] The six precordial leads can be designated as ("CP1", "CP2", "CP3", "CP4", "CP5", and "CP6"), and the composite value is called Wilson's central terminal ("WCT"). "CP(x)" corresponds to any of the six potentials sensed at the anatomical precordial lead location (where "x" is the location number 1-6). For example, CP1 is the ECG measurement sensed at the site where electrodes are placed to measure V1, and this location is approximately in the second intercostal space immediately to the right of the sternum. Therefore, lead V1 = CP1 - WCT.

[0043] WCT is equal to one-third or one-third (RA+LA+LL) of the sum of the potentials sensed in the right upper limb, left upper limb, and left lower leg. In a standard ECG using ten electrodes placed simultaneously, the WCT value is generated simultaneously with the sensing of the precordial leads because RA, LA, and LL, which determine WCT, are sensed simultaneously with CPI, CP2, CP3, CP4, CP5, and CP6.

[0044] In these embodiments, electrode 502 is positioned and configured to simultaneously sense / calculate six limb leads, namely leads I, II, III, aVR, aVL and aVF, when the user touches the first electrode 502A with the right upper limb, the second electrode 502B with the left upper limb and the third electrode 502C with the left lower limb.

[0045] As described herein, the ECG sensing device 500 is configured to sequentially sense six leads V1, V2, V3, V4, V5, and V6 when a user, for example, touches the first electrode 502A with his or her right upper limb, the second electrode 502B with his or her left upper limb, and the third electrode 502C with an area of ​​his or her chest corresponding to the location of the precordial leads.

[0046] In some embodiments of the ECG sensing device 500 described herein, which includes three electrodes, the RA, LA, and LL for determining the WCT are not simultaneously sensed with one or more precordial leads. That is, when one of the three electrodes of the ECG sensing device 500 is held against the user's chest wall, only two electrodes remain free, and the conventional WCT cannot be determined simultaneously. In some of these embodiments, RA is set to 0. When RA=0, WCT=(0+LA+LL) / 3 or ((LA-0)+(LL-0)) / 3 is provided, which can be further expressed as WCT=(Lead I+Lead II) / 3.

[0047] Similarly, in these embodiments, where RA is set to 0, the average WCT = (average lead I + average lead II) / 3. In some embodiments, the average WCT is generated using average lead I and average lead II, which are generated using, for example, an integrated averaging method of the lead I and lead II waveforms sensed by the ECG sensing device described herein. Generating the average WCT is advantageous, for example, in signal filtering, and also simplifies value alignment for subtraction purposes. That is, in some embodiments, the individual CPI, CP2, CP3, CP4, CP5, and CP6 are averaged, and the average WCT is subtracted from each to generate V1, V2, V3, V4, V5, and V6, respectively.

[0048] A complete 12-lead set can also be synthesized from the lead set measured by the ECG sensing device 500 using various machine learning (ML) methods. Compared to training examples used to train the model, ML is well-suited for continuously monitoring one or more criteria to identify large and small anomalies or trends in the input data. The ML models described herein can be trained on user data from user groups and / or on other training examples to suit the model's design needs. ML models that can be used with the embodiments described herein, as examples and not limitations, include: Bayes, Markov, Gaussian processing, clustering algorithms, generative models, kernels, and neural network algorithms. Some embodiments utilize machine learning models based on trained neural networks (e.g., trained recurrent neural networks (RNNs) or trained convolutional neural networks (CNNs)).

[0049] For example, ML models can leverage artificial neural networks (ANNs) for supervised classification, where the model's output represents the probability of an input sample in a particular class of data or that it exhibits some special characteristic. In another example, a data-driven approach based on convolutional neural networks (CNNs) is used. By using convolutional operations, ML models can infer individual output data points by considering the correlations between input samples at temporally closed intervals. More specifically, a single output sample (individual precordial leads) at a general time t is influenced by all input samples (all limb leads) from t–τ to t+τ. The value of τ, representing the network's receptive field, is highly dependent on the model architecture and typically increases with its depth (i.e., the number of consecutive layers). For all data-driven approaches, the ability to generalize to unseen data and avoid overfitting is crucial. Complex models and small datasets may result in excellent performance on the training set but poor performance on unseen data. Models can be optimized using any appropriate regularization methods, such as inter-layer and intra-layer normalization (e.g., batch normalization and layer normalization) and data augmentation techniques. Finally, to improve the effectiveness and efficiency of the model, the use of residual connections can be utilized, that is, the use of identity mappings that allow gradients to flow through layers during backpropagation of gradient-based optimization algorithms.

[0050] Using AI / deep learning with multi-lead ECG sensors can allow patients to monitor their heart's electrical activity themselves (in a hospital or at home) without hospital access or bulky hardware.

[0051] In some embodiments, the memory 510 of the ECG sensing device 500 or other mobile computing device (e.g., computing device 550) may include an instruction software module (not shown) that displays or otherwise sends instructions to the individual instructing how the user should position the ECG sensing device 500 for an ECG (e.g., in a standard precordial lead position) and the position the user should be in for the ECG. For example, displaying an image of a region on the user's chest instructs the user to hold the third electrode against that region while holding electrode 1 and electrode 2 with their left and right hands, respectively.

[0052] In some embodiments, software on the ECG sensing device 500 or computing device 550 is configured to identify whether the first electrode is touched by the left hand and the second electrode by the right hand, and vice versa. For example, in some embodiments, the third electrode is positioned on a different surface of the ECG sensing device 500 than the first and second electrodes, such that a user might need to switch hand positions after their left leg has been in contact with the third electrode to align their precordial lead position on their chest with the third electrode. In some embodiments, software on the ECG sensing device 500 or other mobile computing device receives information from sensors coupled to or integrated with the ECG sensing device 500, wherein the sensors provide information about the device's position in space. Examples of sensor classes that sense this information include, but are not limited to, accelerometers, inclinometers, and gyroscopes.

[0053] In some embodiments, the ECG sensing device 500 is configured to sense an ECG when one or more of the electrodes 502 are not engaged by the user. For example, in some embodiments, the ECG sensing device 500 includes three electrodes, and the ECG sensing device 500 is configured to sense an ECG when all three electrodes are engaged by the user or when any two of the three electrodes are engaged by the user. That is, in this embodiment, the ECG sensing device senses an ECG when the user, for example, contacts the skin surface of their right upper limb with the first electrode and the skin surface of their left upper limb with the second electrode but not the third electrode. In this example, lead I is sensed when two of the three electrodes are engaged by the right upper limb and the left upper limb, respectively. Similarly, lead II is sensed when two of the three electrodes are engaged by the right upper limb and the left lower limb, respectively. Similarly, lead III is sensed when two of the three electrodes are engaged by the left upper limb and the left lower limb, respectively. In this embodiment, the ECG sensing device 500 identifies that the user is not touching one or more electrodes while the user is touching two or more electrodes by sensing electrode potentials from two or more electrodes that are being touched, for example, but not from electrodes that have never been touched by the user.

