System and method for biomagnetic field imaging

Abstract

The apparatus for measuring magnetic fields from a subject's organs comprises multiple unshielded magnetometers in a three-dimensional arrangement. Each pair of magnetometers in the multiple magnetometers has a known separation distance. Each magnetometer in the multiple magnetometers is configured to simultaneously detect the biomagnetic field from at least a portion of the subject's organs and the background magnetic field, and to output a signal indicating the detected biomagnetic field and background magnetic field.

Description

Technical Field

[0001] (Related Applications) This application is a continuation application of U.S. Patent Application No. 18 / 607,317, filed on March 15, 2024, and U.S. Patent Application No. 18 / 607,352, filed on March 15, 2024, each of which claims priority to U.S. Provisional Patent Application No. 63 / 453,038, filed on March 17, 2023, U.S. Provisional Patent Application No. 63 / 453,041, filed on March 17, 2023, and U.S. Provisional Patent Application No. 63 / 502,388, filed on May 15, 2023. Each of the above applications is incorporated herein by reference in its entirety.

[0002] (Technical Field) The disclosed embodiments generally relate to systems and methods for detecting electrical activity in human anatomical structures via magnetic field sensing.

Background Art

[0003] Measurement of the biomagnetic field in human subjects can be used to understand potential disease states. For example, magnetoencephalography (MEG) is a non-invasive technique for examining human brain activity. Magnetocardiography (MCG) is a non-invasive technique for measuring the magnetic field generated by the electric current of the heart. Similar to electrocardiogram, magnetocardiogram has morphological features such as QRS complex, P wave, and T wave.

[0004] Research over the decades has shown that magnetocardiography (MCG) has the potential to improve decision-making in cardiac treatment. However, due to limitations in sensors and systems, widespread adoption in clinical practice has been hindered.

Summary of the Invention

[0005] Magnetocardiography (MCG) is a non-invasive imaging technique that has shown promise for improving clinical diagnosis. MCG records an extracorporeal magnetic field resulting from the electrical activity of myocardial fibers. The magnetic field map (MFM) obtained from MCG can provide detailed information about the propagation of electrical activity in the heart with minimal distortion because the magnetic field propagates through tissue without obstruction.

[0006] MCG is already used to diagnose coronary artery diseases such as ischemic heart disease, in which plaque buildup can narrow the blood vessels, resulting in insufficient and uneven distribution of blood and oxygen within the heart. MCG not only detects subtle changes in the magnetic field caused by ischemia but also localizes the source of the ischemic site, aiding in treatment planning.

[0007] Numerous studies have demonstrated that MCG provides information that complements electrocardiogram (ECG) testing, enabling more reliable identification of ischemic disease with parameterizable magnetic field morphology (MFM) based on various characteristics of the magnetic field pattern. In certain cases, MCG has shown high sensitivity in detecting coronary ischemia even when the corresponding ECG results are inconclusive, raising expectations for its clinical use.

[0008] MCG offers several advantages over electrocardiogram (ECG). Firstly, MCG is less affected by changes in chest conductivity distribution than ECG. Unlike ECG, where electrical signals can be distorted as they pass through body tissues, the magnetic field is not obstructed by the human body. By measuring electrical conduction activity earlier in every heartbeat, physicians can better detect abnormal patterns or find other indicators of potential disease conditions. Because MCG is less affected by bioconductivity, high-sensitivity MCG devices can reveal subtle functional changes or hypoperfusion that would not appear on an ECG. Secondly, MCG devices using sensor arrays can create magnetic maps of cardiac activity, which cannot be generated using a standard 12-lead ECG. Furthermore, MCG offers non-contact measurement, making it valuable in applications where the use of ECG electrodes is not possible, such as evaluating fetuses with life-threatening arrhythmias.

[0009] Current MCG devices have several drawbacks, including high operating costs and large size (for example, an MCG device can occupy the size of an entire room). To date, most MCG devices are based on superconducting quantum interference device (SQUID) technology. SQUID magnetometers are considered one of the most sensitive types of quantitative magnetometers, but they must operate at extremely low temperatures. Furthermore, MEG or MCG measurements (using most modalities) are typically performed in a magnetically shielded room to improve the signal-to-noise ratio of small biomagnetic signals.

[0010] One of the main challenges in detecting biomagnetic signals, such as cardiac signals, using an unshielded magnetometer is that biomagnetic signals tend to have small / weak amplitudes that are masked by magnetic interference from sources outside the target organ. Using the human heart as an example, the strength of the Earth's magnetic field is approximately 50 microtesla, which is about 1 million times the amplitude of the cardiac magnetic field signal expected when the MCG device is placed just outside the chest of a human subject (e.g., a patient). Furthermore, if the MCG device is deployed in a hospital or clinical setting, other medical devices, computers, mobile phones, televisions, or elevators within the building where the MCG device is deployed, as well as other objects near the MCG device, also generate spurious magnetic noise and interference, making it even more difficult to isolate the magnetic signal actually coming from the heart. A typical solution to this is to acquire MCG data in a magnetically shielded environment so that the signal measured by the magnetic sensor actually originates from the heart. However, this solution is costly and impractical, which is a major reason why MCG is not widely adopted in clinical settings.

[0011] Therefore, there is a need for improved systems, methods, and devices that enable the measurement of magnetic fields generated by a target organ (e.g., the heart or brain) while rejecting non-target-related interference. This disclosure describes improved systems, devices, and methods for magnetometric imaging by combining hardware components and signal processing techniques to decompose magnetic fields from a target organ (e.g., the heart or brain) of a human subject.

[0012] As disclosed herein, in some embodiments, the system and / or device comprises a fixed (e.g., rigid) three-dimensional arrangement of magnetometers configured to operate in a large, uniform magnetic field such as that provided by the Earth (i.e., in an environment that is not magnetically shielded). In some embodiments, the magnetometers are mounted on (or within) a panel. During operation of the device, multiple magnetometers detect biomagnetic fields from the subject's organs in the presence of a background magnetic field from the Earth.

[0013] As disclosed herein, in some embodiments, the device is a magnetocardiogram (MCG) device for measuring magnetic signals from a subject's heart. In some embodiments, the device is a magnetoencephalogram (MEG) device for measuring magnetic signals from a subject's brain.

[0014] As disclosed herein, in some embodiments, the device includes a plurality of magnetometers having an average spacing that satisfies constraints in Fourier space. In some embodiments, the magnetometers are spatially distributed in Fourier space to have wave vector coverage that allows for the collection of information from both biomagnetic fields from the subject's organs and background magnetic fields, enabling discrimination between these magnetic fields and inferences about the health characteristics of the human subject based on the organ magnetic fields.

[0015] As disclosed herein, in some embodiments, a plurality of magnetometers comprises an array of magnetic field sensors (e.g., magnetometers), the magnetic field sensors arranged in multiple layers. The combination of array dimensions, sensor arrangement, sensor spacing, and sensor alignment between layers enhances the detection of magnetic field signals from target organs and allows for the application of signal processing techniques to reduce / remove signals from interference sources. Thus, the combination of features enables the disclosed devices and systems to operate without magnetic shielding. Furthermore, the disclosed devices and systems can operate without cryogenic cooling (e.g., at room temperature or ambient operating temperature). This combination of features differentiates the disclosed devices and systems from and improves upon existing MCG systems.

[0016] The disclosed device (and system) has several advantages compared to other commercially available devices. First, the disclosed device and system can be operated at room temperature (e.g., ambient temperature) in an environment that is not magnetically shielded, thus reducing operating costs and improving usability. Second, the disclosed device utilizes a magnetometer, which allows for further opportunities for miniaturization and / or portability of the device.

[0017] Each of the systems, methods, and devices disclosed herein has several embodiments, and not just one of them embodies the desirable attributes disclosed herein.

[0018] According to several embodiments, an apparatus for measuring magnetic fields from a subject's organs comprises a plurality of magnetometers in a three-dimensional arrangement. Each pair of magnetometers in the plurality of magnetometers has a known distance from each other. Each magnetometer in the plurality of magnetometers is configured to simultaneously detect a biomagnetic field from at least a portion of the subject's organs and a background magnetic field, and to output a signal indicating the detected biomagnetic and background magnetic fields. The apparatus is configured to operate without magnetic shielding. In some embodiments, the background magnetic field includes the Earth's magnetic field. In some embodiments, the background magnetic field includes a magnetic field generated by one or more devices separate from the apparatus and located in close proximity to the apparatus (e.g., within 0.5 m, 1 m, or 2 m of the apparatus).

[0019] In some embodiments, each magnetometer in a plurality of magnetometers responds to the total magnetic field adjacent to each magnetometer.

[0020] In some embodiments, each known separation distance has its own fixed length.

[0021] In some embodiments, the plurality of magnetometers includes a first magnetometer having a variable position. Each pair of magnetometers in the plurality of magnetometers has a known separation distance that changes during device operation and is determined by tracking the position of the first magnetometer.

[0022] In some embodiments, the background magnetic field includes a uniform magnetic field. In some cases, the uniform magnetic field originates from the Earth.

[0023] In some embodiments, the plurality of magnetometers have an average spacing that satisfies a constraint condition in Fourier space. In some embodiments, the average spacing provides coverage of wave vectors for recovering information from both the biomagnetic field and the background magnetic field from the subject's organs.

[0024] In some embodiments, the plurality of magnetometers are spatially distributed in Fourier space to have coverage of wave vectors for recovering information from both the biomagnetic field and the background magnetic field from the subject's organs.

[0025] In some embodiments, the plurality of magnetometers comprise an array of magnetometers arranged in a stack of planes. Adjacent magnetometers in each plane of the stack of planes are separated by a first predefined interval along the length of the array and by a second predefined interval along the width of the array. Also, magnetometers in adjacent planes of the stack of planes are separated by a third predefined interval along the thickness direction of the array.

[0026] In some embodiments, a first magnetometer in a first plane of the stack of planes is aligned with a second magnetometer in a second plane of the stack of planes along the length and width of the array.

[0027] In some embodiments, the array of magnetometers includes a first subset of magnetometers disposed in a first plane of the stack of planes and a second subset of magnetometers disposed in a second plane of the stack of planes. The first subset of magnetometers is aligned with the second subset of magnetometers along the length and width of the array.

[0028] In some embodiments, each of the magnetometers in the array comprises an optically pumped magnetometer (OPM).

[0029] In some embodiments, each of the magnetometers in the array comprises a diamond nitrogen vacancy (NV) center sensor.

[0030] In some embodiments, each magnetometer in the array is equipped with a fluxgate sensor.

[0031] In some embodiments, a first subset of magnetometers in the array comprises optical pumping magnetometers. A second subset of magnetometers in the array comprises diamond NV center sensors.

[0032] In some embodiments, a first subset of magnetometers in the array comprises optical pumping magnetometers. A second subset of magnetometers in the array comprises fluxgate sensors.

[0033] In some embodiments, a first subset of magnetometers in the array comprises diamond NV center sensors. A second subset of magnetometers in the array comprises fluxgate sensors.

[0034] In some embodiments, a first subset of magnetometers in the array comprises optical pumping magnetometers, a second subset of magnetometers in the array comprises diamond NV center sensors, and a third subset of magnetometers in the array comprises fluxgate sensors.

[0035] In some embodiments, the distance between the device and the subject is 0.1 to 5 cm during the operation of the device.

[0036] In some embodiments, the apparatus is operable at room temperature.

[0037] In some embodiments, the device is capable of operating without magnetic shielding.

[0038] In some embodiments, each magnetometer in the magnetometer array is

number

number

number

[0039] In some embodiments, each magnetometer in the magnetometer array has a dynamic range (e.g., 50 μT, 100 μT, 150 μT, or 200 μT) sufficient to capture both the MCG signal from the heart and the signal from the Earth's magnetic field.

[0040] In some embodiments, the first predefined interval has a range of 0.5 cm to 2 cm.

[0041] In some embodiments, the second predefined interval has a range of 0.5 cm to 2 cm.

[0042] In some embodiments, the third predefined interval has a range of 0.5 cm to 2 cm.

[0043] In some embodiments, adjacent magnetometers in each plane of a planar stack have a pitch of 1.5 cm to 3.5 cm along the length of the array.

[0044] In some embodiments, adjacent magnetometers in each plane of a planar stack have a pitch of 1.5 cm to 3.5 cm along the width of the array.

[0045] In some embodiments, each magnetometer in a plurality of unshielded magnetometers has a dynamic range of about 50 microtesla.

[0046] In some embodiments, each magnetometer in a plurality of unshielded magnetometers has a dynamic range of about 100 microtesla.

[0047] According to several embodiments, an unshielded magnetometer system for measuring magnetic fields from a subject's organs includes a plurality of magnetometers in a three-dimensional arrangement. Each pair of magnetometers in the plurality of magnetometers has a known separation distance from each other. Each magnetometer in the plurality of magnetometers is configured to simultaneously detect a biomagnetic field from at least a portion of the subject's organs and a background magnetic field, and to output signals indicating the detected biomagnetic field and background magnetic field. The unshielded magnetometer system includes one or more processors and memory. The memory stores instructions for the one or more processors to execute. The stored instructions include instructions for each magnetometer in the plurality of magnetometers to (i) simultaneously measure a biomagnetic field from at least a portion of the subject's organs and a background magnetic field, and (ii) output signals indicating the detected biomagnetic field and background magnetic field signals. The magnetometer system is configured to operate without magnetic shielding.

[0048] In some embodiments, the stored instructions include signal processing instructions for separating the background magnetic field from the biomagnetic signal.

[0049] In some embodiments, multiple magnetometers are arranged within a housing.

[0050] In some embodiments, the magnetometer system includes a positioning arm for supporting multiple unshielded magnetometers. In some cases, the positioning arm is mounted on a base including one or more wheels. In some cases, the positioning arm is mounted on a patient support platform.

[0051] In some embodiments, the magnetometer system includes a panel for mounting multiple unshielded magnetometers. The panel is supported by positioning arms.

[0052] In some cases, the positioning arm is configured to maintain a gap of 0.5 cm to 5 cm between the panel and the subject's chest throughout the entire patient scan.

[0053] In some cases, the positioning arm includes one or more lockable degrees of freedom of movement.

[0054] In some cases, the positioning arm includes three coupled degrees of freedom (DOF).

[0055] According to several embodiments, a method for determining magnetic fields from the organs of a human subject is performed on a computer system. The computer system includes one or more processors and memory. The method includes receiving a plurality of signals corresponding to first time-series magnetic data generated from a plurality of unshielded magnetometers in close proximity to the human subject. The first time-series magnetic data corresponds to magnetic fields generated from the human subject. The first time-series magnetic data further includes environmental magnetic interference. The plurality of signals include contributions from biomagnetic fields from at least a portion of the subject's organs and background magnetic fields. The method includes synchronizing the first time-series magnetic data to a common clock to generate synchronous time-series magnetic data. The method includes applying one or more filters to the synchronous time-series magnetic data to obtain filtered data. The method includes applying one or more noise reduction techniques to the filtered data to generate updated time-series magnetic data.

[0056] In some embodiments, applying one or more filters to synchronous time-series magnetic data includes applying notch filters having a frequency of power line noise (e.g., power supply frequency) (e.g., 60 Hz in the United States, 50 Hz in Europe or Asia).

[0057] In some embodiments, applying one or more filters to synchronous time-series magnetic data includes applying a bandpass filter.

[0058] In some embodiments, the bandpass filter includes a frequency range from 0.5 Hz to 40 Hz.

[0059] In some embodiments, the multiple magnetometers are m × n arrays of magnetometers arranged in stacks of p planes, where m is the number of magnetometers along the length of the array, n is the number of magnetometers along the width of the array, and p is the number of planes in the stack of planes arranged along the height of the array.

[0060] In some embodiments, a stack of p planes includes a first plane and a second plane adjacent to the first plane. Applying one or more noise reduction techniques to the filtered data includes obtaining positional information corresponding to each magnetometer in the array of magnetometers. The method includes calculating the difference between a first filtered signal corresponding to each first magnetometer and a second filtered signal corresponding to each second magnetometer for each pair of magnetometers in the array of magnetometers, each comprising a first magnetometer located in the first plane and a second magnetometer located in the second plane. Each first and second magnetometer is aligned with respect to each other in the length and width directions.