[0054] In some embodiments of the ECG sensing device described herein, exemplary embodiments thereof are as follows: Figures 5A-5E As shown, a mobile computing device (e.g., computing device 550) is configured to run software applications as described herein. In other embodiments, the mobile computing device includes one or more hardware central processing unit (CPU) or general-purpose graphics processing unit (GPGPU) that performs the functions of the device. In further embodiments, the mobile computing device also includes an operating system configured to execute executable instructions. In some embodiments, the mobile computing device is optionally connected to a computer network. In another embodiment, the mobile computing device is optionally connected to the Internet, enabling it to access the World Wide Web. In a further embodiment, the mobile computing device is optionally connected to a cloud computing infrastructure. In other embodiments, the mobile computing device is optionally connected to an intranet. In other embodiments, the mobile computing device is optionally connected to a data storage device.

[0055] Based on the description herein, suitable mobile computing devices include server computers, desktop computers, laptop computers, notebook computers, sub-notebook computers, netbook computers, internet-enabled pad computers, handheld computers, smartphones, smartwatches, digital wearable devices, and tablet computers.

[0056] In some embodiments, the mobile computing device includes an operating system configured to execute executable instructions. For example, an operating system is software that includes programs and data, manages the device's hardware, and provides services for the execution of applications. Non-limiting examples of suitable operating systems include FreeBSD, OpenBSD, and NetBSD. ® Linux, Apple ® Mac OS X Server ® Oracle ® Solaris ® Windows Server ® and Novell ® NetWare ® Those skilled in the art will recognize, by way of non-limiting example, that suitable personal computer operating systems include Microsoft... ® Windows ® Apple ® Mac OSX ® UNIX ® and such as GNU / Linux ® A UNIX-like operating system. In some embodiments, the operating system is provided by cloud computing.

[0057] In some embodiments, a mobile computing device includes a storage and / or memory device. A storage and / or memory device is one or more physical devices for temporarily or permanently storing data or programs. In some embodiments, the device is volatile memory and requires power to maintain the stored information. In some embodiments, the device is non-volatile memory and retains the stored information when the mobile computing device is not powered. In other embodiments, non-volatile memory includes flash memory. In some embodiments, non-volatile memory includes dynamic random access memory (DRAM). In some embodiments, non-volatile memory includes ferroelectric random access memory (FRAM). In some embodiments, non-volatile memory includes phase-change random access memory (PRAM). In other embodiments, the device is a storage device, and non-limiting examples include CD-ROMs, DVDs, flash memory devices, disk drives, magnetic tapes, optical disc drives, and cloud-based storage. In other embodiments, the storage and / or memory device is a combination of devices such as those disclosed herein.

[0058] In some embodiments, the mobile computing device includes a display for transmitting visual information to a user. In some embodiments, the mobile computing device includes an input device for receiving information from a user. In some embodiments, the input device is a keyboard. In some embodiments, the input device is a pointing device, and non-limiting examples include a mouse, trackball, trackpad, joystick, game controller, or stylus. In some embodiments, the input device is a touchscreen or multiple touchscreens. In other embodiments, the input device is a microphone for capturing voice or other sound input. In other embodiments, the input device is a camera or other sensor for capturing motion or visual input. In further embodiments, the input device is a combination of devices such as those disclosed herein.

[0059] In various embodiments, the platforms, systems, media, and methods described herein include cloud computing environments. In some embodiments, a cloud computing environment includes multiple computing processors.

[0060] It should be understood that Figures 5A-5E Exemplary embodiments illustrating the user scenarios described herein are shown. Typically, numerous electrode positions, shapes, and sizes are available in the devices described herein, allowing for comfortable and natural contact with the electrodes by an individual. For example, all three electrodes may be positioned entirely on the side of the computing device or its cover.

[0061] exist Figures 5A-5E In any of the embodiments shown, one or more electrodes may be configured to be removable from the ECG sensing device 500. In these embodiments, the ECG sensing device has, for example, a male or female connector configured to snap-fit ​​couple to a corresponding male or female connector on the removable electrode.

[0062] and Figures 5A-5E The embodiments shown herein illustrate an ECG sensing device comprising three electrodes. It should be understood that other numbers of ECG electrodes may be incorporated into the ECG sensing device described herein.

[0063] In general, any of the technologies, components, and / or subsystems described above can be used or combined with any other examples. For example, any ECG device described herein may include any of the features described above.

[0064] Whether secondary to genetic and / or acquired risk factors, QTc prolongation represents an independent risk factor for SCD and is a predictor of both all-cause and cardiovascular mortality in various clinical settings. Importantly, studies have shown that approximately 1% of all individuals undergoing inpatient or outpatient 12-lead ECG have a QTc ≥ 500 ms. It should be noted that when this threshold is reached or exceeded, the risk of death increases by approximately 2–4 times and, at least in the short term (e.g., 30 days), it serves as a powerful predictor of all-cause mortality, outperforming conventional comorbidity indicators. Although this article discusses the 500 ms threshold, any appropriate threshold (e.g., 450 ms, 475 ms) can be used to indicate QTc prolongation representing a high risk of heart disease-related mortality. Furthermore, an increase in a specific amount of QTc over a specific time period can also indicate QTc prolongation representing a high risk of heart disease-related mortality. For example, if a user's QT is measured at 400 milliseconds and then at 450 milliseconds at a later time (within the threshold time period), this could indicate that QTc has increased, even though 450 may be below the QTc threshold for a single measurement (e.g., 500 ms).

[0065] Furthermore, in many cases, the development of QTc ≥ 500 ms is at least partly driven by the presence of a potentially lethal but highly treatable genetic disorder (i.e., congenital LQTS) and / or modifiable risk factors (e.g., electrolyte abnormalities, use of drugs that prolong ≥ 1 QTc, or underlying QT irritability disorder). Therefore, in many cases, identifying significant QTc prolongation provides important opportunities for: i) identifying host susceptibility at risk, and ii) implementing potentially life-saving modifications required to reduce the risk of TdP and SCD (i.e., initiating beta-blocker use, discontinuing QTc prolonging drugs, or correcting hypokalemia and hypomagnesemia).

[0066] However, due to the reliance on large 12-lead ECG systems and trained ECG technicians, the clinical settings for monitoring QTc are largely limited to “snapshot” assessments in hospitals and outpatient clinics.

[0067] Therefore, in some embodiments, the ECG sensing device 500 can be used to predict a user's QT interval. As discussed herein, machine learning (ML) is well-suited for continuously monitoring one or more criteria to identify large and small anomalies or trends in input data compared to training examples used to train the model. ML models that can be used with the embodiments described herein, by way of example and not limitation, include: Bayes, Markov, Gaussian processing, clustering algorithms, generative models, kernels, and neural network algorithms. Some embodiments utilize machine learning models based on trained neural networks (e.g., trained recurrent neural networks (RNNs) or trained convolutional neural networks (CNNs)).

[0068] ECG sensing device 500 can execute module 510B to predict QT intervals as discussed herein. As described herein, module 510B may include a deep neural network (DNN) for predicting QT intervals; however, any suitable ML model may be used. Users can position ECG sensing device 500 in any suitable manner described herein (e.g., in response to a received instruction, as discussed further in detail herein) and perform ECG using ECG sensing device 500. Processing device 505 can process the recorded ECG signal using an averaging slap algorithm, and the input to the DNN may be the averaging slap in the form of a 2×450 length signal in millivolts. Before passing through the initial convolutional layers (e.g., 2D convolution, leaky rectified linear units, max-merge), and before being processed by three separate pass stages (two separate 2D convolutions, leaky rectified linear units, max-merge, etc.), the averaging slap signal may first be clamped via a hyperbolic tangent function, where the pass input (via trainable parameters) is added before downsampling, pooling, and convolution, and random dropout is applied. After final passing, signals are sent through two separate feedforward network layers, outputting a probability vector of the output class for a given QT value. The expected value of this probability vector is considered the predicted QT interval. The loss function used for training is the sum of the cross-entropy term of the probability vector with the target QT interval and the root mean square error of the QT interval mean. It should be noted that the ML model can utilize any appropriate number and combination of leads (and be trained on any appropriate number and combination of leads). In some embodiments, the ML model can be trained to predict the QT interval based on a combination of lead I, II, or III with any appropriate v lead. For example, the ML model can be trained based on lead I and v5, lead II and v5, or lead III and v5 (or any other appropriate v lead).

[0069] A DNN can be trained using annotated standard 12-lead ECG data from various patients acquired at the same location where the user was during ECG acquisition. In some embodiments, the ECG signals can be preprocessed prior to training. For example, a bandpass filter can be applied to simulate the dynamic bandwidth (0.1-40 Hz) of the ECG sensing device 500, and any suitable filtering algorithm (e.g., Alivecor) can be applied before processing the resulting 12-lead ECG using the average slap algorithm for leads I and II. TM (FDA-approved enhancement filtering algorithm). Source ECGs, including training data, can be classified based on QT intervals, and ECGs can be randomly selected from randomly selected classifications. Since the ECG signals received from the user may include leads generated using, for example, a separate lead conversion ML algorithm (e.g., via module 510A), in some embodiments, additional training data including synthetic leads can be used to further adapt the DNN.

[0070] All the devices described herein are suitable for use in a variety of systems, which may include one or more servers, one or more sensors, electronic data communication networks, and other ECG sensing devices. In some embodiments, multiple ECG sensing devices as described herein transmit ECG data to one or more remote servers via an electronic data communication network. In some embodiments, one or more remote servers are used to analyze the ECG data. In some embodiments, a remote server that analyzes the received ECG data is used to perform arrhythmia detection.

[0071] All devices and systems described herein may also include one or more software modules. In some embodiments, the software includes an application configured to run on a mobile computing device (e.g., a smartphone, smartwatch, or tablet). The software receives and processes ECG data received from an ECG sensing device. The software identifies individual leads within the transmitted data based on, for example, the electrodes from which the ECG data originates. For example, the software may be able to identify lead I based on signals from two electrodes measuring the potential difference between the right and left upper limbs. Once the ECG is identified, the software may also be configured to display a single-lead or multi-lead ECG on a display screen of the mobile computing device. The software may be configured to simultaneously display six leads I, II, III, aVR, aVL, and aVF on the display screen. The software may be configured to display one or more of the six leads I, II, III, aVR, aVL, and aVF on the display screen at a time, wherein the user can manually switch screens to see different one or more leads on different switching screens.

[0072] The software modules described herein include computer-readable and executable code. In various embodiments, a software module includes files, code segments, programming objects, programming structures, or combinations thereof. In other embodiments, a software module includes multiple files, multiple code segments, multiple programming objects, multiple programming structures, or combinations thereof. In various embodiments, as non-limiting examples, the one or more software modules include web applications, mobile applications, and standalone applications. In some embodiments, the software module is within a computer program or application. In other embodiments, the software module is within more than one computer program or application. In some embodiments, the software module is hosted on a single machine. In other embodiments, the software module is hosted on more than one machine. In other embodiments, the software module is hosted on a cloud computing platform. In some embodiments, the software module is hosted on one or more machines at one location. In other embodiments, the software module is hosted on one or more machines at more than one location.

[0073] In some embodiments, two or more sensed leads that are not sensed simultaneously are time-aligned to generate a time-aligned ECG trace that displays the two or more leads in a time-aligned format, such as in a conventional standard 12-lead ECG trace. In some embodiments of the ECG sensing device described herein, one or more ECG sensing electrodes are not simultaneously positioned on the skin of the individual sensing the ECG (i.e., some leads may be sensed sequentially). For example, limb leads (I, II, III, aVR, aVL, and aVF) may be sensed simultaneously, while one or more precordial leads are sensed separately from the limb leads. Thus, in these embodiments, the six limb leads are not automatically time-aligned with the individually and separately sensed precordial leads, but rather further processing is performed by a software application to time-align one or more limb leads with one or more precordial leads. In some embodiments, one or more of the six precordial leads are sensed individually, such that the individually sensed precordial leads are time-aligned by a software application with the six limb leads and with the other precordial leads. In some embodiments, the software application described herein aligns two or more sensed precordial leads to each other and individually time-aligns six sensed limb leads, such that the two sets of six leads are time-aligned separately (i.e., six time-aligned precordial leads and six individually time-aligned limb leads). In some embodiments, the software described herein aligns two or more sensed precordial leads to each other and to the sensed limb leads, such that all twelve sensed leads are time-aligned.

[0074] In some embodiments, one or more average or median waveforms are generated for the first and second leads, such that the waveforms corresponding to different heartbeats are time-aligned. That is, in some embodiments where one or more leads are sensed at different times, an average or median waveform is generated for one or more of these leads, and the average or median waveform is time-aligned such that the P, QRS, and T waveforms / groups are vertically aligned along the X-axis.

[0075] The memory 510 may include a timing alignment software module (not shown) capable of timing alignment of the P, QRS, and T waveforms / groups of the individual leads sensed by the ECG sensing device 500, such that the sensed ECG leads are aligned as waveforms in a conventional ECG trace when displayed. In some embodiments of the ECG sensing device, the ECG sensing device includes a software application configured to time-align two or more sensed ECG leads. In some embodiments of the ECG sensing device, the software application configured to time-align two or more sensed ECG leads is a component of a system that receives data from the ECG sensing device.

[0076] When the first and second electrodes of the ECG sensing device described herein are simultaneously contacted by the user's right and left upper limbs at any of the six precordial leads, Lead I is sensed concurrently with the sensed precordial leads. That is, Lead I equals the voltage sensed in the left upper limb minus the voltage sensed in the right upper limb. Therefore, when the left upper limb, right upper limb, and chest are all contacted by the electrodes of the ECG sensing device described herein, Lead I is sensed in addition to the precordial leads. Thus, when all six precordial leads are sensed sequentially, six corresponding "Precordial Lead I Records" are generated: V1-Lead I, V2-Lead I, V3-Lead I, V4-Lead I, V5-Lead I, and V6-Lead I. These six Precordial Lead I Records are each used to time-align each precordial lead to the limb leads, thereby achieving time alignment with the precordial leads.