[0061] In some embodiments, calculating the difference between a first filtered signal corresponding to each of the first magnetometers and a second filtered signal corresponding to each of the second magnetometers further includes determining the correlation between the first filtered signal and the second filtered signal using linear regression, and calculating the difference according to the determined correlation.

[0062] In some embodiments, applying one or more noise reduction techniques to filtered data includes applying principal component analysis (PCA) to extract a subset of variables from all variables in the filtered data.

[0063] In some embodiments, applying PCA involves determining a number of principal components (PCs) corresponding to the filtered data, and assigning zero values ​​to a subset of the PCs, thereby removing one or more noise contributions from the filtered data.

[0064] In some embodiments, applying one or more noise reduction techniques to filtered data includes applying signal source separation (SSS) to the filtered data.

[0065] In some embodiments, the synchronous time-series magnetic data includes magnetic data recorded over multiple events. The method further includes, for each magnetometer in an array of magnetometers, aligning each subset of the synchronous time-series magnetic data corresponding to each magnetometer over multiple events based on a trigger signal to generate each subset of the aligned signals, and combining each aligned subset of the synchronous time-series magnetic data over multiple events.

[0066] In some embodiments, the method further includes acquiring positional information corresponding to each magnetometer in a plurality of magnetometers, and correlating each of the signals of a plurality of signals with the acquired positional information.

[0067] In some embodiments, the updated time-series magnetic data includes multiple updated signals corresponding to each magnetometer in an array of magnetometers. The method includes acquiring positional information corresponding to each magnetometer in the array of magnetometers, correlating each updated signal of the multiple updated signals to its respective magnetometer based on the acquired positional information, generating a magnetic field map that spatially correlates the magnetic field distribution of a human subject's organs with the positional information of each magnetometer in the array of magnetometers, and displaying the magnetic field map on a display device.

[0068] In some embodiments, the magnetometer comprises multiple vector magnetometers. The method further includes receiving multiple signals corresponding to vector magnetic field signals from the multiple vector magnetometers, calibrating and processing the multiple signals according to the physical sensor orientation to obtain first time-series magnetic data.

[0069] According to some embodiments, the computer system includes one or more processors and memory. The computer memory stores one or more programs configured for execution by the computer system. The one or more programs include instructions for executing any of the methods described herein.

[0070] According to some embodiments, a non-temporary computer-readable storage medium stores one or more programs configured to be executed by a computer system having one or more processors and memory. The one or more programs include instructions for executing any of the methods described herein.

[0071] It should be noted that the various embodiments described above can be combined with any other embodiments described herein. The features and advantages described herein are not exhaustive, and many additional features and advantages will be apparent to those skilled in the art in light of the drawings, specification and claims. Furthermore, it should be noted that the language used herein has been chosen primarily for readability and instruction, and may not have been chosen to define or encompass the inventive subject matter. [Brief explanation of the drawing]

[0072] [Figure 1A] This figure shows the operating environment according to several embodiments.

[0073] [Figure 1B] This figure shows a device for measuring magnetic fields from a subject's organs according to several embodiments.

[0074] [Figure 2] This is a block diagram of a device for measuring the magnetic field of a target organ in a human subject according to several embodiments.

[0075] [Figure 3] This is a block diagram showing computer systems in several embodiments.

[0076] [Figure 4A] This figure shows exemplary embodiments of a device for measuring the magnetic field of a target organ in a human subject, according to several embodiments. [Figure 4B] This figure shows exemplary embodiments of a device for measuring the magnetic field of a target organ in a human subject, according to several embodiments. [Figure 4C] This figure shows exemplary embodiments of a device for measuring the magnetic field of a target organ in a human subject, according to several embodiments. [Figure 4D] This figure shows exemplary embodiments of a device for measuring the magnetic field of a target organ in a human subject, according to several embodiments. [Figure 4E] This figure shows exemplary embodiments of a device for measuring the magnetic field of a target organ in a human subject, according to several embodiments. [Figure 4F] This figure shows exemplary embodiments of a device for measuring the magnetic field of a target organ in a human subject, according to several embodiments. [Figure 4G] This figure shows exemplary embodiments of a device for measuring the magnetic field of a target organ in a human subject, according to several embodiments.

[0077] [Figure 5A] This document describes workflows for measuring magnetic fields from target organs in human subjects, using several embodiments. [Figure 5B] This document describes workflows for measuring magnetic fields from target organs in human subjects, using several embodiments.

[0078] [Figure 6A]This figure shows time-series magnetic field data from a multi-channel magnetometer array. [Figure 6B] This figure shows time-series magnetic field data from a multi-channel magnetometer array. [Figure 6C] This figure shows time-series magnetic field data from a multi-channel magnetometer array. [Figure 6D] This figure shows time-series magnetic field data from a multi-channel magnetometer array. [Figure 6E] This figure shows time-series magnetic field data from a multi-channel magnetometer array. [Figure 6F] This figure shows time-series magnetic field data from a multi-channel magnetometer array. [Figure 6G] This figure shows time-series magnetic field data from a multi-channel magnetometer array.

[0079] [Figure 7] This graph shows the actual signals from two geometric sensors arranged in an axial gradiometer configuration relative to the heart, according to several embodiments.

[0080] [Figure 8] This figure shows a general linear model (GLM) for linear regression in several embodiments.

[0081] [Figure 9A] This figure shows magnetic field signals from an array having two layers of magnetometers, according to several embodiments. [Figure 9B] This figure shows magnetic field signals from an array having two layers of magnetometers, according to several embodiments.

[0082] [Figure 10] This figure shows stacked epoch plots illustrating a single trial MCG that is visible in unobstructed operation, according to several embodiments.

[0083] [Figure 11A]This figure shows spatial magnetic field maps of time slices according to several embodiments. [Figure 11B] This figure shows spatial magnetic field maps of time slices according to several embodiments.

[0084] [Figure 12] This figure shows exemplary data collected from participants in MCG studies under several embodiments.

[0085] [Figure 13] This provides a summary of the maximum signal-to-noise ratio separated by experimental conditions for participants in MCG studies, based on several embodiments.

[0086] [Figure 14A] A flowchart of a method for determining magnetic fields from the organs of a human subject, according to several embodiments, is provided. [Figure 14B] A flowchart of a method for determining magnetic fields from the organs of a human subject, according to several embodiments, is provided. [Figure 14C] A flowchart of a method for determining magnetic fields from the organs of a human subject, according to several embodiments, is provided. [Figure 14D] A flowchart of a method for determining magnetic fields from the organs of a human subject, according to several embodiments, is provided.

[0087] Next, an embodiment is shown in the accompanying drawings with reference to the embodiment. In the following description, a number of specific details are given in order to fully understand the present invention. However, it will be apparent to those skilled in the art that the present invention can be carried out without requiring these specific details. [Modes for carrying out the invention]

[0088] Several methods, devices, and systems disclosed herein improve existing devices for bioimaging by advantageously providing devices and / or systems that combine hardware components and signal processing techniques for decomposing magnetic fields from target organs of human subjects. In this disclosure, the terms “human subject” and “patient” are used interchangeably.

[0089] According to several embodiments disclosed herein, the MCG system of the present invention is constructed around an array of unshielded scalar optical pumping magnetometers (OPMs) that effectively reject ambient magnetic interference without environmental magnetic shielding. Figure 1A shows an exemplary operating environment 100 according to several embodiments. In some embodiments, the operating environment 100 includes one or more devices 200. In some embodiments, device 200 is an MCG device for measuring magnetic signals from the heart of a human subject. In some embodiments, device 200 is a MEG device for measuring magnetic signals from the brain of a human subject. In the example of Figure 1A, each of the devices 200 (e.g., device 200-1 and device 200-2) is located at each (e.g., separate) site (e.g., site A and site B) corresponding to each (e.g., separate) city, state, or country. In some embodiments, the sites correspond to a hospital, clinic, or medical facility.

[0090] In some embodiments, device 200 (e.g., a device) is communicably coupled to computer system 120 via a communication network 110. In some embodiments, computer system 120 and device 200 are jointly installed in the same location (e.g., the same room in which device 200 is employed). For example, Figure 1A shows device 200-1 and computer system 120-1 located at site A (e.g., different rooms in the same hospital, or within the same room in the hospital). In some embodiments, device 200 and computer system 120 are located in different locations. For example, Figure 1A shows device 200-2 located at site B, while computer device 120-2 is located at site N. In some embodiments, the computer system is a system located on the cloud.

[0091] As disclosed herein, the device 200 includes one or more magnetometers 201 configured to measure (e.g., sense, acquire, collect, retrieve, etc.) magnetic field signals from target organs of a human subject 122 while the device 200 is operating in an environment that is not magnetically shielded. In some embodiments, the magnetic field signal is magnetic flux density. In some embodiments, the measured signals are transmitted to a computer system 120 for post-processing. In some embodiments, the data (e.g., actual data or post-processed data) is made to be displayed on the computer system 120 or on a display device 140 that is communicably connected to the computer system 120.

[0092] In some embodiments, device 200 is part of a self-contained device / system that also includes sensors (e.g., a magnetometer 201, and optionally, sensors of different sensor types such as optical sensors and motion sensors), one or more positioning arms (e.g., mechanical positioning arms) (e.g., positioning arm 142-1 or positioning arm 142-2, Figure 1A), a computer (e.g., computer system 120, display device 140, etc.), and a power supply with backup. In some embodiments, the magnetometer 201 is housed (e.g., contained) within a housing 143. The housing 143 can be easily sterilized for repeated use in different environments. In some embodiments, device 200 is made of non-magnetic material and is robust against vibration.

[0093] In some embodiments, the self-contained device / system 200 is a portable mobile device / system configured to be movable within the operating environment 100. For example, Figure 1A shows that in some embodiments, the device / system 200 includes a standalone positioning arm 142-1 mounted on a base 144 including one or more wheels 146 (e.g., wheels 146-1 and 146-2) for moving the device / system 200 between different examination rooms within the surgical environment 100. The device / system 200 can be positioned next to a patient lying on a patient support platform 148-1 or next to a patient sitting in a chair. Figure 1A also shows that in some embodiments, the device / system 200 includes a positioning arm 142-2 that can be directly mounted on the patient support platform 148-2. The device / system 200 is designed to allow data (e.g., imaging data, scan data, etc.) to be acquired from the patient without requiring the patient to move from their current position, thereby minimizing disruption to the emergency medical workflow.

[0094] As disclosed, the device / system 200 is easier to maneuver compared to existing MCG or MEG systems that require shielding and / or operate under cryogenic conditions. The portability of the device / system 200, coupled with its ability to operate without environmental magnetic shielding, makes it a suitable candidate for use in external environments, including military, unconventional, non-medical / hospital, and poor areas with limited infrastructure.

[0095] In some embodiments, the device 200 includes a panel (e.g., a rigid panel) on which a magnetometer 201 is mounted (or inside). The magnetometer 201 is modular and can be replaced and / or swapped if one or more fail. During operation of the device, the panel maintains a fixed position relative to the patient. For example, if the device 200 is an MCG device, the device 200 (e.g., the panel including the magnetometer 201) can maintain a position with a parallel gap of 0.5 cm to 5 cm relative to the patient's chest throughout the patient scan, with minimal vibration (e.g., no vibration) and / or minimal slip (e.g., no slip, slip / movement not exceeding 1 mm, etc.). In some embodiments, the stability of the device is aided by locking out unnecessary mechanical degrees of freedom (DOF) (e.g., no rotation around the y and / or z axes).

[0096] To elaborate further on embodiments in which the device 200 is an MCG device, in some embodiments, with the base of the device 200 stabilized against the floor and parallel to the patient's sagittal plane, the operator can easily and conveniently (e.g., within a few seconds of manual or automatic operation) operate the positioning arm 142 (e.g., positioning arm 142-1) to bring the sensor panel substantially close to and parallel to the patient's chest. In some embodiments, the device 200 allows for easy and precise positioning of the panel so that it is substantially centered on the patient's heart. For example, in some embodiments, the device 200 allows for fiducial alignment to standard anatomical landmarks to enable data consistency between patients.

[0097] In some embodiments, the positioning arm 142 (e.g., positioning arm 142-1 or 142-2) includes lockable degrees of freedom of movement (DOF) (e.g., three coupled DOFs, i.e., a double-joint pivot that allows for sliding along the y-axis and fine rotation around the x-axis following a coarse rotation around the x-axis).

[0098] In some embodiments, the device / system 200 is configured to acquire a biomagnetic field scan of a patient and provide information about potential disease conditions in the patient (including adults, children, and fetuses), such as coronary artery disease, ischemia, and arrhythmias. In some embodiments, the patient scan and diagnosis are provided during the same patient examination. For example, in some embodiments, the device / system 200 is configured to acquire a biomagnetic field scan of a patient within 1-2 minutes of the start of the scan. In some embodiments, the device / system 200 is configured to process the scan data locally (e.g., in real time) or to send the scan data to the cloud for processing (e.g., in real time). In some embodiments, the care report results are then returned to the operator within 2-5 minutes (e.g., by the computer system 120).

[0099] In the example shown in Figure 1A, the computer system 120 includes a data receiving module 124 configured to receive data from one or more devices 200. In some embodiments, the data corresponds to magnetic field data generated (e.g., measured, collected, acquired) from each magnetometer 201 of the devices 200. The computer system 120 also includes a data processing module 126 configured to process the received data. Further details of the data processing process are described with reference to Figures 5 to 11.

[0100] In some embodiments, the computer system 120 includes a database 128. Details of the database 128 will be described with reference to Figure 3.

[0101] In some embodiments, the computer system 120 includes a machine learning database 130 for storing machine learning information. In some embodiments, the machine learning database 130 is a distributed database. In some embodiments, the machine learning database 130 includes a deep neural network database. In some embodiments, the machine learning database 130 includes a supervised training and / or reinforcement training database.

[0102] Figure 1B shows a device 200 for measuring magnetic fields from a subject's organs in several embodiments. In this example, the device 200 occupies a three-dimensional space defined by three orthogonal axes (x, y, and z axes). The device 200 includes a plurality of magnetometers 201 (e.g., magnetometer 201-A to magnetometer 201-N). Each magnetometer 201 occupies a position in the three-dimensional space of the device 200, with x, y, and z axis coordinates, respectively. Each pair of magnetometers 201-i and 201-j in the plurality of magnetometers 201 are spaced apart by a distance dij. For example, Figure 1B shows that the pair of magnetometers 201-E and 201-F are spaced apart by a distance dEF, and the pair of magnetometers 201-C and 201-N are spaced apart by a distance d(CN). Each of the multiple magnetometers 201 is configured to simultaneously detect the biomagnetic field from at least a portion of the subject's organs and the background magnetic field.

[0103] In some embodiments, each magnetometer has a fixed (e.g., rigid) position in the apparatus 200. Each pair of magnetometers 201-i and 201-j in the plurality of magnetometers 201 are spaced apart by their respective fixed distances dij.

[0104] In some embodiments, the device 200 may include at least one magnetometer 201 whose position is variable (e.g., continuously variable). The position of the magnetometer can be tracked during device operation.

[0105] In some embodiments, each of the magnetometers 201 within the device 200 responds to the total magnetic field adjacent to the magnetometer 201. During device operation, the magnetometer 201 detects not only the biomagnetic field from the subject's organs but also the background magnetic field (e.g., from the Earth and / or other interference sources). In some embodiments, the magnetometer 201 is a scalar magnetometer that measures the total strength of the magnetic field it is subjected to (e.g., including the Earth's magnetic field) but does not measure its direction.

[0106] In some embodiments, each magnetometer in the magnetometer array has a dynamic range (e.g., 50 μT, 100 μT, 150 μT, 200 μT, or greater than 200 μT) sufficient to capture both the MCG signal from the heart and the signal from the Earth's magnetic field.

[0107] In some embodiments, the magnetometer 201 is a vector magnetometer capable of measuring the magnetic field component in a specific direction with respect to the spatial direction of the magnetometer.

[0108] In some embodiments, the magnetometers 201 within the apparatus have an average spacing that satisfies constraints in Fourier space. For example, in some embodiments, the average magnetometer spacing is determined by the Nyquist sampling rate in Fourier space of the wave vector of the target organ's magnetic field. In some embodiments, the magnetometers 201 are spatially distributed within the apparatus 200 (e.g., within the apparatus) so as to have wave vector coverage in Fourier space for collecting information from both the biomagnetic field and the background magnetic field from the target organ.

[0109] Developing a non-magnetically shielded (i.e., not magnetically shielded), portable MCG device capable of eliminating ambient / environmental noise and providing clinically useful magnetic field imaging of disease state effects presents numerous technical challenges. Because transient human cardiac signals are 10⁵–10⁶ times smaller than background interference, to detect heart rate signals in an unshielded environment, the magnetometer needs both a large dynamic range and excellent sensitivity. To obtain a clear magnetic field image of the entire torso, the sensor needs to have a range of 1 × 10⁶. 7It is necessary to demonstrate a high level of accuracy exceeding [a certain threshold], and as a result, slight differences in the magnetic field measurements across the entire chest are attributed to cardiac activity rather than sensor inaccuracy.

[0110] Table 1 compares candidate sensor types for which MCG recording has been demonstrated. Point-of-care infrastructure cost is defined as the monetary and practical costs of installing and maintaining the infrastructure associated with sensing technology. For example, cryogenic devices or magnetic shielding enclosures represent high capital costs for hospitals or clinics, and if patients must be transferred to the emergency department for treatment, high space and logistical costs are incurred within the patient's care route. Practical demonstration of clinical utility means literature-supported evidence of disease classification that is equivalent to or better than the sensitivity and specificity of standard treatments. Table 1. Comparison of MCG sensor technologies and readiness for clinical adoption. JPEG2026511267000005.jpg60164

[0111] Table 1 shows that clinically usable magnetic field imaging of cardiac disease states has only been demonstrated through sensor technologies that have high financial and practical costs associated with cryogenic and / or magnetically shielded environments. These challenges limit the widespread adoption of MCG in clinical settings.

[0112] The majority of clinical MCG studies are conducted using SQUID-based MCG systems due to their extremely high magnetic sensitivity and small voxel size. SQUID arrays were used in the early demonstration of MCG imaging for detecting cardiac lesions. Unfortunately, these high-performance systems are expensive due to the need for cryogenic cooling and their large hospital footprint. Compact, spin-exchange relaxation-free (SERF) optical pumping magnetometers (OPMs) have clinical utility for MCG data acquisition and can operate without cryogenic equipment. However, SERF OPMs are also expensive because they must operate within large room-scale magnetic shielding to eliminate the Earth's magnetic field and ambient power line noise. Induction coil sensors sensitive to time derivatives of magnetic fields have been employed for MCG recording, but further research is needed to determine their clinical utility. Other types of commercially available magnetometers, including fluxgate sensors and tunnel magnetoresistance (TMR) sensors, are low-cost and can operate without shielding or cryogenic equipment. However, demonstrations of the sensitivity required for MCG have been limited to proof-of-concept experiments.

[0113] Figure 2 is a block diagram of a device 200 (e.g., apparatus) for measuring the magnetic field of a target organ in a human subject, according to several embodiments.

[0114] Device 200 includes one or more magnetometers 201 (e.g., magnetic field sensors).

[0115] In some embodiments, the magnetometer 201 comprises an electron spin defect-based magnetometer, such as a diamond nitrogen vacancy (NV) center magnetometer (e.g., a solid-state sensor). The diamond NV center magnetometer is a quantum sensor that utilizes the occurrence of electron spin defects in a solid lattice, where the spin can be initialized and read out optically or electronically. In some cases, the defects may occur as atomic-level vacancies in the lattice structure, such as vacancies occurring near nitrogen atoms substituted in place of carbon atoms in diamond.

[0116] In some embodiments, the magnetometer 201 includes an optical pumping magnetometer (OPM) (e.g., a vapor cell sensor). The OPM is a quantum sensor containing heated alkaline vapor (including, but not limited to, cesium vapor or potassium vapor) through which a laser beam passes. Due to the quantum properties of atoms, the amount of light passing through the atomic vapor is modulated at a frequency proportional to the ambient magnetic field.

[0117] In some embodiments, the magnetometer 201 includes a scalar OPM. Scalar OPMs are highly sensitive to weak magnetic fields, have a large dynamic range (e.g., 50 μT or more), and do not require magnetic shielding or cryogenic operation. Furthermore, commercially available scalar OPMs are small, lightweight, and low-power, making them ideal for integration into bedside medical devices. Given the vast difference in magnitude between the Earth's magnetic field (~10⁻⁶ T) and the cardiac magnetic field (up to ~100 pT), a scalar OPM responds only to the cardiac magnetic field component aligned with the total magnetic field vector, which is primarily composed of the Earth's magnetic field, in an unshielded environment.

[0118] According to some embodiments disclosed herein, the combination of excellent sensitivity, large dynamic range, high precision, and technological maturity makes scalar OPMs particularly promising for integration into multi-channel high-resolution MCG devices. In some embodiments, the scalar OPM comprises a full-field OPM capable of operating in the Earth's magnetic field.

[0119] In some embodiments, the magnetometer 201 includes a fluxgate sensor.

[0120] In some embodiments, the magnetometer 201 responds to the total magnetic field in the vicinity of the magnetometer 201. During device operation, the magnetometer detects the total magnetic field, including the biomagnetic field from the subject's organs as well as the background magnetic field (e.g., magnetic fields from the Earth, other equipment in the vicinity of the magnetometer, etc.). In some embodiments, the magnetometer 201 is a scalar magnetometer that measures the total intensity of the magnetic field to which it is exposed, but does not measure its direction.

[0121] In some embodiments, the magnetometer 201 is a vector magnetometer capable of measuring not only the magnitude of the magnetic field but also the direction of each magnetic field. In this case, the total strength of the magnetic field can be obtained by calibrating and processing multiple signals according to the direction of the physical sensor (e.g., by calculating the dot product of the vector components).

[0122] In some embodiments, the use of a vector magnetometer as the magnetometer 201 can enable new capabilities of device 200. Cardiac magnetic signals are, by their nature, directional. By using a vector magnetometer, it becomes possible to completely decompose the axial components (e.g., x, y, and z components), thereby resulting in a higher information density and enhancing the source reconstruction method. Denoising algorithms can also be improved by decomposing noise according to the vector sensor axis. In some embodiments, a sensor calibration method is provided with the vector magnetometer. For example, the vector magnetometer can be calibrated periodically or before use of device 200 to ensure that orthogonality between sensors is maintained, and high-performance gradient geometry noise subtraction can be achieved.

[0123] In some embodiments, one or more magnetometers 201 comprises at least two magnetometers arranged in an array. In some embodiments, the array of magnetometers is arranged in a planar stack. Further details of the array of magnetometers are shown in Figures 4A to 4D.

[0124] In some embodiments, the device 200 includes a positioning arm 142, as described with respect to Figure 1A. In some embodiments, the device 200 includes a housing 143 (e.g., housing 251, see Figure 4F). In some embodiments, the device 200 includes one or more wheels 146 for supporting the device 200 (and the positioning arm 142), as described with respect to Figure 1A.

[0125] In some embodiments, device 200 is coupled to a power supply 147. In some embodiments, power supply 147 is an external power supply. In some embodiments, power supply 147 is a battery pack that provides backup power to protect data and / or protect device 200 in the event of a power outage. In some embodiments, the battery pack facilitates the movement of device 200 between rooms without losing power to device 200.

[0126] In some embodiments, the device 200 includes an input interface 210 for facilitating user input, such as a display 212, buttons 214, a keyboard and / or a mouse 216.

[0127] In some embodiments, the input interface 210 includes a display device 212. In some embodiments, the device 200 includes input devices such as buttons 214 and / or a keyboard / mouse 216. Alternatively or in addition, in some embodiments, the display device 212 includes a touch-sensitive surface, in which case the display device 212 is a touch-sensitive display. In some embodiments, the touch-sensitive surface is configured to detect various swipe gestures (e.g., vertical and / or horizontal sequential gestures) and / or other gestures (e.g., single / double taps). In computer devices with a touch-sensitive display, the physical keyboard is used optionally (e.g., a soft keyboard may be displayed when keyboard input is required). The input interface 210 also includes an audio output device 218, such as a speaker or an audio output connection connected to a speaker, earphone, or headphones. In some embodiments, the device 200 includes an audio input device 220 (e.g., a microphone) for capturing voice (e.g., utterances from the user). In some embodiments, device 200 uses a microphone and voice recognition to complement or replace the keyboard.

[0128] Device 200 also includes one or more processors (e.g., CPUs) 202, one or more communication interfaces 204 (e.g., network interfaces), memory 206, and one or more communication buses 208 (sometimes called a chipset) for interconnecting these components.

[0129] In some embodiments, device 200 includes a radio 220. The radio 220 enables one or more communication networks, allowing device 200 to communicate with other devices such as a computer system (e.g., computer system 120 in Figures 1 and 3) or a server. In some embodiments, the radio band 220 includes any other suitable communication protocol, including a variety of custom or standard radio protocols (e.g., IEEE 802.15.4, Wi-Fi, ZigBee, 6LoWPAN, Thread, Z-Wave, Bluetooth Smart, ISA 100.5A, WirelessHART, MiWi, ultrawideband (UWB), and / or software-defined radio (SDR)), custom or standard wired protocols (e.g., Ethernet or HomePlug), and / or communication protocols not yet developed as of the filing date of this patent application.

[0130] Memory 206 includes high-speed random-access memory such as DRAM, SRAM, DDR RAM, or other random-access solid-state memory devices. In some embodiments, memory includes non-volatile memory such as one or more magnetic disk storage devices, one or more optical disk storage devices, one or more flash memory devices, or one or more other non-volatile solid-state storage devices. In some embodiments, memory 206 includes one or more storage devices located remotely from one or more processors 202. Memory 206, or alternatively, non-volatile memory within memory 206, includes a non-temporary computer-readable storage medium. In some embodiments, memory 206, or the non-temporary computer-readable storage medium of memory 206, stores the following programs, modules, and data structures, or subsets or supersets thereof: Operational logic 222 includes procedures for handling various basic system services and performing hardware-dependent tasks; A communication module 224 (e.g., a wireless communication module) connects to and communicates with other network devices (e.g., local networks such as routers providing internet connectivity, network storage devices, network routing devices, server systems, computer systems 120, and / or other connection devices) connected to one or more communication networks via a communication interface 204 (e.g., wired or wireless); An application 230 acquires magnetic field data (e.g., via magnetometer 201) and / or processes the acquired data. In some embodiments, application 230 controls one or more components of device 200 and / or other connected devices (e.g., according to the acquired data). In some embodiments, application 230: ○ Acquisition module 232 that acquires magnetic data (e.g., magnetic field data) from magnetometer 201. In some embodiments, the magnetic data includes time-series magnetic data. ○ Processing module 234 that processes the magnetic field data. Data 242 for device 200, including, but not limited to, the following: ○Magnetic field data 244 ○Device settings 246 for device 200 (default options, acquisition settings, preferred user settings, etc.), and ○User setting 248.

[0131] In some embodiments, data collected during the data acquisition process is added as data 242.

[0132] Figure 2 shows a device (e.g., apparatus) 200, but Figure 2 is intended to be a functional description of various possible features rather than a structural schematic of the embodiments described herein. In practice, and as will be recognized by those skilled in the art, items shown separately can be combined, and some items can be separated.

[0133] Each of the sets of executable modules, applications, or procedures identified above may be stored in one or more of the aforementioned memory devices and correspond to a set of instructions for performing the above functions. The modules or programs (i.e., sets of instructions) identified above do not need to be mounted as separate software programs, procedures, or modules; therefore, various subsets of these modules can be combined or otherwise rearranged in various embodiments. In some embodiments, memory 206 stores a subset of the modules and data structures identified above. Furthermore, memory 206 may store additional modules or data structures not described above. In some embodiments, the programs, modules, and / or data subsets stored in memory 206 are stored in a server system and / or external devices (e.g., computer system 120) and / or executed by the server system.

[0134] Figure 3 is a block diagram showing a computer system 120 in several embodiments.

[0135] According to some embodiments, the computer system 120 includes one or more processors 302 (e.g., CPU processing units), one or more network interfaces 304, computer memory 306, and one or more communication buses 308 (sometimes called chipsets) for interconnecting these components.

[0136] The computer system 120 optionally includes one or more input devices 310 that facilitate user input, such as a keyboard, mouse, voice command input unit or microphone, touchscreen display, touch-sensitive input pad, gesture capture camera, or other input buttons or controls. In some embodiments, the computer system 120 optionally uses a microphone and voice recognition, or a camera and gesture recognition, to complement or replace the keyboard. The computer system 120 optionally includes one or more output devices 312 that enable the presentation of a user interface and display content, such as one or more speakers and / or one or more visual displays.

[0137] Memory 306 includes high-speed random-access memory such as DRAM, SRAM, DDR RAM, or other random-access solid-state memory devices, and optionally includes non-volatile memory such as one or more magnetic disk storage devices, one or more optical disk storage devices, one or more flash memory devices, or one or more other non-volatile solid-state storage devices. Memory 306 optionally includes one or more storage devices located remotely from one or more processors 302. Memory 306, or alternatively the non-volatile memory within Memory 306, includes a non-temporary computer-readable storage medium. In some embodiments, Memory 306, or the non-temporary computer-readable storage medium of Memory 306, stores the following programs, modules, and data structures, or subsets or supersets thereof: An operating system that includes procedures for handling various basic system services and performing hardware-dependent tasks; A user interface module 323 for enabling the presentation of information (e.g., graphical user interfaces for presenting applications, widgets, websites and their web pages, games, audio and / or video content, text, etc.) on any computer system or device, for example, device 200; A data receiving module 124 for receiving data (for example, from device 200); A data processing module 126 for calculating data received by the computer system 120. Further details of the data processing are shown in Figures 5 to 11; A training module 324 for generating and / or training a model, which can be used to identify and filter interference sources to improve the signal-to-noise ratio of magnetic field data, and / or to identify disease states of target organs from magnetic field measurements. In some embodiments, the training module 324 uses training data 334 to generate and / or train a model. In some embodiments, the training data 334 (e.g., training dataset) is generated from measurements from a device 200 (e.g., magnetometer 201). For example, in some embodiments, supervised machine learning is used to assist in denoising and signal and / or feature extraction. In some embodiments, a deep learning network such as a convolutional neural network (CNN) or autoencoder network is trained on signal components that can be classified as noise or heartbeat signals, and these models can be used to classify signal components in the data. Further transformations of the training data, including derivatives obtained from PCA or ICA, and data augmented with simulated signals or noise, can be used to further refine the model; Database 128 stores data used, received, and / or created by computer system 120. ;

[0138] In some embodiments, the memory 306 includes a machine learning database 130 for storing machine learning information. In some embodiments, the machine learning database 130 includes the following datasets or subsets or supersets thereof: The neural network data 328 includes information corresponding to the operation of one or more neural networks. In some embodiments, the neural network data 328 includes one or more models trained on a combination of realistic physical simulations of magnetic interference sources and real-world datasets. In some embodiments, the models can identify noise sources and remove them from the magnetic signal. In some embodiments, the models are trained on a combination of time-series data and magnetic field map data to classify different cardiac disease states. In some embodiments, the neural network data 328 includes: ○Training data 330, for example, a training dataset for training one or more models to identify differences between disease states (e.g., labeled by a trained expert). In some embodiments, the training data 330 includes signal data (e.g., from a target organ) and noise data.

[0139] In some embodiments, the computer system 120 includes a device registration module for registering devices to be used with the computer system 120 (e.g., computer devices, device 200, etc.).

[0140] Each of the elements identified above may be stored in one or more of the memory devices described herein and correspond to an instruction set for performing the functions described above. The modules or programs identified above do not need to be mounted as separate software programs, procedures, modules, or data structures, and therefore, in various embodiments, various subsets of these modules can be combined or otherwise rearranged. In some embodiments, memory 306 optionally stores a subset of the modules and data structures identified above. Furthermore, memory 306 optionally stores additional modules and data structures not described above. In some embodiments, a subset of programs, modules, and / or data stored in memory 306 is stored on and / or executed by device 200.

[0141] Devices for measuring biomagnetic signals Design considerations In magnetic measurement applications in unshielded environments, it is desirable to separate signals originating from the target source (signal) from signals originating from environmental sources (noise). This important distinction utilizes the fact that the amplitude of a magnetic field decreases as the distance from the magnetic source increases, following the magnetic field lines.