[0077] In some embodiments, the timing alignment software module times-aligns precordial leads V1, V2, V3, V4, V5, and V6 by utilizing precordial lead I records simultaneously sensed with each of the waveforms V1, V2, V3, V4, V5, and V6. That is, precordial lead I records V1-lead I, V2-lead I, V3-lead I, V4-lead I, V5-lead I, and V6-lead I (senses of these leads simultaneously with the precordial lead records) are each time-aligned with a precordial lead record. For example, by moving the precordial lead I records a certain distance along the Y-axis, each precordial lead I record is time-aligned with the lead I records sensed together with the limb leads, and because each precordial lead I record is time-aligned with the precordial leads, when the same distance is moved along the Y-axis, each precordial lead V1, V2, V3, V4, V5, and V6 will also be aligned with the precordial lead I records sensed together with them. For example, “V1-Lead I” refers to a Lead I record that is time-aligned with V1. “V1-Lead I” is different from “Lead I,” which is a Lead I record simultaneously sensed with five other limb leads using the ECG sensing device described herein. “V1-Lead I” is also not necessarily time-aligned with “Lead I,” as these two different Lead I records are typically not sensed simultaneously using the ECG sensing device described herein. However, since both “V1-Lead I” and “Lead I” are Lead I records, they can be time-aligned in a fairly intuitive way because, on average, they are expected to have very similar (if not identical) morphology and timing between the waveforms. For example, if the peak of the R wave in the averaged "Lead I" occurs at 1 second, and the peak of the R wave in the averaged "V1-Lead I" occurs at 1.5 seconds, then the averaged "V1-Lead I" will be repositioned or shifted 0.5 seconds along the Y-axis so that its R wave peak occurs at 1 second, just as in the averaged "Lead I". Since V1 is time-aligned with "V1-Lead I", it must also be shifted 0.5 seconds along the Y-axis to align with the averaged "Lead I". When V1 is time-aligned with "Lead I", it will also be time-aligned with the other five limb leads that are already time-aligned with "Lead I". A similar alignment will occur for V2, V3, V4, V5, and V6 by aligning V2, V3, V4, V5, and V6 with "Lead I" respectively.

[0078] Figure 6This is a flowchart of a method 600 for predicting QT intervals according to some embodiments of the present invention. Method 600 can be performed by processing logic that may include hardware (e.g., circuitry, dedicated logic, programmable logic, processor, processing device, central processing unit (CPU), system-on-a-chip (SoC), etc.), software (e.g., instructions that run / execute on the processing device), firmware (e.g., microcode), or a combination thereof. In some embodiments, method 600 can be performed by, for example... Figure 5A The ECG sensing device 500 shown is used (via processing device 505).

[0079] In block 605, the ECG sensing device 500 may provide the user with instructions regarding the location where their body should be positioned (in some embodiments, the computing device 550 may provide these instructions), and instructions to position the ECG sensing device 500 on their body to record ECG. In block 610, the processing device 505 may perform ECG and acquire the user's ECG signal. In block 615, the processing device 505 may process the recorded ECG signal using an averaging spurs algorithm, and in block 620, the processing device 505 may execute module 510B to predict the user's QT interval. As described herein, module 510B may include a deep neural network (DNN) for predicting the QT interval; however, any suitable ML model may be used. The input to the DNN may be, for example, an averaging spurs in the form of a 2×450 length signal in millivolts. Before passing through the initial convolutional layers (e.g., 2D convolution, leaky rectified linear units, max pooling), and before being processed by three separate pass stages (two separate 2D convolutions, leaky rectified linear units, max pooling, etc.), the average spurious signal can first be clamped via a hyperbolic tangent function, where the pass input (via trainable parameters) is added before downsampling, pooling, and convolution, and random deactivation is applied. After the final pass, the signal is sent through two separate feedforward network layers, outputting a probability vector of the output class for a given QT value, the expectation of which is considered the predicted QT interval. The loss function used for training is the sum of the cross-entropy term of the probability vector with the target QT interval and the root mean square error of the average QT interval.

[0080] In some embodiments, one or more electrodes may be external to the mobile computing device. In such embodiments, one or more external electrodes are wirelessly or hardwired to the mobile computing device. Non-limiting examples of wireless connectivity may include, for example, a WiFi connection between one or more external electrodes and the device, or a Bluetooth connection between one or more external electrodes and the device. ®The connection may include a low-power Bluetooth connection between one or more external electrodes and the device, an NFC (Near Field Communication) connection between one or more external electrodes and the device, or a near-field ultrasonic communication connection between one or more external electrodes and the device. Those skilled in the art will understand that other components that wirelessly communicate with the device are suitable for use with the systems, devices, and methods described herein.

[0081] Additionally, software incorporated with any system, apparatus, or method described herein can be configured to analyze ECG data received from an ECG sensing device. Analysis may include generating QRS and T-axis values ​​using the six leads I, II, III, aVR, aVL, and aVF described herein. Furthermore, software incorporated with any system, apparatus, or method described herein can determine the QRST angle by calculating the difference between the QRS and T-axis as described herein. Analysis may also include rhythm analysis, which may include determining heart rate variability, QT interval, or corrected QT interval.

[0082] Furthermore, software used in conjunction with any of the systems, devices, or methods described herein can be used to determine diagnoses or abnormalities associated with ECG. For example, as described, axis deviation may be associated with abnormalities such as right or left ventricular hypertrophy. For example, heart rate variability may be associated with a diagnosis of atrial fibrillation. For example, QT interval changes may indicate certain arrhythmias.

[0083] Any system, device, or method described herein can also be combined with sensors that measure physiological parameters. For example, any system, device, or method described herein can be combined with a blood pressure sensor. For example, any system, device, or method described herein can be combined with a photoplethysmography (PPG) sensor. For example, any system, device, or method described herein can be combined with a temperature sensor. For example, any system, device, or method described herein can be combined with a pulse oximeter sensor. For example, any system, device, or method described herein can be combined with an accelerometer. Those skilled in the art will understand that other sensors for monitoring or detecting physiological parameters are suitable for use with the systems, devices, and methods described herein.

[0084] In some embodiments, in any system, apparatus, and method described herein, sensed physiological data is sent to a processor. Software combined with any system, apparatus, and method described herein can use the sensed physiological data in combination with a sensed ECG for analysis. For example, blood pressure data can be combined with ECG data by the software to provide analysis identifying the presence of ventricular tachycardia (an immediately life-threatening condition).

[0085] The systems, apparatus, and methods described herein may include either or both of a transmitter and a receiver for transmitting and receiving wireless signals.

[0086] In some embodiments, when an anomaly analysis result is determined, the software described herein also causes a signal to be sent to a server. For example, the anomaly analysis result includes an abnormal ECG. For example, the anomaly analysis result includes an abnormal QRS axis. For example, the anomaly analysis result includes an abnormal QRST angle. In some embodiments, the anomaly analysis result includes an abnormal ECG. For example, the anomaly analysis result includes an abnormal heart rate variability value. For example, the anomaly analysis result includes an abnormal physiological parameter value. The signal sent may include a signal to an emergency care provider. For example, if an immediately life-threatening condition, such as VT, is determined, the software described herein may send an emergency signal to an emergency operator, an emergency care provider (e.g., a caregiver), or another third-party detector.