[0142] Based on the above, there are three parameters to consider when designing a sensor array: range, density, and regularity.

[0143] Range refers to the overall scale of the sensor array, i.e., the maximum distance between two magnetometers. A wider range allows for longer baseline measurements, enabling the identification of sources at various distances, such as distant (e.g., noise) sources and nearby (e.g., signal) sources. In particular, a larger gradient can better separate signals and noise, as insufficient spatial gradient field across the entire sensor array can impair signal detection. Applications in unshielded environments require a minimum array size (e.g., minimum extent) so that the magnetometer can sample magnetic fields (e.g., magnetic flux density) from distant sources. In contrast, sensor arrays operating in magnetically shielded (i.e., noise-free) environments have no constraints on array size / extent.

[0144] Density (or packing density) refers to the number of sensing locations per unit volume, which is determined by the spacing between sensors. The minimum density of a sensor array is determined by factors such as the size, weight, and power (SWaP) of each sensor type, the size and resolution of the spatial signal features to be captured, and potential crosstalk between sensors. For example, if a current dipole light source, such as a heart, is placed 10 cm below the sensor array, the spacing between sensors must be less than 10 cm to capture the broad features corresponding to the dipole pattern, and less than 4 cm to resolve the light source position with uncertainty bubbles smaller than a heart.

[0145] Regularity refers to how uniform the sensor arrangement is. Regularity is determined by the arrangement and orientation of the sensors. Regular arrangement is primarily a matter of convenience. In the case of a magnetometer that measures the vector component of a magnetic field, it is easiest if the sensor orientations are the same or vary regularly, because the magnetic field components must be combined later to correctly analyze the magnetic field. Furthermore, the irregularities and uncertainties resulting from this variation can affect calibration and spatial magnetic field information density. While these challenges may be overcome, the solutions may be complex. In the case of full-field sensors, the regularity of arrangement and orientation is insignificant for signal analysis, and in some cases, sensors can be densely packed at a density of one or more if necessary. When varying the density, one can consider increasing the density near the expected signal source and decreasing the density near the outer edge of the array.

[0146] Reference sensors (e.g., reference magnetometers) are commonly used in unshielded applications to provide measurements of background / noise magnetic fields without signal fields. For this purpose, the magnetometer is positioned at a certain distance from the sensor array so that the magnetic field from the signal source is expected to be smaller than the sensor's noise floor. For example, the magnetic field from a cardiac signal source should be well below 1 pT at a distance of approximately 30 cm from the source. However, a noise source 1 m away may be visible in both the sensor array and the reference sensor. Therefore, the signal from the reference sensor can be used as a regression of the sensor array data. To capture the highest-dimensional noise sources, the reference sensor should be positioned along three orthogonal axes at the shortest distance from the sensor array. In large sensor arrays, sensors along the edges of the array are considered reference sensors.

[0147] Table 2 shows that the following devices can operate at room temperature in an environment without magnetic shielding:

number

[0148] As shown in Table 2, each sensor type has specific advantages and disadvantages. In some cases, zero-field OPMs have a low dynamic range (e.g., ~10 nT) compared to other sensor types, which generally have a dynamic range about 10,000 times higher, making them unsuitable for use in non-magnetically shielded environments (i.e., where shielding is required). Zero-field OPMs require magnetic shielding (active or passive devices), making them unsuitable for use in portable / bedside devices. SQUIDs require cryogenic cooling, and considering the infrastructure required for their supply and maintenance, they are unsuitable for use in portable / bedside devices. NV diamond sensors meet SWaP requirements while offering the following sensitivities.

number

[0149] As shown in Table 2, full-field OPM sensors and fluxgate sensors are suitable candidates for bedside MCG systems because they satisfy all the attributes (i.e., "+") listed in all columns. Table 2 Comparison of bedside magnetocardiography sensor types JPEG2026511267000008.jpg73168

[0150] Figures 4A to 4E show exemplary embodiments of a device 200 in which magnetometers 201 are arranged in an array, according to several embodiments.

[0151] Figure 4A shows that device 200 includes one or more planes 402 (e.g., layers). In some embodiments, device 200 includes multiple planes (e.g., plane 402-a, plane 402-b, and / or plane 402-n, etc.) (e.g., multiple layers) arranged in a stack.

[0152] In some embodiments, each plane 402 in the planar stack has the same dimensions (e.g., the same length, width, and / or thickness).

[0153] In some embodiments, the planar stack includes a first plane and a second plane. The first plane has different dimensions (e.g., different length, width, and / or thickness) from the second plane.

[0154] In some embodiments, each plane 402 (e.g., plane 402-a) is aligned with the other planes 402 in the stack (e.g., planes 402-b and 402-n) in the length direction (e.g., x-axis direction). In some embodiments, each plane 402 (e.g., plane 402-a) is aligned with the other planes 402 in the stack (e.g., planes 402-b and 402-n) in the width direction (e.g., y-axis direction). In some embodiments, each of the planes 402 in the stack has the same x and y coordinates but different z coordinates.

[0155] Referring again to Figure 4A, in some embodiments, adjacent planes in the stack of planes (e.g., plane 402-a and plane 402-b) are separated by a spacing D4(408). In some embodiments, the spacing D4(408) is in the range of 0.5 cm to 2 cm.

[0156] Figure 4A shows that in some embodiments, one or more magnetometers 201 are positioned on (e.g., within or within) their respective planes 402. The magnetometers 201 are arranged in an array (e.g., having an ordered arrangement). For example, Figure 4A shows that the magnetometers are arranged in a rectangular array. In some embodiments, the magnetometers are arranged in a circular array, an elliptical array, a hexagonal array, or any polygonal array. In some embodiments, the array of magnetometers has an area that covers the entire cardiac activity area of ​​the patient.

[0157] In some embodiments, each magnetometer 201 comprises a diamond NV center magnetometer (e.g., a sensor). In some embodiments, each magnetometer 201 comprises an OPM (e.g., an OPM sensor). In some embodiments, each magnetometer 201 comprises a fluxgate sensor.

[0158] In some embodiments, each magnetometer 201 is as follows:

number

number

number

number

number

number

[0159] In some embodiments, each magnetometer 201 within the device 200 has a cubic shape having a dimension A1 (402) in the length direction (e.g., x direction or x-axis), a dimension A2 (404) in the width direction (e.g., y direction, y-axis or height direction), and a dimension A3 (406) in the thickness direction (e.g., z direction, z-axis, plane normal direction). In some embodiments, dimension A1 is in the range of 1 cm to 3 cm. In some embodiments, dimension A2 is in the range of 1 cm to 3 cm. In some embodiments, dimension A3 is in the range of 1 cm to 3 cm.

[0160] In some embodiments, the device 200 is also referred to as having a multi-planar (e.g., multilayer) array of magnetometers or a three-dimensional array of magnetometers. For example, in some embodiments, the device 200 comprises an array of m × n × p magnetometers, where m corresponds to the number of magnetometers in the longitudinal direction (e.g., x-axis) of the device 200, n corresponds to the number of magnetometers in the width direction (e.g., y-axis) of the device 200, and p corresponds to the number of layers or faces (e.g., plane 402) of the device 200. In some embodiments, m is a positive integer at least 3. In some embodiments, n is a positive integer at least 3. In some embodiments, p is a positive integer at least 2.

[0161] In some embodiments, the number of magnetometers 201 in device 200 is determined (e.g., optimized) according to the expected spatial characteristics of the signal source, including signals from other external sources (e.g., the Earth's magnetic field, external objects near the device) as well as signals from the target organ. For example, in some embodiments, the number of magnetometers 201 in device 200 is selected (e.g., optimized) based not only on the signal amplitude from the target organ (e.g., magnetic field strength) but also on the signal amplitude from external sources that may interfere with the signal from the target organ. For example, in MCG applications, two sensors are sufficient for signal acquisition in a relatively quiet, unshielded environment. Twenty-six magnetometers are used for disease biomarker extraction in noisy, unshielded environments. Thirty-six or more magnetometers are desirable for accurate source detection in unshielded environments.

[0162] As a general rule, MEG requires more sensors than MCG, and the actual number of sensors depends on the application. For example, observing alpha wave activity with MEG requires about 5 magnetometers. Observing motor cortex activity or vocal cortex activity requires about 5 magnetometers each (i.e., observing both activities requires a total of about 10 sensors). Detecting epileptic sources requires 256 or more magnetometers.

[0163] In some embodiments, the device 200 includes a first magnetometer located in a first plane of a planar stack and a second magnetometer 201 located in a second plane of a planar stack. The first magnetometer is aligned with the second magnetometer along the length and width of the magnetometer array. For example, referring to Figure 4A, the device 200 includes a magnetometer 201-1a located in plane 402-a. The device 200 includes a magnetometer 201-1b located in plane 402-b. In some embodiments, magnetometer 201-1a is aligned with magnetometer 201-1b in the x and y directions. In other words, magnetometers 201-1a and 201-1b have the same x and y coordinates but different z coordinates.

[0164] In some embodiments, the array of magnetometers includes a first subset (e.g., one or more) of magnetometers arranged in a first plane of a planar stack, and a second subset of magnetometers arranged in a second plane of the planar stack. The first subset of magnetometers cooperates with the second subset of magnetometers along the length and width of the array (e.g., the first subset and the second subset of magnetometers have the same x and y coordinates).

[0165] In some embodiments, the device 200 includes a first plane and a second plane adjacent to the first plane in the thickness direction (e.g., the z-axis). An array of magnetometers is arranged on the first plane (e.g., within the plane). An array of magnetometers is arranged on the second plane (e.g., within the second plane). Each magnetometer in the first plane is precisely registered with the corresponding magnetometer in the second plane (i.e., having the same x and y coordinates but different z coordinates). For example, referring to Figure 4A, the device 200 includes a plane 402-a and a plane 402-b adjacent to plane 402-a along the z-axis. Magnetometers such as magnetometers 201-1a, 201-3a, 201-qa, and 201-pa are arranged in plane 402-a. The second plane 402-b includes magnetometers such as magnetometers 201-1b, 201-3b, 201-qb, and 201-pb. Magnetometer 201-1a has the same x and y coordinates (and different z coordinates) as magnetometer 201-1b. Magnetometer 201-3a has the same x and y coordinates (and different z coordinates) as magnetometer 201-3b. Magnetometer 201-qa has the same x and y coordinates (and different z coordinates) as magnetometer 201-qb. Magnetometer 201-pa has the same x and y coordinates (and different z coordinates) as magnetometer 201-pb.

[0166] Figure 4E shows an exemplary arrangement of magnetometers in a three-dimensional coordinate system. Each data point in the coordinate system represents a magnetometer. In the example in Figure 4E, magnetometers MAG002B, MAG003B, MAG004B, MAG005B, and MAG007B are arranged on a first layer (e.g., a first plane) (e.g., plane 402-a), while magnetometers MAG002A, MAG003A, MAG004A, MAG005A, and MAG007A are arranged on a second layer (e.g., a second plane) (e.g., plane 402-b) that is different from the first layer. Magnetometers indicated by the same reference number (e.g., MAG002A and MAG002B, MAG003A and MAG003B, etc.) have the same x and y coordinates but different z coordinates.

[0167] Figure 4B shows a plan view of device 200 according to several embodiments. Figure 4C shows a front view of device 200 having an exemplary plan 402 according to several embodiments.

[0168] In some embodiments, the stack of planes 402 includes a total thickness D1(426). In some embodiments, D1(426) is in the range of 1 cm to 11 cm.

[0169] In some embodiments, adjacent planes in a stack of planes 402 (e.g., planes 402-a and 402-b) have a pitch D3(430) (e.g., the distance between centers in the z direction). In some embodiments, D3(430) is in the range of 1 cm to 3 cm.

[0170] In some embodiments, each plane 402 has a length B4(416) in the longitudinal direction. In some embodiments, B4(416) is in the range of 35 cm to 45 cm.

[0171] In some embodiments, each plane 402 has a width C4(424) in the width direction. In some embodiments, C4(424) is in the range of 35 cm to 45 cm.

[0172] In some embodiments, each plane 402 has a thickness D2(428) in the thickness direction. In some embodiments, D2(428) is in the range of 1 cm to 3 cm.

[0173] In some embodiments, adjacent magnetometers in each plane 402 (e.g., magnetometers 201-3 and 201-4 in Figure 4C) have a pitch B1(410) in the longitudinal direction (e.g., measured from the center of one magnetometer to the center of an adjacent magnetometer). In some embodiments, the pitch B1(410) is in the range of 1.5 cm to 3.5 cm.

[0174] In some embodiments, adjacent magnetometers in each plane 402 (e.g., magnetometers 201-1 and 201-2 in Figure 4C) are spaced apart in the longitudinal direction by a distance B2(412). In some embodiments, the distance B2(412) is in the range of 0.5 cm to 2 cm.

[0175] In some embodiments, each magnetometer 201 within the plane 402 has a combined length B3(414) in the longitudinal direction. In some embodiments, the combined length B3(414) is in the range of 30 cm to 40 cm.

[0176] In some embodiments, adjacent magnetometers in each plane 402 (e.g., magnetometers 201-5 and 201-7 in Figure 4C) have a pitch C1(418) in the width direction. In some embodiments, the pitch C1(418) is in the range of 1.5 cm to 3.5 cm.

[0177] In some embodiments, adjacent magnetometers in each plane 402 (e.g., magnetometers 201-5 and 201-6 in Figure 4C) are spaced apart in the width direction by a distance C2(420). In some embodiments, the distance C2(420) is in the range of 0.5 cm to 2 cm.

[0178] In some embodiments, each magnetometer 201 within the plane 402 has a combined width C3(422) in the width direction. In some embodiments, the combined width C3(422) is in the range of 30 cm to 40 cm.

[0179] Figure 4D shows the placement of the device 200 on a target organ (e.g., heart 442) of a human subject 440 according to several embodiments. In some embodiments, during device operation, the human subject 440 is positioned at a distance H1(444) from the device 200. In some embodiments, the distance H1(444) is between 0.1 cm and 5 cm. In some embodiments, the distance H1(444) is at least 1 cm (e.g., between 1 cm and 5 cm, between 1 cm and 8 cm, etc.).

[0180] According to some embodiments, the disclosed device 200 is an improvement over existing biomagnetic field sensing systems because, by arranging magnetometers in an array, it can sample magnetic fields over a spatial range, thereby distinguishing between signals originating closer to a target organ (e.g., the heart or brain) and signals originating further away from the target organ.

[0181] Furthermore, the arrangement and alignment of magnetometers (e.g., sensors) across multiple planes (e.g., same x and y positions) also improves the noise reduction process. Using device 200 in Figure 4D as an example, during device operation, plane 402-a is the closest to the target organ (e.g., heart 442), so magnetometers placed on plane 402-a (e.g., within plane 402-a or within plane 402-a) are primarily used to measure the magnetic field from the target organ, while magnetometers placed on other planes (e.g., planes 402-b, plane 402-n) are used to measure the magnetic field from the target organ, and magnetometers placed on planes 402-b, 402-n, etc.) are primarily used to measure the magnetic field from other external sources (note that magnetometers placed on each plane 402 simultaneously detect signals from both the target organ and other external sources). The configuration and isolation of the magnetometers facilitate signal processing using gradient geometry (e.g., z direction). In other words, in some embodiments, a signal from a target organ can be obtained by subtracting the signal detected by a magnetometer in a first plane (e.g., plane 402-a) from the signals detected by magnetometers in other planes (e.g., planes 402-b and / or plane 402-n).

[0182] Figure 4F shows an exemplary diagram of the MCG system 250 (e.g., device 200) according to several embodiments. The MCG device 250 utilizes an array of scalar OPMs (e.g., magnetometer 201) to eliminate the need for magnetic shielding.

[0183] Panel i of Figure 4F shows a photograph of the MCG system 250 with some of its components. The MCG system 250 includes a sensor head assembly 252 and its associated electronics, which are housed in a housing 251 (e.g., housing 143, Figure 2). Sensors such as a magnetometer 201 and their control modules are housed in the sensor head assembly 252 via the housing 251. The sensor head assembly 252 and the head support arm 254 can pivot around a point indicated by a circulating red arrow 255, allowing the operator to optimally position the device on the patient's chest. In some embodiments, the head support arm 254 includes a double-joint pivot that allows both coarse and fine rotation of the sensor head 252. The electronics rack 262 includes data acquisition electronics and other support components.