[0087] A six-lead ECG can be displayed on the display screen, the six-lead ECG including lead I, lead II, lead III, lead aVR, lead aVL and lead aVF.

[0088] While preferred embodiments of the systems, apparatuses, and methods described herein have been shown and described, it will be apparent to those skilled in the art that these embodiments are provided by way of example only. Many variations, modifications, and substitutions will now be apparent to those skilled in the art without departing from the user's questions described herein. It should be understood that various alternatives may be employed in implementing the embodiments of the systems, apparatuses, and methods described herein. The scope of the invention is intended to be defined by the following claims, and the methods and structures within the scope of these claims and their equivalents are thereby covered.

[0089] In one embodiment, the methods and systems described herein can be combined with methods and systems for guiding and / or calibrating electrode placement on a patient. These methods can be specifically used to guide the placement of electrodes for ECG measurements. Generally, the systems, apparatus, and methods contemplated herein process images of a patient to output electrode positions on the patient. Typically, this may include presenting an image of the patient (e.g., a modified version of a patient image) showing the positions of the electrodes relative to the actual patient image.

[0090] For example, a system or apparatus for guiding electrode placement, as described herein, may include control logic for controlling a processor (e.g., a microprocessor of a computing device such as a handheld computing device) to receive images of the patient, analyze the patient to determine the correct placement of the electrodes, and output an image of the patient with the correctly marked predetermined electrode positions. Generally, the control logic may be configured as software, hardware, or firmware and may control a general-purpose computing device (e.g., a computer, tablet, etc.) or a mobile telecommunications device (a smartphone, such as an iPhone™, Android™, etc.) to receive or acquire images and output images of the patient. Processing steps may be performed remotely or locally. Generally, processing steps may include comparing the patient's images with a database of various body types (e.g., an electrode placement database) and the corresponding predetermined, conventional, or standard positions of the electrodes associated with each body type. Images may also be normalized before being compared with a patient database by resizing the images and / or, in some cases, adjusting the aspect ratio, brightness, contrast, or other image features to allow direct comparison with the database. Normalization may be performed using markers that are part of the image. For example, images of the patient may be acquired using markers of known or known size on the patient, and these markers may be used as normalization markers to normalize the images before comparison with the database. Standardization can also be performed to balance the brightness, contrast, sharpness, or other imaging qualities of the image. The markings can be placed or applied directly to the patient (e.g., the patient's torso), for example, using adhesives.

[0091] Methods performed by devices and systems for guiding electrode placement, such as methods for guiding a patient to place electrodes, are also described.

[0092] For example, this paper envisions a method for guiding the correct placement of electrodes on a patient, comprising: comparing images of the patient with an electrode placement database to determine the location of the electrodes on the patient, wherein the electrode placement database includes representations of multiple body types and predetermined electrode placement locations corresponding to each body type; and presenting an image of the patient showing the location of the electrodes on the image of the patient.

[0093] These methods are particularly well-suited for guiding the placement of ECG electrodes on patients using standard or routine configurations. Therefore, the database can be configured to include multiple body types, and corresponding routine / standard electrode placement locations for each body type in the database.

[0094] In some variations, the electrode locations of all electrodes can be identified and indicated (e.g., all 10 electrode locations for a standard 12-lead electrode system). However, some variations can be identified and / or displayed on a subset of electrode locations. For example, a method for guiding the placement of standard / conventional 12-lead electrodes may identify and display only six electrode locations on the patient's chest. In some variations where other electrode locations can be identified relative to one or more key electrode locations, only the locations of the key electrodes may be displayed.

[0095] Generally, any suitable image of the subject can be used. In some variations, the system, apparatus, or method may include acquiring or obtaining images. In some variations, images may be acquired by a system or apparatus performing the method (e.g., a smartphone or other handheld computer device). The systems, apparatus, and methods described herein may instruct the user how to acquire images of the patient, including positioning the patient (facing forward, standing, sitting, lying down, etc.), approximately how far away from the patient to acquire images, positioning standardized markers on or near the patient, etc. Images may be received as digital images. Images may include images of the patient, particularly areas of the patient's body to which electrodes will be applied. For example, when applying ECG electrodes, the image may include the patient's torso or chest. Other areas of the patient's body may be included, such as the patient's head, legs, etc. The patient may be standing, sitting, or lying down. The areas of the patient to which electrodes will be applied are typically exposed (e.g., may be bare-chested or at least partially bare, allowing the skin to be visible). As previously described, in some variations, standardized markers may be included as part of the image. For example, a reference marker may be placed on the patient; the reference / standardized marker typically has a known or standard size (such as a coin, e.g., a US cent, penny, etc.). In some variations, a reference marker is provided, and the reference marker may be of a different shape or color. In some variations, the marker is automatically recognized by the device. For example, the marker may include a readable code (e.g., a barcode, alphanumeric code, QR code, etc.); alternatively, the device may recognize the marker by color, shape, etc.

[0096] In variations of methods, systems, or devices that guide a user to collect or acquire images, the images may be identified by the system or device. Identifying images may include examining them to confirm that they are suitable and can be analyzed (e.g., compared) relative to a database.

[0097] As used herein, the phrases “user” and “patient” are intended to broadly encompass any subject to whom methods, devices, and systems can be used to aid in the positioning of electrodes. A patient may include animals (in systems and devices specifically configured for that type of animal) or humans, and may include healthy or unhealthy subjects. As used herein, “user” can be a person using the systems, methods, and devices described herein. In some variations, “user” is the same as “patient” because the systems, devices, and methods described herein can be used by a patient to guide the placement of electrodes thereon.

[0098] In some variations, comparing images with an electrode placement database may include determining the standard placement of electrodes for a patient's 12-lead ECG.

[0099] Generally, comparing a patient's image with an electrode placement database may involve determining a match (e.g., the closest match) between the image and one or more representative body types in the patient database. Once one or more closest matching representative body types have been identified, the electrode placement corresponding to the matching representative body type can be mapped to the patient's image. In cases where more than one match is identified, electrode placements can be determined from standard electrode placements corresponding to multiple representative body matches using weighted, averaged, or other appropriate statistical methods for finding consensus criteria in the closest matches, and these standard electrode placements can be mapped to the patient's image.

[0100] As described in more detail below, electrode placement databases typically include multiple (e.g., >10, >100, >1000, >10000, etc.) representations of standard / conventional electrode placements for different bodies. Body type representations may include images of the body (e.g., pictures, portions of pictures, etc.) or information extracted from images of that body including body-specific electrode placements, where the electrode placements have been identified or verified as being within standard / conventional limits. Various body types may include different shapes and sizes (height, weight, morphology), sex (male / female), age (infant, child, adult, elderly), body morphology (shoulder width, chest circumference, waist circumference, etc.). Individual body type representations may be unique, but may include similar body types, thus creating body type clusters around more common body types. All body types in the database may be pre-normalized to allow comparisons between different representations. Multiple different electrode placement databases may be used. For example, a separate database may be used for different patient positions (lying down, sitting, standing, etc.), or for different patient sexes, ages, etc. Furthermore, different electrode placement databases may be used for different standard / conventional electrode placements.