[0184] In some embodiments, the subject bed 260 is a magnetic resonance imaging (MRI)-compatible hospital-grade bed constructed from non-magnetic PVC. The sensor head 252 is mounted on a sliding gantry 258 and can be adjusted to conform to the patient's body length. In some embodiments, the gantry support (e.g., the sliding gantry 258) is an assembly made from extruded aluminum.

[0185] Panel ii in Figure 4F is a photograph of the bottom layer of sensors within the sensor housing 251. In some embodiments, the housing 251 contains up to one, two, three, four, five, or six or more layers of sensors. In some embodiments, the sensor mount 270 is constructed by 3D printing. In some embodiments, the sensor mount 270 is non-magnetic. In some embodiments, a magnetometer is mounted on the sensor mount 270. The example in Figure 4F shows that the sensor mount 470 can accommodate up to nine sensors per layer, but it will be apparent to those skilled in the art that in practice, the sensor mount can be constructed to contain nine or more sensors per layer or fewer than nine sensors per layer.

[0186] The right image in panel ii of Figure 4F shows a schematic diagram of the two sensor layers, indicating dimensions and gradient baselines. In the example shown in panel ii of Figure 4F, the housing 251 contains a two-layer scalar OPM array including five pairs of commercially available magnetometers arranged as shown. To ensure a high common-mode rejection ratio (CMRR), the magnetometers were paired as gradient meters with baselines perpendicular to the sensor array plane. The sensor spacing was chosen to maximize common-mode dispersion of distant noise sources and minimize dispersion of nearby signal sources.

[0187] Panel iii of Figure 4F shows a photograph of the sensor head 252 centered above the patient for data acquisition. The gap between the sensor head 252 and the patient is in the range of approximately 0.5 cm to 5 cm. The approximate direction of the Earth's magnetic field is indicated by the arrow Bearth. The scalar OPM measures small magnetic field fluctuations on the total magnetic field vector governed by the Earth's geomagnetic field in an unshielded environment. In some embodiments, the MCG system 250 (e.g., device 200) and the patient are positioned so that the Earth's magnetic field is approximately perpendicular to the chest surface. As a result, the measurement primarily records the cardiac magnetic field in the direction of the Earth's magnetic field.

[0188] In some embodiments, magnetometers in a plane (or layer, or array) can have varying packing densities. Figure 4G shows an exemplary plane 450 of a device 200 having magnetometers 201 with different packing densities according to some embodiments. Plane 450 includes a first set 452 (e.g., one or more) of magnetometers 201 with a higher packing density and a second set 454 (e.g., one or more) of magnetometers 201 with a lower packing density. During device use, the magnetometers with higher packing density (e.g., the first set 452) are positioned on the patient's chest (e.g., on the chest, aligned with the chest, close to the chest) to capture more signals, while the magnetometers with lower packing density (e.g., the second set 454) are positioned further away from the patient's chest to capture background (e.g., noise) signals.

[0189] In some embodiments, varying the packing density of the magnetometers can be achieved by varying the spacing between magnetometers in a plane. In other words, the magnetometers in a plane can be spaced irregularly. Referring again to the example in Figure 4G, the magnetometers 201 arranged along the x-axis can have different spacings E1, E2, E3, and E4, where E1 is the smallest and E4 is the largest. The magnetometers 201 arranged along the y-axis can have different spacings such as F1 and F2, where F1 is smaller than F2.

[0190] According to the sensor array shape design principles disclosed herein, magnetometers 201 within an array (or layer or plane) can have irregular spacing, as long as the spacing optimizes the magnetometer density based on their form factor. In some embodiments, magnetometers within an array can have irregular spacing, as long as sampling requirements are met (for example, the irregular spacing satisfies the Nyquist sampling rate in Fourier space of the wave vector of the magnetic field of the target organ, and the irregular spacing provides sufficient wavenumber coverage to retrieve information from both the biomagnetic field and the background magnetic field from the subject's organs).

[0191] In some embodiments, the MCG device / system 200 includes one or more reference sensors (e.g., magnetometers 201) for acquiring long-range (e.g., baseline or background) magnetic field signals. The reference sensors should be positioned far enough away from the source of interest (e.g., the patient's heart) so as not to capture the source signal. For example, in some embodiments, the reference sensors are distributed around the sensor array. In some embodiments, the reference sensors are positioned on the periphery (e.g., corners or edges) of a plane (e.g., plane 402) of the device 200. In some embodiments, the reference sensors can be positioned beyond the array (e.g., at distances of about 0.5 meters and 1 meter from the array, on the positioning arm 142, or on parts of the device 200 other than the array).

[0192] According to this disclosure, the dimensions / spacing discussed in the exemplary embodiments of Figures 4A to 4G satisfy the constraints in Fourier space when the myocardial source is located approximately 10 cm below the panel.

[0193] Other design considerations There are several relationships to consider when determining the dimensions of an array.

[0194] In some embodiments, the distance H1(444) and spacing D4(408) are determined (e.g., optimized) considering two factors: denoising and source localization. For denoising, the optimal gradiometer baseline distance (interplane and inplane spacing) is a function of the distance to the source of interest, as well as the spatial properties of noise from distant sources. The array should sample space in such a way that it can capture small variations in the source field and large features of the noise. In other words, for denoising, the values ​​of distance H1(444) and spacing D4(408) are chosen to maximize the dispersion of source grounds between sensors from distant noise sources while minimizing the dispersion of nearby signal sources. For source localization, the values ​​of distance H1(444) and spacing D4(408) are chosen to obtain the highest possible resolution, although the yield for tighter packing decreases given the sensor-source distance.

[0195] Generally speaking, smaller sensor sizes (e.g., smaller values ​​for A1, A2, and / or A3) allow more sensors to be placed on a plane, which translates to better resolution, better field mapping, and better source localization. However, given the distance between the source and the sensor (e.g., distance H1), there is little benefit in increasing the resolution beyond a certain point.

[0196] When designing array spacing (e.g., B1, B2, C1, C2), the optimal gradiometer baseline (i.e., the relative difference in magnetic field gradients between nearby and distant sources) must also be considered for optimal denoising. The optimal array spacing is determined by the expected spatial gradient of the signal magnetic field compared to the noise magnetic field. In an ideal world, we would have a densely packed array of sensors that completely fills the volume surrounding the object, providing both high resolution for source detection and large spatial sampling for denoising.

[0197] There is a correlation between the distance H1 (444) and the array parameters (e.g., dimensions such as B1, B2, B3, B4, C1, C2, C3, C4). As the distance H1 increases, the magnetometers in the array may not be packed as densely to fully characterize the signal field. Furthermore, as the distance H1 increases, the amplitude of the signal field decreases sharply, and its spatial gradient becomes smaller, reducing the sensor array's ability to discriminate between the signal field and other noise sources. Consequently, as the distance H1 increases, the overall size of the array (e.g., B3, B4, C3, C4) needs to be increased to sample larger spatial sites.

[0198] Signal processing One of the major challenges in detecting biomagnetic signals in environments that are not magnetically shielded is the extremely low signal-to-noise ratio of biomagnetic signals. For example, in the case of the human heart, the amplitude of the electrocardiogram signal from the heart is very small compared to magnetic interference sources outside the heart. The strength of the Earth's magnetic field is about 50 microtesla, which is about 1 million times the amplitude of the cardiac magnetic field signal expected when a biomagnetic field sensing device is placed just outside the patient's chest. Clinical environments are often noisy due to power line frequencies and their harmonics, as well as electrical equipment that radiates magnetic interference over a wide frequency range.

[0199] As described above, in some embodiments, the apparatus 200 uses a magnetometer 201, each configured to simultaneously detect a biomagnetic field and a background magnetic field from at least a portion of the subject's organs and to output signals indicating the detected biomagnetic and background magnetic fields. To decompose the electrocardiogram signal, it is important to reduce the contribution of noise in the collected signal and improve the signal-to-noise ratio of the electrocardiogram signal. Several aspects of this disclosure describe systems and methods for denoising cardiac signals collected using an unshielded (e.g., unmagnetically shielded) multichannel magnetometer (e.g., device 200). The technical effects of the denoising processes disclosed herein enable the system / apparatus to operate in a magnetically unshielded environment.

[0200] Figures 5A and 5B illustrate a workflow 500 for measuring magnetic fields from a target organ (e.g., heart or brain) of a human subject, according to several embodiments. In some embodiments, the steps of the workflow 500 are performed on a computer device (e.g., CPU 302 of computer system 120). In some embodiments, the computer device is communicably connected to a device (e.g., device 200).

[0201] In some embodiments, workflow 500 includes receiving data acquired by a device comprising multiple magnetometers (e.g., device 200 as shown in Figure 1B) (step 510). In some embodiments, the data comprises N-channel time-series magnetic data (512) acquired by a multi-channel magnetometer device, where N is the number of channels in the magnetometer device (e.g., device 200 as shown in Figures 1, 2, and 4A-4E). In some embodiments, each magnetometer is a single-channel sensor, and N corresponds to the total number of magnetometers in the device. In some embodiments, the device comprises magnetometers arranged in an array, such as device 200 shown in Figures 4A-4G. For example, if the array of magnetometers comprises an A×B array of magnetometers arranged in a C-plane stack, then N = A×B×C. In some embodiments, each magnetometer is a multi-channel sensor (e.g., a multi-axis sensor) that can contribute up to 3 channels. In some embodiments, each dataset is tagged with a unique identifier. In some embodiments, each dataset is cataloged according to its associated metadata.

[0202] In some embodiments, the computer device receives data in real time (e.g., automatically, without user intervention) as data is acquired by the magnetometer array. In some embodiments, the computer device receives data asynchronously at a time after the data has been acquired.

[0203] Figure 6A is a graph showing actual raw time-series magnetic field data (e.g., collected and generated by the array of magnetometers) from a multi-channel magnetometer array over a 10-second period, in several embodiments. The data in Figure 6A was collected by placing the array of magnetometers over a human subject at room temperature in a magnetically unshielded environment. Each line in the graph corresponds to the magnetic field signal collected by each channel of the array (the lines are offset for clarity). The magnetic signals include signals from the subject's target organs and signals from interference sources.

[0204] Referring again to Figure 5A, workflow 500 includes synchronizing N channels of time-series magnetic data to a common clock (e.g., a common time reference). This is illustrated as step 520 in Figure 5A. The channel synchronization step ensures that each channel maintains the same clock and timing. Time synchronization is also important for the epoch step (step 560).

[0205] In some embodiments, the computer device synchronizes the channels by interpolating and sampling clock drift (522).

[0206] In some embodiments, channel synchronization is performed by aligning a common signal injected into all channels.

[0207] In some embodiments, the computer device synchronizes the channels by using electromagnetic-based techniques (e.g., by using a common signal appearing in all channels) (524). An exemplary electromagnetic-based technique is a clapper coil 525 that provides a well-characterized calibration source. In some embodiments, the clapper coil 525 is configured to emit a unique electromagnetic signature (e.g., a signal) that is captured by the magnetometer channel while the biomagnetic field is being measured (e.g., during measurement, during measurement). The unique electromagnetic signature of the pulsating coil 525 can be used to precisely synchronize (e.g., align) signals from different channels of the device 200.

[0208] In some embodiments, the clapper coil 525 includes a coil of conductive wire (e.g., copper wire) positioned on the magnetometer device 200 and rigidly connected to the sensor array. For example, the clapper coil 525 can be fabricated by winding copper wire around a 3D-printed mounting core with a predetermined number of turns within a given volume. A periodic (e.g., sinusoidal) waveform can be driven through the coil using a current driver, and the resulting electromagnetic field can be detected by the magnetometer. Since the coil parameters are well known, a forward model of the expected magnetic field at each sensor position can be calculated and compared with the measurement results. For calibration to function properly, it is important to know the precise distance and orientation of the pulsating coil 525 relative to each magnetometer on the device 200. In some embodiments, the precise predetermined position of the clapper coil 525 can be specified in the same computer-aided design (CAD) drawing as the device.

[0209] In some embodiments, the computer device synchronizes channels using correlation-based offset correction (526) (for example, by cross-correlating the signals from the channels to determine the offset between channels).

[0210] In some embodiments, workflow 500 includes applying one or more filters to synchronized time-series magnetic data (step 530) to obtain filtered data. In some embodiments, the computer device applies an infinite impulse response (IIR) filter (e.g., a notch filter) at expected line noise (e.g., 60 Hz frequency) (532). In some embodiments, the computer device applies a bandpass filter (534), such as a bidirectional Butterworth digital filter, having a frequency between 0.5 and 40 Hz (or a similar frequency range applied in electrocardiogram (ECG) testing). Figure 6B is a graph of the same 10-second time-series raw magnetic field data as Figure 6A after the synchronization and filtering steps.

[0211] Continuing to refer to Figure 5A, in some embodiments, workflow 500 includes removing defective channels / epochs (step 540). Removal of defective channels / epochs can be automatic (542) or manual (544). For example, in some embodiments, a computer device automatically detects channels and / or data segments that are defective and / or have artifacts and removes those channels and / or data segments from subsequent analysis. Artifacts may include motion artifacts (e.g., a subject moving, someone bumping into the array or patient bed, or moving a support supporting the magnetometer array) and magnetic artifacts (e.g., an unexpected large magnetic disturbance).

[0212] Referring to Figure 5B, in some embodiments, workflow 500 includes applying one or more denoising (e.g., denoising) techniques to the filtered data (step 550). Up to this point, unless bad channels are removed in step 540, the N channels of filtered data in the N-channel magnetometer device remain.

[0213] Noise reduction approach In some embodiments, as shown in Figure 5B, the noise reduction technique includes one or more of the following: principal component analysis (PCA) (552), general linear regression gradometry (554), scaling / banding gradometry (556), independent component analysis (ICA) (557), signal spatial separation (SSS) (558), and signal spatial projection (SSP) (559).

[0214] PCA (or PCA filtering) is a technique for analyzing datasets containing a large number of dimensions / features per observation. In some embodiments, a computer device extracts a subset of variables from the filtered data, from all the variables in the filtered data. The subset of variables reframes the sensor space data into several signal spaces, each variable being data from the sensor channel and describing a portion of the statistical variance between all signals. In this disclosure, a variable that explains the largest variance in the sensor array (e.g., the first X component) is selected and considered common-mode noise from a distant source. These common-mode noise contributions are removed by setting these variables to zero before converting them to sensor space variables (i.e., time-series signals from the original set of magnetometers).

[0215] In some embodiments, PCA filtering is performed in the time domain. The PCA transformation reconstructs the real signal channels into a new set of channels (e.g., variables) that capture the largest variance in the total set of original signals. For example, if there are 10 original channels and 60Hz noise (e.g., wire noise) appears in all of the original channels, one of the new “channels” will be entirely 60Hz, while the other new “channels” will ideally contain nothing. As a result, if this 60Hz channel is set to zero before the transformation, the 60Hz noise is removed across the entire sensor array. Using this approach, for example, one such variable (principal component) can account for the variance between sensor signals originating from a large air conditioning unit in the room / building where the magnetometer array is located. Figure 6C is a graph of the same 10-second time-series raw magnetic field data as in Figure 6A, after synchronization, filtering, and PCA application.

[0216] In some embodiments, the noise reduction technique involves a non-scaled grapheometry in which the signals from two sensors are assumed to be calibrated and are simply subtracted from each other.

[0217] In some embodiments, the denoising technique includes linear regression gra-geometry (e.g., linear regression scaling gra-geometry). Linear regression is a technique for fitting known features to target data and has an overall scaling factor (and is therefore linear). Linear regression of gra-geometry works as follows: Typically, two signals have some kind of linear relationship. These can be called X and Y. In signal processing, the dependent signal Y is the target signal. In a particular case, Y is the signal recorded by the sensor closest to the heart. X is a signal feature representing unwanted interference; for example, a second sensor has a smaller amplitude heart signal, but the environmental interference should be equivalent. In a complete gra-geometry scenario, Y is the sum of the environmental noise X and the heart signal ε, and can also be interpreted as the residue from the linear relationship.

number

[0218] In equation (1), each index i (i=1,2,3,...) refers to a data point in time. The weight coefficient β is the same as the scale coefficient used to balance the gradiometer. In grageometry, the same weight coefficient β can be used for each index because the magnetic field measured by one sensor is linearly related to the magnetic field measured elsewhere. This collapses rapidly in the case of multiple nearby interference sources because magnetic fields originating from different directions have different scales. To solve this problem, a lateral sensor array is very useful because each orientation sensor has a principal feature that regresses from a central target sensor. In some cases, linearly regressively scaled grageometry is useful in situations where the greatest contribution to the signal is noise and is expected to be uniform across the two sensors in question (otherwise, the scaling of the result may be incorrect).

[0219] In some embodiments, the scale factor β for balancing the gradient meter can be calculated as the ratio between the mean or median values ​​of the two signals.

[0220] Figure 7 is a graph showing the actual signals from two geometric sensors positioned in an axial gradiometer configuration toward the heart. This graph shows that the relationship between sensor 1 and sensor 2 is linear.

[0221] The generalized linear model (GLM) of linear regression retains the same basic form (Y = Xβ + ε). However, now Y is considered to be the sum of one or more experimental signals (X), each multiplied by a weight coefficient (β) and with a random error (ε) added. As a result, X becomes an N × M matrix, where N is the number of time points and M is the number of signals to be regressed.

[0222] In GLM, both Y and ε remain single-column vectors containing the signal and heart rate from a single sensor, respectively, at consecutive time points (i=1 to n). However, the experimental design matrix (X) can consist of multiple columns, not just one. Each new column in X reflects a specific interference source ("regressor") that we want to eliminate. In the case of gra-geometry, these would be other sensor outputs, but they could also include the fluxgate signal, or the mean of the sensor array, or the subject's motion, or, in the case of a diamond NV center magnetometer, the reference photodiode output.

[0223] Figure 8 shows a GLM with Y, X, and ε represented as signals. This figure is taken from "Questions and Answers in MRI," available at mriquestions.com / general-linear-model.html, and is incorporated herein by reference in its entirety.

[0224] According to some embodiments of this disclosure, linear regression graphometry is applied to filtered data (e.g., N channels of filtered data). This process takes front and rear sensors (e.g., sensors having the same x and y coordinates but different z coordinates) and subtracts them as pairs using a scaling factor that considers linear regression for each of these signals. The processed signals for each pair of sensors after linear regression graphometry are: Signal sensor_1 minus β × Signal sensor_2 This can be expressed as follows. In this way, the scale of one sensor signal is matched to the signal of the other sensor. This reduces the maximum amount of signal originating from noise, almost always, when there is no magnetic shielding. In some embodiments, this value is β, which is determined (e.g., calibrated) for all MCG datasets corresponding to the patient scan. Figure 9A shows magnetic field signals from an array having two layers of magnetometers, where the magnetometers of the first layer are aligned with the magnetometers of the second layer in the length (x-axis) and width (y-axis) directions. Figure 9B shows the signal subtraction from the second layer to the first layer.

[0225] Figure 6D is a graph of the same 10-second time-series raw magnetic field data as in Figure 6A, after synchronization, filtering, and linear regression gradometry. Figure 6E compares the time-series magnetic field data after synchronization and filtering (i) with the data after synchronization, filtering, and linear regression gradometry (ii). A clear improvement in the signal-to-noise ratio of the magnetic field data after applying linear regression gradometry is observed.

[0226] In some embodiments, the denoising process includes applying linear regression graphometry to a synchronized and filtered dataset, and then applying PCA. Figure 6F shows the synchronized and filtered time-series magnetic data in Figure 6B after linear regression graphometry and PCA have been applied.

[0227] In some embodiments, the denoising process includes applying PCA followed by linear regression graphometry to the synchronized and filtered dataset. Figure 6G shows the synchronized and filtered time-series magnetic data in Figure 6B after PCA followed by linear regression graphometry.

[0228] The y-axis scale in the transition from Figure 6A to Figure 6B and Figure 6G clearly demonstrates strong background reduction. Noise is reduced by four orders of magnitude between the raw data in Figure 6A and the processed data in Figures 6F and 6G.

[0229] Unlike PCA, which does not change the number of channels when applied to a dataset, linear regression graphometry reduces the number of channels by half when applied to a dataset.

[0230] When the final number of signal channels is not critical (e.g., when looking for MCG waveforms instead of maps), linear regression graphometry provides a more direct and reliable method for removing common-mode noise, especially when electronic noise may be shared between sensor pairs. In lower-noise environments, linear regression graphometry is likely sufficient to recover the MCG signal from the background.

[0231] Regarding whether to apply linear regression or PCA (and which to apply first), the design of the magnetometer array (e.g., whether the magnetometer array is optimized to maximize the gradiometric signal) and the presence of magnetic common-mode noise should be considered. However, this largely depends on the nature of the noise and can vary from dataset to dataset in the high-noise environment in which the device operates. Many properties of the noise field can be considered, including amplitude, first and second-order spatial gradients, divergence and curl of the noise field, and more. In some situations, knowing these factors is impractical. In some embodiments, a heuristic approach is taken, in which different denoising pipelines are applied and the results are evaluated based on the resulting signal-to-noise ratio.

[0232] In some embodiments, the denoising process includes applying a scaled / banded gradometry 556. The scaled gradometry is similar to linear regression gradometry, except that instead of fitting a scale factor (e.g., β) based on linear regression, the signals from a pair of sensors are scaled based on the ratio between their central absolute deviations (MADs). In band gradometry, the reference signal is filtered into frequency bands that vary between 1–10 Hz segments, and each band is then used as a regression from the signal channel.

[0233] In some embodiments, applying banded graphometry involves filtering a reference channel by n different bidirectional bandpass filters, where the width of the bandpass window can vary between 1 and 10 Hz, resulting in n new time-series data from a given reference. Each of these n new time-series data is used as a regression in the GLM model of the target signal.

[0234] In some embodiments, the denoising process includes applying Independent Component Analysis (ICA)557. ICA is a blind source separation technique that separates multivariate signals into statistically maximally independent new components. The ICA algorithm assumes that the source signals are independent of each other and that the values ​​of each source have a non-Gaussian distribution. ICA is useful for denoising MCG signals because it can be reasonably assumed that the myocardial signal sources are independent of external noise sources. However, the effectiveness of ICA in separating source signals is limited because the cardiac sources are not stationary and can translate / rotate in space as cardiac sources.

[0235] Applying ICA after filtering and other denoising methods such as PCA is effective in decomposing the signal into cardiac and non-cardiac signal sources, even if the cardiac signal is dispersed across multiple components. While ICA components are not sorted by useful metrics, they can be epoched and averaged to aid in the classification of components containing myocardial signals. In some embodiments, the denoising techniques disclosed herein include discarding component sources that obviously do not contain cardiac waveform features.

[0236] In some embodiments, the noise reduction process includes applying signal space separation (SSS) 558 to the filtered data. SSS is a technique based on the physics of electromagnetic fields. SSS separates the measured signal into components originating from sources within the measurement volume of the magnetometer array (internal components) and components originating from sources outside the measurement volume (external components). Since the internal and external components are linearly independent, it is possible to simply discard the external components to reduce environmental noise. Typically, SSS and Maxwell filtering are performed together. Maxwell filtering is an associated procedure that removes higher-order components of the internal subspace dominated by sensor noise.

[0237] Signal Space Projection (SSP) 559 is another technique for removing noise from the MCG signal by projecting the signal onto a lower subspace. The subspace is selected by calculating the average pattern across the entire sensor when noise is present, treating that pattern as the "orientation" of the sensor space, and constructing the subspace to be orthogonal to the orientation of the noise. SSP can remove noise from the MCG signal when the noise originates from an environmental source (such as an electromagnetic field from nearby electrical equipment, or a source outside the subject's body and the MCG system) and the noise is stationary (does not change much during the recording period).

[0238] In some embodiments, the noise reduction technique is selected depending on the sensor type (e.g., whether the sensor is an OPM, a diamond NV center magnetometer, or a fluxgate sensor), or more specifically, whether the sensor is scalar or vector (e.g., uniaxial or multiaxial vector). Scalar OPMs in an array inherently have good relative alignment with each other, assuming the azimuthal magnetic field is spatially uniform and calibration is based on physical constants. Multiaxial vector sensors (such as diamond NV center magnetometers) may not have an external reference for aligning the sensing axes from multiple sensors, but their high channel density per volume suggests that statistical methods such as PCA and other extensional methods based on the spatial characteristics of the field (e.g., SSP or SSS) are more appropriate.

[0239] Continuing to refer to Figure 5B, after applying the noise reduction technique, workflow 500 proceeds to the epoch step (step 560). An epoch is a time period centered around a particular heartbeat. Specifically, an epoch can be defined as x ms before and y ms after the R-wave peak of a particular heartbeat. While x and y can be adjusted according to the desired analysis objective, typical values ​​are x=300 and y=500, resulting in an epoch window of 800 ms. This window length allows us to capture the characteristics of the P wave, QRS wave, and T wave for most subjects.

[0240] In some embodiments, the time series magnetic data is collected over a time period (e.g., 5 minutes or 10 minutes) and corresponds to a number of heartbeat events. Each heartbeat has a respective signal-to-noise ratio, and better data can be obtained by combining data (e.g., from one channel) and averaging the data. Channel synchronization of the data as described in step 520 is key to performing the subsequent steps of combining and averaging the data. In some embodiments, the computer device aligns subsets of the synchronized time series magnetic data corresponding to each magnetometer channel over a plurality of heartbeat events based on a trigger signal (e.g., the trigger signal can be a signal that identifies the start of a heartbeat, ventricular contraction, etc.) and generates respective subsets of the aligned signals. The computer device combines (e.g., aggregates, averages, etc.) each aligned subset of the synchronized time series magnetic data over the heartbeat events. Since the noise varies between the plurality of heartbeat events, the aggregation and averaging of the heartbeat signals reduces the noise while improving the signal. In some embodiments, the trigger signal is a signal detected by an electrocardiogram (ECG) (562). In some embodiments, the trigger signal is a signal detected by a magnetocardiogram (MCG) (564). In some embodiments, the trigger signal is a pulse oximeter. FIG. 10 shows a stacked epoch plot showing a single trial MCG visible in an unshielded operation. According to some embodiments of the present disclosure, the combination of sensor selection (e.g., scalar OPM, full-field OPM, fluxgate sensors, and / or NV diamond sensors), array design, and number of sensors on the multi-channel device / system 200, in combination with the disclosed noise removal process, enables the system / device to effectively reject ambient magnetic interference without magnetic shielding.

[0241] Continuing to refer to FIG. 5B, in some embodiments, after the epoxying, the workflow 500 proceeds to a field mapping step (step 570). Here, the computer device generates a magnetic field map by correlating (e.g., mapping) the magnetic field measurements to the real-space positions (e.g., spatial positions) of the magnetometers in the array. The computer device can then construct a spatial "heat map" for each slice of time across the sensor array. FIG. 11A shows a spatial magnetic field map at time slice t = 500 ms. FIG. 11B is a line graph showing the change in magnetic field per unit length over time for the pair of magnetometers MAG004A and MAG004B.

[0242] In some embodiments, the computer device uses prior information about the (e.g., fixed) array configuration (572) to map the magnetic field data of a channel to the corresponding position information of the magnetometers (e.g., spatial position, position coordinates, magnetometer element number in the magnetometer array, separation distance between magnetometers, etc.). In some embodiments, the computer device determines (e.g., acquires) the position information (574) of the magnetometers in the array in real time (e.g., when the computer device receives data, when the magnetic field data is collected by the magnetometer array, etc.) and correlates the position information with the magnetic field data collected by the channel.

[0243] Investigation Design and Setup According to several embodiments, the MCG system 250 shown in Figure 4F is deployed in an office environment close to operational railway tracks, power lines, and roads. To evaluate the system 250 in this magnetically unshielded environment, MCG data was acquired from 30 adult participants (multiple MCG sessions per participant, totaling 104 MCG datasets). Three-lead ECGs were recorded simultaneously to provide a high-signal-noise timing trigger for averaging the MCG data. Participants gave written informed consent before MCG data was collected in each session. Individuals with a history of heart disease, pacemakers or metal implants in the torso, or who were pregnant or breastfeeding were excluded from participation. Demographic information was not tracked or managed as this was a non-interventional feasibility study without demographic group comparisons.

[0244] The majority of participants (n=23) completed four different MCG sessions, each lasting 300 seconds (5 minutes), with varying measurement conditions: i) In resting (RS) sessions, participants sat quietly for 5 minutes before scanning and remained fully compliant (no speaking or moving) during scanning. ii) In sessions involving elevated heart rate or stress (SS), participants exercised for approximately 3 minutes before the scan and remained at rest during the scan. iii) In contrast, in non-compliant (NC) sessions, participants were asked to sit quietly for at least 5 minutes before the scan and to engage in active conversation with the device operator during the scan. iv) Finally, in the magnetic interference (MI) session, a steel plate was attached to the bed. This generated significant magnetic field artifacts, which were amplified by unintentional movements of the participants, including those related to heartbeat. The instructions given to the participants were the same as for the RS measurement.

[0245] Prior to MCG data acquisition, participants were instructed to lie supine on a gantry bed with the backrest adjusted 10 degrees above horizontal. Electrocardiograms were acquired using three non-magnetic surface electrodes (Ambu Neuroline 715) positioned on the right wrist (RA), left wrist (LA), and the lower right section of the thoracic cage (Ref). Subsequently, the MCG sensor head was positioned on each participant's thoracic cage using a repeatable protocol. Specifically, the central notch of the lid of the patient-facing sensor array was aligned with the participant's left eye and centered on the line connecting the armpits along the vertical axis. After centering the sensor array, an average distance of approximately 4 cm was maintained between the front of the sensor panel and each participant's chest, ensuring sufficient space for normal breathing during MCG acquisition without participants physically touching the sensor panel.

[0246] Figure 12 shows exemplary data collected from participants in MCG studies under several embodiments.

[0247] Panel i of Figure 12 shows a signal processing pipeline flowchart that includes the processing steps for time-series data acquired from a multi-channel sensor array in the study. After the data is loaded, channel synchronization is performed by aligning a common signal injected into all channels, including an upsampled ECG, to optimize trigger timing. Filtering is then performed, resulting in a 60Hz IRR notch filter and a 0.5-45Hz bandpass filter using bidirectional Butterworth digital filters. Next, faulty channels and segments are identified and removed from the data using automatic power thresholding and basic data checks. In the denoising step, noisy signal components are removed using a combination of gradometry and principal component analysis (PCA). MCG epochs are identified triggered by the ECG, and automatic epoch removal is performed based on signal power and timing criteria. Finally, the epochs are averaged, and the epoch average is visualized.

[0248] In the GraGeometry analysis, signals from adjacent sensors in the vertical direction (perpendicular to the chest) were subtracted to form a GraGeometry signal at a baseline of 6.5 cm. To ensure the calibration accuracy of the sensors, the common-mode field between sensors was used to measure possible deviations in the balance of the weights for GraGeometry. Linear regression was performed on the sensor pairs for this purpose. In practice, within the expected calibration limits of the sensors, all balance coefficients were equal to 1, confirming that the sensors were faithfully reporting the absolute magnetic field. In an unshielded environment, the common-mode dispersion between sensors is largely dominated by signals common to both the original and the gradiometer arrays. The original 10-channel PCA can be enhanced by making five GraGeometry channels available for training the PCA filter, while limiting the number of components to 10. This allows capturing common-mode signals that are incompletely canceled by the GraGeometry filtering. Typically, the first two to three principal components are identified as noise and removed before being reprojected into signal space.

[0249] In this study, by using gradometry and PCA, it became possible to distinguish cardiac signals from environmental noise, and we were able to obtain epoch-averaged MCG waveforms showing clear QRS and T-wave features under four different experimental conditions (resting state, high heart rate / stress state, participant non-compliance, and increased magnetic interference).