[0101] Therefore, in addition to standardizing the images before comparing them with an electrode placement database, the images can be processed to make them ready for comparison with the database. In variations where comparisons are made by extracting features from the images and comparing these extracted features with representations of body types in a database, feature extraction can be performed on the images before (or as part of) the comparison. For example, when comparing images of a patient with an electrode placement database involves identifying anatomical landmarks from the images and comparing these landmarks with the database, anatomical landmarks can be extracted from the images first. Images can also be processed to remove patient-identifying features that might be relevant to protecting patient privacy (e.g., all or part of the patient's face).

[0102] As mentioned above, the comparison between images and the database can include interpolation between the closest match to the image and two or more representative body types in the patient database.

[0103] In some variations, comparing patient images with a database involves using pattern recognition to determine the closest match between the image and a representative body type in the database. In other variations, comparing patient images with an electrode placement database involves comparing standardized images of the patient with the electrode placement database.

[0104] The methods, apparatus, and systems envisioned herein may also include presenting an image of the patient showing the location of the electrodes on the patient's image. Any suitable image of the patient can be presented, including a modified version of an image of the patient showing the location of the electrodes determined by comparison with a database. In some variations, the image of the patient is displayed digitally (e.g., on a handheld computing device) and can be enlarged (zoomed out) or manipulated so that the user can see the location of the electrodes. In some variations, the image may include additional guidance, including measurements for the patient (scales in inches, mm, distances, etc.), patient landmarks such as anatomical landmarks, and / or for other electrodes.

[0105] The presentation of patient images showing the conventional / standard positions of electrodes can show all electrodes or some electrodes. In some variations, the presentation of images may include a series of images showing the patient with the indicated different electrode positions, to better allow the user to step through the process of applying or repositioning the electrodes. Generally, the presentation of patient images can be visual (showing images) and may also include textual (written / oral) instructions for applying the electrodes. For example, in a variation of the system and method described herein intended for use with a handheld computer device (such as a smartphone), the device can be controlled to guide the user through the steps by both acquiring images of the patient and positioning (or repositioning) the electrodes by viewing the smartphone screen.

[0106] In some variations, the methods, apparatus, and systems described herein can be used to correct and / or verify the position of electrodes already present on a patient. For example, a user can acquire or receive images of a patient already having ECG electrodes on their chest. Comparing the patient's images to an electrode placement database can also compare the position of the electrodes already on the patient to a defined standard / conventional position. Therefore, comparing the patient's images to an electrode placement database can include comparing images of patients with one or more electrodes already placed on their chest to the database. The position of one or more electrodes already placed on the patient's chest can then be verified automatically (indicating when one or more electrodes have been placed) or passively by overlaying the correct position (in a specific way, e.g., by color) onto the patient's image to form the presented image. In some variations, the presented image includes an image of the patient showing the corrected positioning of the electrodes on the patient's image.

[0107] This document also envisions a method for guiding ECG electrode placement, the method comprising: receiving an image of a patient, including the patient's chest; comparing the patient's image with an electrode placement database to determine the positioning of the electrodes on the patient, wherein the electrode placement database includes representations of multiple body types and predetermined conventional ECG electrode placement locations corresponding to each body type; and presenting an image of the patient showing the locations of the conventional ECG electrode positions on the patient's image. According to the method of this application, comparing the patient's image with the electrode placement database includes determining the closest match between the image and representative body types in the electrode placement database.

[0108] As described above, comparing patient images with an electrode placement database involves identifying anatomical landmarks from the images and comparing these landmarks with the electrode placement database. In some variations, comparing patient images with a database involves using pattern recognition to determine the closest match between the images and representative body types in the electrode placement database.

[0109] In any of the variations described herein, the comparison of patient images with an electrode placement database can be performed remotely relative to other steps. For example, images of the patient can be acquired using a smartphone (e.g., using an app downloaded to the phone) and presented as images of the patient showing the conventional positions of the electrodes; the comparison of images with a database can be performed remotely using a remote server. Therefore, the database can be maintained separately from the application on the smartphone (or other device). This allows for changes, updates, or other modifications to the database and / or the mechanisms used to compare patient images with it. The generated image can then be presented on the handheld computer device after the handheld computer device receives information (or the generated image) from the remote database. Alternatively, in some variations, all steps are performed at a local level (e.g., using a handheld computing device, such as a smartphone or tablet).

[0110] As described above, the patient's image may include standardized markers. Therefore, the step of receiving the patient's image may include receiving the patient image including standardized markers. In some variations, the received patient image may include electrodes on the patient's chest; the method, apparatus, or system can verify the placement of electrodes on the chest relative to conventional ECG electrode placement positions.

[0111] This document also describes a method for determining the placement of ECG electrodes, comprising: receiving an image showing a patient and standardized markers; standardizing the image using the standardized markers; comparing the standardized image with an electrode placement database including representations of various body types and predetermined ECG electrode placement locations for each body type to determine the location of the electrodes on the patient; and presenting an image of the patient showing the location of the ECG electrodes on the patient's image.

[0112] The system or device may be configured to perform any or all of the steps described above for receiving an image of a patient, including the area of ​​the patient to which electrodes will be applied, analyzing the image, and providing an image of the patient (or any other patient-specific mapping) showing the location of one or more electrodes on the patient based on predetermined, conventional, and / or standard electrode positions.

[0113] While many of the examples described herein are specific to systems, devices, and methods for placing ECG electrodes according to standard or conventional 12-lead ECG electrode placement (e.g., device 500), these systems, devices, and methods can be used (or are suitable for use) with any predetermined, conventional, and / or standard electrode placement system, including electrodes for electroencephalography (EEG), electromyography (EMG), current skin reflex (GSR), electrooculography (EOG), bioimpedance (BI), etc. For example, an electrode placement database may include various body types and corresponding predetermined, conventional, and / or standard electrode locations for each body type for EEG, EMG, GSR, EOG, BI, etc. In some variations, different electrode placement areas (different conventional and / or standard electrode placements) may be linked to individual body types in the database, and the user may select which placement areas to display. In other variations, multiple placement schemes may be shown sequentially or simultaneously for the same patient. For example, for ECG electrode placement, electrode placement may correspond to 3-lead, 5-lead, and 12-lead ECGs.

[0114] Systems for guiding electrode placement typically include control logic, which can be executed as software, hardware, or firmware (or a combination thereof) that receives images of the patient, uses an electrode placement database to determine the patient's routine and / or standard electrode placement, and outputs a map or image of the patient showing where the electrodes should be positioned on the patient. The system can also be configured to guide or lead a user through the processes of acquiring images of the patient and / or positioning electrodes on the patient. In some variations, the system is configured to guide the user via audible, written, and / or visual instructions. The system can be configured to operate according to (e.g., control) a handheld computing device, including a smartphone (e.g., iPhone, Android, etc.), to receive (and in some cases acquire) images of the patient and output an image of the patient with the determined electrode locations marked. For example, the system can be configured as a smartphone-specific application downloadable to a smartphone.

[0115] Any system described herein may be a dedicated system pre-configured to receive patient images and output images of the patient showing electrode placement locations, without requiring an application (e.g., software) to be downloaded to a separate device. For example, the system may include a camera for acquiring patient images and control logic for receiving images, controlling the analysis of the images to determine electrode placement using an electrode placement database, and outputting a map or image of the patient showing the location of one or more electrodes according to conventional and / or standard electrode positioning schemes. The system may include all or part of the electrode placement database, or the system may communicate with a remote electrode placement database. Furthermore, the system may include a comparison unit that includes comparison logic for comparing the patient images with body types in the electrode placement database to find one or more close matches between the patient and the body types in the database, from which predetermined conventional and / or standard electrode positions can be extrapolated to the patient images.

[0116] The system can also be configured to use (and may be included as part of the system) standardized markers incorporated into patient images. Standardized markers are typically distinctive marks that the system / device described herein can distinguish in an image and can be used to provide scale and / or orientation for reference within the image. For example, a standardized marker could be a label that can be attached to the patient's skin; the label could be a bright color and could have a known size (e.g., an orange circle with a diameter of 1 inch). Thus, the system / device can distinguish a marker of that specified size in an image and can use the standardized marker to standardize the image. In some variations, the standardized marker can also provide a reference location that the system can use to provide instructions for electrode placement. In some variations, more than one marker of a specified size can be used. Standardized markers can be public objects of known size, such as coins. The user can indicate what a standardized marker is within the system / device (e.g., from a menu of possible standardized markers).

[0117] As described above, a patient image showing the electrode positioning can be displayed to the user on a handheld computer device. For example, the handheld computer device can be a mobile phone, smartphone, tablet computer, or camera with network connectivity.

[0118] Figure 7This is a block diagram of an example computing device 700 capable of performing one or more operations described herein, according to some embodiments. In some embodiments, computing device 700 may represent the internal hardware of ECG device 500. Computing device 700 may be connected to other computing devices in a LAN, intranet, extranet, and / or the Internet. The computing device may operate as a server machine in a client-server network environment or as a client in a peer-to-peer network environment. The computing device may be provided by a personal computer (PC), set-top box (STB), server, network router, switch, or bridge, or any machine capable of executing a set of instructions (sequential or otherwise) specifying actions to be taken by the machine. Furthermore, although only a single computing device is illustrated, the term "computing device" should also be considered as including any collection of computing devices that individually or collectively execute a set (or more) of instructions to perform the methods discussed herein.

[0119] Example computing device 700 may include processing device (e.g., general-purpose processor, PLD, etc.) 702, main memory 704 (e.g., synchronous dynamic random access memory (DRAM), read-only memory (ROM)), static memory 706 (e.g., flash memory and data storage device 718), which can communicate with each other via bus 730.

[0120] Processing device 702 may be provided by one or more general-purpose processing devices such as microprocessors, central processing units, etc. In illustrative examples, processing device 702 may include a Complex Instruction Set Computing (CISC) microprocessor, a Reduced Instruction Set Computing (RISC) microprocessor, a Very Long Instruction Word (VLIW) microprocessor, or a processor implementing other instruction sets or combinations of instruction sets. Processing device 702 may also include one or more special-purpose processing devices, such as application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), network processors, etc. Processing device 702 may be configured to perform the operations described herein according to one or more aspects of the present invention to perform the operations and steps discussed herein.

[0121] The computing device 700 may also include a network interface device 708 capable of communicating with the network 720. The computing device 700 may also include a video display unit 710 (e.g., a liquid crystal display (LCD) or a cathode ray tube (CRT)), an alphanumeric input device 712 (e.g., a keyboard), a cursor control device 714 (e.g., a mouse), and a sound signal generation device 716 (e.g., a speaker). In one embodiment, the video display unit 710, the alphanumeric input device 712, and the cursor control device 714 may be combined into a single component or device (e.g., an LCD touchscreen).

[0122] According to one or more aspects of the invention, the data storage device 718 may include a computer-readable storage medium 728 on which one or more sets of ECG instructions 725, such as instructions for performing the operations described herein, may be stored. During the execution of the ECG instructions 725 by the computing device 700, the main memory 704, and the processing device 702, which also constitutes a computer-readable medium, the ECG instructions 725 may also reside wholly or at least partially in the main memory 704 and / or the processing device 702. The instructions 725 may also be transmitted or received on a network 720 via a network interface device 708.

[0123] Although computer-readable storage medium 728 is shown as a single medium in the illustrative example, the term "computer-readable storage medium" should be considered to include a single medium or multiple media (e.g., a centralized or distributed database and / or associated caches and servers) that store one or more instruction sets. The term "computer-readable storage medium" should also be considered to include any medium capable of storing, encoding, or carrying instruction sets executable by a machine and enabling the machine to perform the methods described herein. Therefore, the term "computer-readable storage medium" should be considered to include, but is not limited to, solid-state memory, optical media, and magnetic media.

[0124] The methods and illustrative examples described herein are not inherently related to any particular computer or other device. Various general-purpose systems can be used in accordance with the teachings described herein, or it can be demonstrated that it is convenient to construct more specialized devices to perform the required method steps. The necessary structures for various such systems will be as set forth in the description above.

[0125] The foregoing description is intended to be illustrative and not restrictive. Although the invention has been described with reference to specific illustrative examples, it will be appreciated that the invention is not limited to the described examples. The scope of the invention should be determined by referring to the following claims and the full range of their equivalents.

[0126] As used herein, unless the context clearly indicates otherwise, the singular forms “a,” “an,” and “the” are intended to include the plural forms as well. It is further understood that the terms “comprises,” “comprising,” “includes,” and / or “including” as used herein specify the presence of the stated feature, integer, step, operation, element, and / or component, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. Therefore, the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting.

[0127] It should also be noted that in some alternative embodiments, the functions / actions mentioned may not occur in the order shown in the figures. For example, two figures shown successively may actually be performed substantially simultaneously, or sometimes in reverse order, depending on the functions / actions involved.

[0128] Although the method operations are described in a specific order, it should be understood that other operations may be performed between the described operations, the described operations may be adjusted so that they occur at slightly different times, or the described operations may be distributed across a system that allows processing operations to occur at various intervals associated with the processing.

[0129] Various units, circuits, or other components may be described or claimed to be "configured to" or "configurable to" perform one or more tasks. In such a context, the phrase "configured to" or "configurable to" is used to indicate a structure by indicating that the unit / circuit / component includes a structure (e.g., a circuit) performing one or more tasks during operation. Thus, even when a specified unit / circuit / component is currently inoperable (e.g., not switched on), it can be said that the unit / circuit / component is configured to perform a task or is configurable to perform a task. Units / circuit / components used with the language "configured to" or "configurable to" include hardware (e.g., circuits, memory storing program instructions executable to perform operations, etc.). Enumerating units / circuit / components that are "configured to" or "configurable to" perform one or more tasks is explicitly intended not to refer to the unit / circuit / component in connection with 35 USC 112, paragraph 6. Additionally, "configured to" or "configurable to" can include a general-purpose structure (e.g., a general-purpose circuit) manipulated by software and / or firmware (e.g., an FPGA or general-purpose processor executing the software) to operate in a manner capable of performing the task of interest. "Configured to" can also include adjusting a manufacturing process (e.g., a semiconductor manufacturing facility) to produce means (e.g., an integrated circuit) suitable for implementing or performing one or more tasks. "Configurable to" is explicitly intended not to be applied to blank media, unprogrammed processors or unprogrammed general-purpose computers, or unprogrammed programmable logic devices, programmable gate arrays, or other unprogrammed devices, unless accompanied by programming media that endow the unprogrammed device with the ability to be configured to perform the disclosed functions.