[0250] In the epoch / averaging step, ECG data was collected and analyzed in parallel with MCG to identify the time segments or epochs in which heartbeats occurred. ECG read 1 (LA-RA) was bandpass filtered between 0.5 Hz and 45 Hz using a bidirectional Butterworth digital filter. The filtered ECG was automatically thresholded to detect potential QRS times (false triggers were excluded based on relative timing). These trigger times were further refined by evaluating a 200 ms window around the initial estimate using a peak-finding algorithm. Using these triggers, the MCG data was divided into 1000 ms epochs. Epochs were excluded based on integrated signal power (epochs with up to 20% of signal power were excluded), and the remaining epochs were averaged together (Figure 12, panels ii and iii).

[0251] In panel ii of Figure 12, the epoch average of each gladometer channel is displayed based on its approximate relative position on the participant's chest. The upper right and lower left sensors show inverter characteristics. The upper plot in panel iii of Figure 12 is a superimposed epoch average of five gladometric signals. In panel iii, the epoch averages of five gladometric signals are superimposed. With the data in Figure 12, a total of 214 epochs can be averaged, but as shown in the inset in panel iii of Figure 12, an SNR > 10 is possible with only 60 epochs (~1 minute of data). ECG read I trace acquired simultaneously with MCG data. The corresponding ECG trace averaged over 214 epochs is shown in the lower plot in panel iii of Figure 2.

[0252] The averaged number of epochs per dataset was 191 (median 192), with a standard deviation of 41. This is consistent with achieving over 90% of the peak SNR within an average of 144 epochs, and with noise decreasing at a rate of 1 / sqrt(N) (where N is the average number).

[0253] In currently reported systems, epochs performed without the aid of ECG typically recover less than 30% of the heart rate, while ECG allows for the identification of 100% or very close to 100% heart rate. Unshielded interference is difficult to systematically remove from magnetometer data without distorting the heart rate morphology. Therefore, even after the noise reduction steps described above, the SNR of the time-series data for individual channels remains insufficient to consistently identify heart rate timing with the accuracy required for epoch averaging. In ischemia triage, the added information and reliability make the trade-off of adding an ECG subsystem worthwhile.

[0254] The resulting epoch averages for the five gradiometer channels were fitted using a general-purpose multi-peak fitting function designed to capture the feature amplitudes and inter-feature intervals of the QRS and T waves. Using these amplitudes, the signal-to-noise ratio (SNR) for each gradiometer channel was calculated, with the peak-to-peak of the QRS composite segment as the signal and the standard deviation of the segment 200-400 ms prior to the Q wave onset as the noise. This definition results in a lower SNR report than more standard power-based estimations, but better captures the feasibility of feature extraction. SNRmax is defined as the highest signal-to-noise ratio (SNR) across all five gradiometer channels in each dataset.

[0255] Each dataset was loaded via an automated pipeline that performed the steps described above with parameters selected by the algorithm. Datasets where the automated pipeline could not complete due to technical issues or excessive noise (approximately 30%) were manually analyzed using the same processing steps, but with manually selected parameters for identifying bad segments and denoising.

[0256] After data collection for this study was completed, group statistics were performed to understand the SNR distribution between subjects and experimental conditions. In particular, since it was found that the T-wave SNR depends on the relative position of each magnetic sensor relative to the heart, statistics on QRS-SNR are reported.

[0257] Mixed modeling (python statsmodels=0.14.1) was used to evaluate the hypothesis that there was no significant difference in SNR across the experimental conditions. The resting state was used as the reference group, and each of the other three experimental conditions was compared to the reference group using dummy coding. The dummy-coded variables (0, 1) were input as fixed effects at level 1 of a hierarchical linear model to predict the SNR outcome, while participants were input as random effects at level 2 of the same model. This configuration naturally yields three comparisons with the resting state (as the reference group), so a Bonferonni post-hoc correction was applied to adjust the critical alpha to 0.017.

[0258] In this technical feasibility test, 95 (91%) of the 104 MCG datasets acquired from the device showed a clear (SNR > 3) QRS complex between the gradiometer channels in the epoch-averaged MCG signal. In datasets with clear MCG signal visibility and in datasets before epoch averaging, the background white noise rejection ratio exceeded 100 at 1 Hz, while the common-mode rejection ratio of narrowband noise overlapping the MCG signal spectral content exceeded 1000. The epoch-averaged signals achieved an SNR of over 20 in 55% of the datasets, enabling the distinction between QRS and T-wave features corresponding to simultaneously acquired ECG features, as shown in panel iii of Figure 12.

[0259] Figure 13 shows the distribution of participant SNR results across conditions. Mixed modeling analysis revealed no significant differences between any condition or between the resting state (all p-values ​​were less than 0.056, and the critical α after post-hoc correction was 0.017). This result supports the hypothesis that the controlled experimental conditions did not significantly affect the SNR distribution.

[0260] Figure 13 summarizes the mean heart rate SNRmax for all participants by experimental condition. This summary is based on 23 participants and 92 observational data points. The mean heart rate for each participant was 191, and examining the scaling of SNR against heart rate reveals that this cannot explain the spread of SNR values. The wings of each violin plot represent the empirical distribution of participant results calculated by kernel density estimation (KDE). The mean SNR for each experimental condition is shown as a gold dot, and the asymmetric standard deviation from the mean of the participant-level distribution is reported as a thick black line. Comparing the reordered datasets using mixed modeling revealed no statistically significant differences in the distributions of each condition, indicating that the control factors in this study did not significantly affect SNR.

[0261] The average number of epochs across all datasets was 191, with a standard deviation of 41. In all datasets, background noise reduction was observed as 1 / sqrt(N) (where N is the number of epochs). This means that, for a constant signal amplitude, the SNR can vary by up to 11% based on the difference in the average number. Furthermore, it was observed that the SNR reached 90% of its final value (after averaging across all epochs) after an average of 141 epochs. On the other hand, the standard deviation of the SNR for resting-state data was approximately 50%. Therefore, the variability in averaged heart rate cannot account for the variability in SNR values.

[0262] Due to the low spatial resolution, a magnetic field map that can accurately estimate the dipole current source cannot be obtained. However, when qualitatively evaluating the MCG deflection from panel ii of FIG. 12, an inversion of the magnetic field amplitude from the upper right sensor to the lower left sensor is observed. This spatial feature is derived from the dipole angle expected during repolarization in healthy individuals. The bipolar angle is related to the angle of the equivalent current dipole during the T wave and has been shown to be sensitive to the ischemic state of the heart.

[0263] In this reported system operating in a harsh environment, for data processing without ECG triggering, the SNR is lower than 3 in most datasets, but including the ECG significantly improves the SNR. Since this device aims at deployment in clinical settings where environmental interference will almost certainly be high, even if sensor channels are added and noise removal algorithms are improved, the challenge of extracting a single heartbeat will become even greater.

[0264] In applications of the disclosed device, such as ischemic triage in the emergency treatment room, including the simultaneous ECG is known to bring a substantial improvement in the MCG SNR without inhibiting the clinical workflow.

[0265] Demonstration under a set of conditions for volunteers of the MCG device / system 250 showed that using a shieldless magnetometer, in a non-magnetically shielded environment, even when the participants are not completely calm and compliant and in the presence of magnetized devices, MCG recordings can be reliably obtained. The research published this time opens up the possibility of deploying this system at the bedside clinical setting.

[0266] Flowchart Figures 14A to 14D provide flowcharts of Method 700 for determining magnetic fields from organs of a human subject (e.g., the heart or brain of a human subject) according to several embodiments. Part of the process disclosed in Method 700 reduces / eliminates signals from interference sources, enabling the disclosed devices and systems to operate without magnetic shielding. Method 700 is performed on a computer system (e.g., computer system 120) having one or more processors (e.g., CPU 302) and memory (e.g., memory (306)). The memory stores one or more programs configured for execution by one or more processors. In several embodiments, the operations shown in Figures 1A, 1B, 4A to 4G, 5A, 5B, 6A to 6G, 7A, 7B, 8, 9A, 9B, 10, 11A, 11B, 12, and 13 correspond to instructions stored in memory 206 or other non-temporary computer-readable storage medium. The computer-readable storage medium may include magnetic or optical disk storage devices, solid-state storage devices such as flash memory, or other non-volatile memory devices or devices. In some embodiments, the instructions stored in the computer-readable storage medium include one or more of the following: source code, assembly language code, objective lens assembly, or other instruction formats interpreted by one or more processors. Some operations in Method 700 may be combined, and / or the order of some operations may be changed.

[0267] The computer system receives multiple signals corresponding to a first time-series magnetic data (e.g., magnetic field data) generated from multiple magnetometers (e.g., magnetic field sensors) located in close proximity to the human subject (e.g., 2 cm from the human subject) (702). The first time-series magnetic data corresponds to the magnetic field generated by the subject.

[0268] In some embodiments, the computer system and the magnetometers are jointly installed in the same location. For example, the computer system and the magnetometers are located in the same room where the magnetometers are deployed (e.g., employed). In some embodiments, the computer system and the magnetometers are located in different locations. In some embodiments, the computer system resides in the cloud.

[0269] In some embodiments, multiple magnetometers are not magnetically shielded.

[0270] In some embodiments, multiple magnetometers operate (e.g., are located) in an environment that is not magnetically shielded (e.g., a room that is not magnetically shielded).

[0271] In some embodiments, multiple magnetometers operate at room temperature.

[0272] In some embodiments, each of the multiple magnetometers includes an optical pumping magnetometer (OPM).

[0273] In some embodiments, each magnetometer in a plurality of magnetometers comprises a diamond nitrogen vacancy (NV) center magnetometer.

[0274] In some embodiments, each magnetometer in a plurality of magnetometers is equipped with a fluxgate sensor.

[0275] In some embodiments, the magnetometer array includes at least one OPM and at least one diamond NV center magnetometer.

[0276] In some embodiments, the plurality of magnetometers is an m × n array of magnetometers arranged in a stack of (704)p planes, where m is the number of magnetometers arranged in the length direction of the array (e.g., the x-axis direction), n is the number of magnetometers arranged in the width direction of the array (e.g., the y-axis direction), and p is the number of planes in the stack of planes arranged in the height direction of the array (e.g., the z-axis direction).

[0277] The computer system synchronizes the first time-series magnetic data to a common clock (e.g., a common time reference) (706) to generate synchronized time-series magnetic data.

[0278] The computer system applies one or more filters to the synchronous time-series magnetic data (708) to obtain filtered data.

[0279] In some embodiments, applying one or more filters to synchronous time-series magnetic data includes applying notch filters (e.g., infinite impulse response (IIR) filters or bandstop filters) having a frequency of line noise (e.g., 50 Hz or 60 Hz) (710).

[0280] In some embodiments, applying one or more filters to synchronous time-series magnetic data includes applying a bandpass filter (712).

[0281] In some cases, the bandpass filter includes a frequency range of (714) 0.5 Hz to 40 Hz.

[0282] Referring to Figure 14B, in some embodiments, the computer system applies one or more denoising (e.g., denoising) techniques to the filtered data (716) to generate updated time-series magnetic data. The updated time-series magnetic data has an improved signal-to-noise ratio compared to the first time-series magnetic data.

[0283] One or more noise reduction techniques may include linear regression graphometry. In some embodiments, the stack of p-planes includes a first plane and a second plane adjacent to the first plane. Applying one or more noise reduction techniques to filtered data includes obtaining positional information corresponding to each magnetometer in an array of magnetometers (718). In the array of magnetometers, for each pair of magnetometers consisting of each first magnetometer located in the first plane and each second magnetometer located in the second plane, each first magnetometer and each second magnetometer are aligned with respect to each other in the length and width directions, and a computer device calculates the difference between a first filtered signal corresponding to each first magnetometer and a second filtered signal corresponding to each second magnetometer (720).

[0284] Generally, in linear regression gra-geometry, this technique combines signals from a pair of magnetometers and calculates their difference, thus reducing the number of channels by half.

[0285] In some embodiments, calculating the difference between a first filtered signal corresponding to each first magnetometer and a second filtered signal corresponding to each second magnetometer further includes determining the correlation between the first filtered signal and the filtered second signal using linear regression (722) and calculating the difference according to the determined correlation (724).

[0286] As an example, each first magnetometer is magnetometer A, and each second magnetometer is magnetometer B. Magnetometers A and B are aligned along the length and width of the magnetometer array (i.e., they have the same x and y positions). The calculated difference is Signal A -β×Signal B Here, Signal A This corresponds to the filtered signal from magnetometer A, and Signal Bθ corresponds to the filtered signal from magnetometer B, and β is the scaling factor between the signal from magnetometer A and the signal from magnetometer B.

[0287] In some embodiments, applying one or more noise reduction techniques to filtered data includes applying principal component analysis (PCA) to extract a subset of variables from all variables in the filtered data (726). Unlike linear regression graphometry, where the number of signal channels is halved, PCA does not change the number of signal channels. In some embodiments, applying PCA includes determining a plurality of principal components (PCs) corresponding to the filtered data (727) and assigning zero values ​​to a subset of the plurality of PCs, thereby removing one or more noise contributions from the filtered data.

[0288] In some embodiments, applying one or more noise reduction techniques to filtered data includes applying source separation (SSS) to filtered data (728).

[0289] Continuing to refer to Figure 14C, in some embodiments, the synchronized time-series magnetic data comprises magnetic data recorded over multiple events (e.g., epoch or heartbeat events). Method 700 further includes, for each magnetometer in the array of magnetometers: aligning each subset of the synchronized time-series magnetic data corresponding to each magnetometer over multiple events based on a trigger signal (e.g., the trigger signal may be a signal that identifies the onset of a heartbeat, ventricular contraction, etc.) to generate each subset of the aligned signal; and combining (730) (e.g., aggregating, averaging, etc.) each aligned subset of the synchronized time-series magnetic data over multiple events.

[0290] In some embodiments, the computer system acquires positional information corresponding to each magnetometer in the array of magnetometers (e.g., position coordinates, magnetometer element number in the array of magnetometers, spacing between magnetometers, etc.) (732), and correlates (734) (e.g., maps) each signal of a plurality of signals with the acquired positional information.

[0291] In some embodiments, the updated time-series magnetic data includes a plurality of updated signals corresponding to each magnetometer in the array of magnetometers. The computer device (736) acquires positional information (e.g., spatial position, positional cooperating coordinates, magnetometer element number in the array of magnetometers, spacing between magnetometers, etc.) corresponding to each magnetometer in the array of magnetometers. Based on the acquired positional information, the computer device correlates (e.g., maps) each updated signal of the plurality of updated signals to its respective magnetometer. The computer device generates a magnetic field map (e.g., a two-dimensional map, data visualization) that spatially correlates the positional information of each magnetometer in the array of magnetometers with the magnetic field distribution around the organs of a human subject (740), and displays the magnetic field map on a display device (e.g., output device 312, or a user device communicably connected to the computer system) (742).

[0292] Referring to Figure 14D, in some embodiments, the multiple magnetometer comprises multiple vector magnetometers (744). Method 700 further includes receiving multiple signals corresponding to vector magnetic field signals from the multiple vector magnetometers, and then calibrating and processing the multiple signals according to the physical sensor orientation of the multiple vector magnetometers. In the case of two-axis or three-axis vector measurements, the method includes employing the methods disclosed above to take advantage of a higher information density. A vector magnetometer reporting two-axis or three-axis vector measurements from a single location in space has an inherently higher spatial information density than a single-axis or scalar magnetometer. Statistical methods such as PCA and SSS rely on the spatial sampling frequency of the sensor array, and the ability to separate nearby sources from distant sources benefits from a higher spatial information density.

[0293] The methods disclosed herein comprise one or more steps or actions for achieving the described method. The steps and / or actions of the method can be interchanged with one another without departing from the claims. In other words, unless a particular order of steps or actions is required for the proper operation of the described method, the order and / or use of any particular steps and / or actions can be changed without departing from the claims.

[0294] As used herein, the term “plural” refers to two or more. For example, “plural components” refers to two or more components. The term “determine” encompasses a wide variety of actions, and therefore “determine” can include calculation, computing, processing, derivation, investigation, lookup (e.g., looking up in a table, database or other data structure), confirmation, etc. “Determine” can also include receiving (e.g., receiving information), accessing (e.g., accessing data in memory), etc. “Determine” can also include resolving, selecting, choosing, establishing, etc.

[0295] The expression "based on" does not mean "based solely on" unless explicitly stated otherwise. In other words, the expression "based on" can mean both "based solely on" and "based on at least."