[0130] For illustrative purposes, the above description has been given with reference to specific embodiments. However, the illustrative discussion above is not intended to be exhaustive or to limit the invention to the precise forms disclosed. Many modifications and variations are possible in light of the above teachings. These embodiments were chosen and described in order to best explain the principles of these embodiments and their practical application, thereby enabling others skilled in the art to best utilize these embodiments and various modifications that may be adapted to the particular uses contemplated. Therefore, these embodiments are to be considered illustrative rather than restrictive, and the invention is not limited to the details given herein but may be modified within the scope and equivalents of the appended claims.

Claims

1. A mobile electrocardiogram sensor, i.e., a mobile ECG sensor, comprising: An electrode assembly including electrodes, wherein the electrode assembly senses heart-related signals and generates electrical signals representing the sensed heart-related signals when in contact with a user's body. Processing device, operatively coupled to the electrode assembly, the processing device being used for: A training dataset is provided, which includes multiple electrocardiogram measurement results, i.e., ECG measurement results, and corresponding QT interval markers for each of the multiple ECG measurement results; The following methods can be used to train machine learning (ML) models: The ML model is used to analyze each ECG measurement result among the multiple ECG measurement results to generate a corresponding probability vector for each ECG measurement result among the multiple ECG measurement results; and For each probability vector among the multiple probability vectors: Compare the probability vector with the corresponding QT interval marker; and The ML model is updated based on a loss function, wherein the loss function is the sum of the cross-entropy of the probability vector and the corresponding QT interval marker, and the root mean square error of the average QT interval. The ML model is used to analyze the sensed cardiac-related signals to predict the 12-lead QT interval value, i.e., the 12-lead QTc value, based on the sensed cardiac-related signals, wherein the sensed cardiac-related signals include fewer than 12 leads; and The predicted QTc value is analyzed using the ML model to determine whether any health abnormalities exist; and A housing that houses the electrode assembly and the processing device.

2. The mobile ECG sensor according to claim 1, wherein, The training dataset includes multiple ECG measurements from one or more users, comprising multiple 12-lead ECG measurements, wherein the corresponding QT interval marker for each ECG measurement includes 12-lead QTc interval data from one or more users.

3. The mobile ECG sensor according to claim 1, wherein, The ML model is a deep neural network ML model.

4. The mobile ECG sensor according to claim 1, wherein, If the predicted QTc value indicates that QTc is prolonged, the processing device determines that the health abnormality exists.

5. The mobile ECG sensor according to claim 1, wherein, The processing device is also configured to send a notification to the client device in response to the presence of the health abnormality.

6. The mobile ECG sensor according to claim 4, wherein, QTc extension corresponds to an increase in QTc that is above the threshold amount and within the threshold time amount.

7. A mobile electrocardiogram system, i.e., a mobile ECG system, comprising: An electrode assembly including electrodes, wherein the electrode assembly senses heart-related signals and generates electrical signals representing the sensed heart-related signals when in contact with a user's body. Processing device, operatively coupled to the electrode assembly, the processing device being used for: A training dataset is provided, which includes multiple electrocardiogram measurement results, i.e., ECG measurement results, and corresponding QT interval markers for each of the multiple ECG measurement results; The following methods can be used to train machine learning (ML) models: The ML model is used to analyze each ECG measurement result among the multiple ECG measurement results to generate a corresponding probability vector for each ECG measurement result among the multiple ECG measurement results; and For each probability vector among the multiple probability vectors: Compare the probability vector with the corresponding QT interval marker; and The ML model is updated based on a loss function, wherein the loss function is the sum of the cross-entropy of the probability vector and the corresponding QT interval marker, and the root mean square error of the average QT interval. The ML model is used to analyze the sensed cardiac-related signals to predict the QT interval value (i.e., the 12-lead QTc value) from 12 leads based on the sensed cardiac-related signals, wherein the sensed cardiac-related signals include fewer than 12 leads; and The predicted QTc value is analyzed using the ML model to determine whether any health abnormalities exist. A display operatively connected to the processing device; and The memory includes instructions for the processing device to process the sensed heart-related signals and to display the heart-related signals on the display.

8. The mobile ECG system according to claim 7, wherein, The training dataset includes multiple ECG measurements from one or more users, comprising multiple 12-lead ECG measurements, wherein the corresponding QT interval marker for each ECG measurement includes 12-lead QTc interval data from one or more users.

9. The mobile ECG system according to claim 7, wherein, The ML model is a deep neural network ML model.

10. The mobile ECG system according to claim 7, wherein, If the predicted QTc value indicates that QTc is prolonged, the processing device determines that the health abnormality exists.

11. The mobile ECG system according to claim 7, wherein, The processing device is also configured to send a notification to the client device in response to the presence of the health abnormality.

12. The mobile ECG system according to claim 10, wherein, QTc extension corresponds to an increase in QTc that is above the threshold amount and within the threshold time amount.

13. A method performed by a processing device, comprising: A training dataset is provided, which includes multiple electrocardiogram measurement results, i.e., ECG measurement results, and corresponding QT interval markers for each of the multiple ECG measurement results; The following methods can be used to train machine learning (ML) models: The ML model is used to analyze each ECG measurement result among the multiple ECG measurement results to generate a corresponding probability vector for each ECG measurement result among the multiple ECG measurement results; as well as For each probability vector among the multiple probability vectors: Compare the probability vector with the corresponding QT interval marker; and The ML model is updated based on a loss function, wherein the loss function is the sum of the cross-entropy of the probability vector and the corresponding QT interval marker, and the root mean square error of the average QT interval. Process the acquired ECG signal; The ML model is used to analyze the processed ECG signal to predict the 12-lead QT interval, i.e., the 12-lead QTc value, based on the processed ECG signal, wherein the ECG signal includes fewer than 12 leads; and The predicted QTc values ​​are analyzed using the ML model to determine if any health abnormalities exist.

14. The method according to claim 13, wherein, The training dataset includes multiple ECG measurements from one or more users, comprising multiple 12-lead ECG measurements, wherein the corresponding QT interval marker for each ECG measurement includes 12-lead QTc interval data from one or more users.

15. The method of claim 13, further comprising determining the presence of the health abnormality if the predicted QTc value indicates a QTc extension.

16. The method of claim 13, further comprising sending a notification to a client device in response to the presence of the health abnormality.

17. The method according to claim 15, wherein, QTc extension corresponds to an increase in QTc that is above the threshold amount and within the threshold time amount.

Images