[0296] As used herein, the term “exemplary” means “to serve as an example, case, or illustration,” and does not necessarily imply that this example takes precedence or superiority over other configurations or embodiments.

[0297] In this specification, the term "and / or" includes any combination of the enumerated elements. For example, "A, B, and / or C" includes the following sets of elements: A only, B only, C only, A and B without C, A and C without B, B and C without A, and combinations of the three elements A, B, and C.

[0298] The terms used in this specification to describe the present invention are intended to describe specific embodiments and are not intended to limit the invention. Where used in the description of the present invention and in the appended claims, the singular forms "a," "an," and "the" are intended to include the plural form unless the context clearly indicates otherwise. Furthermore, where used herein, the terms "and / or" will be understood to refer to and include any and all possible combinations of one or more of the related enumerated items. In addition, where used herein, the terms "equipped with" and / or "equipped with" specify the presence of the described features, steps, operating states, elements, and / or components, but will not be understood to exclude the presence or addition of one or more other features, steps, operating states, elements, components, and / or groups thereof.

[0299] The above description has been based on reference to specific embodiments for illustrative purposes. However, the above exemplary discussion is not intended to be exhaustive or to limit the invention to the exact form disclosed. In view of the above teachings, many modifications and variations are possible. The embodiments have been selected and described to best illustrate the principles of the invention and its practical applications, thereby enabling those skilled in the art to best utilize the invention and its various embodiments with various modifications to suit specific intended uses. [Explanation of symbols]

[0300] 201 Magnetometer 250 MCG System 251 Housing 252 Sensor head assembly 254 Head support arm 256 Balance weight 258 Slide Gantry 260 subject beds 262 Electronic Equipment Rack 270 Sensor Mount

Claims

1. A device for measuring magnetic fields from a subject's organs, Equipped with multiple unshielded magnetometers in a three-dimensional arrangement, Each pair of magnetometers in the plurality of magnetometers has a known separation distance, Each of the plurality of magnetometers is configured to simultaneously detect the biomagnetic field from at least a portion of the subject's organs and the background magnetic field, and to output a signal indicating the detected biomagnetic field and the detected background magnetic field. The device is configured to operate without magnetic shielding. Device.

2. The apparatus according to claim 1, wherein each of the plurality of magnetometers responds to the total magnetic field adjacent to each magnetometer.

3. The apparatus according to claim 1 or 2, wherein each of the known separation distances has a fixed length.

4. The plurality of magnetometers include a first magnetometer having a variable position, and each pair of magnetometers of the plurality of magnetometers has a known separation distance that changes during device operation and is determined by tracking the position of the first magnetometer. The apparatus according to any one of claims 1 to 3.

5. The apparatus according to any one of claims 1 to 4, wherein the background magnetic field includes a uniform magnetic field.

6. The apparatus according to any one of claims 1 to 5, wherein the plurality of magnetometers have an average interval that satisfies constraints in Fourier space, and the average interval provides wave vector coverage for recovering information from both the biomagnetic field and the background magnetic field from the subject's organs.

7. The apparatus according to any one of claims 1 to 6, wherein the plurality of magnetometers are spatially distributed in Fourier space such that the plurality of magnetometers have wave vector coverage for recovering information from both the biomagnetic field from the subject's organs and the background magnetic field.

8. The plurality of magnetometers comprises an array of magnetometers arranged in a planar stack, The adjacent magnetometers in each plane of the stack of the plane are spaced apart by a first predetermined interval along the length of the array and by a second predetermined interval along the width of the array. The magnetometers in adjacent planes of the aforementioned stack are spaced apart by a third predetermined interval along the thickness direction of the array. The apparatus according to any one of claims 1 to 7.

9. The apparatus according to claim 8, wherein a first magnetometer in the first plane of the stack of planes is aligned with a second magnetometer in the second plane of the stack of planes along the length and width of the array.

10. The apparatus according to claim 8 or 9, wherein the array of magnetometers includes a first subset of magnetometers arranged in a first plane of the planar stack and a second subset of magnetometers arranged in a second plane of the planar stack, the first subset of magnetometers being aligned with the second subset of magnetometers along the length and width of the array.

11. The apparatus according to any one of claims 8 to 10, wherein each of the magnetometers in the array comprises an optical pumping magnetometer.

12. The apparatus according to any one of claims 8 to 10, wherein each of the magnetometers in the array comprises a diamond NV center magnetometer.

13. The apparatus according to any one of claims 8 to 10, wherein each of the magnetometers in the array is equipped with a fluxgate sensor.

14. A first subset of magnetometers in the array comprises an optical pumping magnetometer, A second subset of magnetometers in the array comprises a diamond NV center magnetometer. The apparatus according to any one of claims 8 to 10.

15. A first subset of magnetometers in the array comprises an optical pumping magnetometer, A second subset of magnetometers in the array comprises fluxgate sensors. The apparatus according to any one of claims 8 to 10.

16. The first subset of magnetometers in the array comprises a diamond NV center magnetometer, A second subset of magnetometers in the array comprises fluxgate sensors. The apparatus according to any one of claims 8 to 10.

17. A first subset of magnetometers in the array comprises an optical pumping magnetometer, A second subset of magnetometers in the array comprises fluxgate sensors, A third subset of magnetometers in the array comprises a diamond NV center magnetometer. The apparatus according to any one of claims 8 to 10.

18. The apparatus according to any one of claims 8 to 17, wherein the distance between the apparatus and the subject is 0.1 to 5 cm during the operation of the apparatus.

19. The apparatus according to any one of claims 8 to 18, wherein the apparatus is operable at room temperature.

20. The apparatus according to any one of claims 8 to 19, wherein the apparatus is capable of operating without magnetic shielding.

21. Each magnetometer in the aforementioned magnetometer array is as follows: [Number 13] The apparatus according to any one of claims 8 to 20, having superior sensitivity.

22. The apparatus according to any one of claims 8 to 21, wherein the first predefined interval is in the range of 0.1 cm to 2 cm.

23. The apparatus according to any one of claims 8 to 22, wherein the second predefined interval is in the range of 0.5 cm to 2 cm.

24. The apparatus according to any one of claims 8 to 23, wherein the third predefined interval is in the range of 0.5 cm to 2 cm.

25. The apparatus according to any one of claims 8 to 24, wherein adjacent magnetometers in each plane of the stack of planes have a pitch of 1.5 cm to 3.5 cm along the length of the array.

26. The apparatus according to any one of claims 8 to 25, wherein adjacent magnetometers in each plane of the stack of planes have a pitch of 1.5 cm to 3.5 cm along the width of the array.

27. The apparatus according to any one of claims 1 to 26, wherein each of the plurality of unshielded magnetometers has a dynamic range of about 50 microtesla.

28. The apparatus according to any one of claims 1 to 26, wherein each of the plurality of unshielded magnetometers has a dynamic range of about 100 microtesla.

29. A magnetically unshielded magnetometer system for measuring magnetic fields from the organs of a subject, Equipped with multiple unshielded magnetometers in a three-dimensional arrangement, Each pair of magnetometers in the plurality of magnetometers has a known separation distance, Each of the plurality of magnetometers is configured to simultaneously detect the biomagnetic field from at least a portion of the subject's organs and the background magnetic field, and to output a signal indicating the detected biomagnetic field and the background magnetic field. The magnetometer system is configured to operate without magnetic shielding. The aforementioned magnetometer system further, One or more processors, A memory for storing instructions to be executed by the one or more processors, Equipped with, The stored command is sent to each of the plurality of magnetometers: To simultaneously measure the biomagnetic field from at least a portion of the subject's organs and the background magnetic field, To output signals indicating the measured biomagnetic field and the measured background magnetic field, Including an order to perform, Magnetometer system.

30. The magnetometer system according to claim 29, wherein the stored command includes a signal processing command for separating the background magnetic field from the biomagnetic signal.

31. The magnetometer system according to claim 29 or 30, wherein the plurality of magnetometers are arranged within a housing.

32. A magnetometer system according to any one of claims 29 to 31, further comprising a positioning arm for supporting a plurality of unshielded magnetometers.

33. The magnetometer system according to claim 32, wherein the positioning arm is mounted on a base including one or more wheels.

34. The magnetometer system according to claim 32 or 33, wherein the positioning arm is mounted on a patient support platform.

35. A magnetometer system according to any one of claims 32 to 34, comprising a panel for mounting the plurality of unshielded magnetometers, wherein the panel is supported by the positioning arm.

36. The magnetometer system according to claim 35, wherein the positioning arm is configured to maintain a gap of 0.5 cm to 5 cm between the panel and the subject's chest during the entire scan of the patient.

37. The magnetometer system according to any one of claims 32 to 36, wherein the positioning arm includes one or more lockable degrees of freedom of movement.

38. The magnetometer system according to any one of claims 32 to 37, wherein the positioning arm includes three coupled degrees of freedom (DOF).

39. The magnetometer system according to any one of claims 29 to 38, wherein each of the plurality of magnetometers responds to the total magnetic field adjacent to each of the magnetometers.

40. The magnetometer system according to any one of claims 29 to 39, wherein each of the known separation distances has a fixed length.

41. The plurality of magnetometers include a first magnetometer having a variable position, A magnetometer system according to any one of claims 29 to 40, wherein each pair of magnetometers in the plurality of magnetometers has a known separation distance that changes during device operation and is determined by tracking the position of the first magnetometer.

42. The magnetometer system according to any one of claims 29 to 41, wherein the background magnetic field includes a uniform magnetic field.

43. The magnetometer system according to any one of claims 29 to 42, wherein the plurality of magnetometers have an average interval that satisfies constraints in Fourier space.

44. The magnetometer system according to any one of claims 29 to 43, wherein the plurality of magnetometers are spatially distributed in Fourier space to have wave vector coverage for collecting information from both the biomagnetic field from the subject's organs and the background magnetic field.

45. The plurality of magnetometers comprises an array of magnetometers arranged in a planar stack, The adjacent magnetometers in each plane of the stack of the plane are spaced apart by a first predetermined interval along the length of the array and by a second predetermined interval along the width of the array. The magnetometers in adjacent planes of the aforementioned stack are spaced apart by a third predetermined interval along the thickness direction of the array. A magnetometer system according to any one of claims 29 to 44.

46. The magnetometer system according to claim 45, wherein a first magnetometer in a first plane of the stack of planes is aligned with a second magnetometer in a second plane of the stack of planes along the length and width of the array.

47. The array of magnetometers includes a first subset of magnetometers arranged in a first plane of the planar stack and a second subset of magnetometers arranged in a second plane of the planar stack. The first subset of the magnetometers is aligned with the second subset of the magnetometers along the length and width of the array. The magnetometer system according to claim 45 or 46.

48. The magnetometer system according to any one of claims 45 to 47, wherein each of the magnetometers in the array comprises an optical pumping magnetometer.

49. The magnetometer system according to any one of claims 45 to 47, wherein each of the magnetometers in the array includes a diamond NV center magnetometer.

50. The magnetometer system according to any one of claims 45 to 47, wherein each of the magnetometers in the array includes a fluxgate sensor.

51. A magnetometer system according to any one of claims 45 to 47, wherein a first subset of magnetometers in the array includes an optical pumping magnetometer, and a second subset of magnetometers in the array includes a diamond NV center magnetometer.

52. A magnetometer system according to any one of claims 45 to 47, wherein a first subset of magnetometers in the array includes an optical pumping magnetometer, and a second subset of magnetometers in the sensor array includes a fluxgate sensor.

53. A magnetometer system according to any one of claims 45 to 47, wherein a first subset of magnetometers in the array includes a diamond NV center magnetometer, and a second subset of magnetometers in the array includes a fluxgate sensor.

54. A magnetometer system according to any one of claims 45 to 47, wherein a first subset of magnetometers in the array includes an optical pumping magnetometer, a second subset of magnetometers in the sensor array includes a fluxgate sensor, and a third subset of magnetometers in the array includes a diamond NV center magnetometer.

55. Each magnetometer in the array of magnetometers, [Number 14] A magnetometer system according to any one of claims 45 to 54, having superior sensitivity.

56. The magnetometer system according to any one of claims 45 to 55, wherein the first predefined interval is in the range of 0.1 cm to 2 cm.

57. The magnetometer system according to any one of claims 45 to 56, wherein the second predefined interval is in the range of 0.5 cm to 2 cm.

58. The magnetometer system according to any one of claims 45 to 57, wherein the third predefined interval is in the range of 0.5 cm to 2 cm.

59. The magnetometer system according to any one of claims 45 to 58, wherein adjacent magnetometers in each plane of the stack of planes have a pitch of 1.5 cm to 3.5 cm along the length of the array.

60. The magnetometer system according to any one of claims 45 to 59, wherein adjacent magnetometers in each plane of the stack of planes have a pitch of 1.5 cm to 3.5 cm along the width of the array.

61. The magnetometer system according to any one of claims 29 to 60, wherein the distance between the device and the subject is 0.1 to 5 cm during the operation of the magnetometer system.

62. The magnetometer system according to any one of claims 29 to 61, wherein the magnetometer system is operable at room temperature.

63. The magnetometer system according to any one of claims 29 to 62, wherein the magnetometer system is capable of operating without magnetic shielding.

64. The magnetometer system according to any one of claims 29 to 63, wherein each of the plurality of unshielded magnetometers includes an optical pumping magnetometer, a diamond NV center magnetometer, or a fluxgate sensor.

65. The magnetometer system according to any one of claims 29 to 64, wherein each of the plurality of unshielded magnetometers has a dynamic range of about 50 microtesla.

66. The apparatus according to any one of claims 29 to 64, wherein each of the plurality of unshielded magnetometers has a dynamic range of approximately 100 microtesla.

67. A method for determining the magnetic field from the organs of a human subject, In a computer system having one or more processors and memory, The method involves receiving multiple signals corresponding to first time-series magnetic data generated from multiple unshielded magnetometers located in close proximity to the human subject, The first time-series magnetic data corresponds to the magnetic field generated by the human subject, and the plurality of signals include contributions from the biomagnetic field from at least a portion of the subject's organs and the background magnetic field, to be received. Synchronizing the first time-series magnetic data with a common clock to generate synchronized time-series magnetic data, The process involves applying one or more filters to the aforementioned synchronous time-series magnetic data to obtain filtered data. The filtered data is subjected to one or more noise reduction techniques to generate updated time-series magnetic data. Methods that include...

68. The method according to claim 67, wherein applying one or more filters to the synchronous time-series magnetic data includes applying a notch filter having the frequency of power line noise.

69. The method according to claim 67 or 68, wherein applying one or more filters to the synchronous time-series magnetic data includes applying a bandpass filter.

70. The method according to claim 69, wherein the bandpass filter includes a frequency range of 0.5 Hz to 40 Hz.

71. The method according to any one of claims 67 to 70, wherein the plurality of magnetometers comprises m × n arrays of magnetometers arranged in a stack of p planes, where m is the number of magnetometers in the length direction of the array, n is the number of magnetometers in the width direction of the array, and p is the number of planes in the stack of planes arranged in the height direction of the array.

72. The stack of p planes includes a first plane and a second plane adjacent to the first plane. Applying one or more noise reduction techniques to the filtered data is: To acquire positional information corresponding to each magnetometer in the array of magnetometers, In the array of magnetometers, for each pair of magnetometers consisting of a first magnetometer arranged in the first plane and a second magnetometer arranged in the second plane, the first magnetometer and the second magnetometer are aligned with respect to each other in the length and width directions, and the difference between a first filtered signal corresponding to each first magnetometer and a second filtered signal corresponding to each second magnetometer is calculated. The method according to claim 71, including the method described in claim 71.

73. Further, the difference between the first filtered signal corresponding to each of the first magnetometers and the second filtered signal corresponding to each of the second magnetometers is calculated. The correlation between the first filtered signal and the second filtered signal is determined using linear regression. Calculate the difference according to the determined correlation, The method according to claim 72, including the method described in claim 72.

74. The method according to any one of claims 67 to 73, wherein applying one or more noise reduction techniques to the filtered data includes applying principal component analysis (PCA) to extract a subset of variables from all variables of the filtered data.

75. The method according to claim 74, wherein applying PCA includes determining a plurality of principal components (PCs) corresponding to the filtered data, and removing one or more noise contributions from the filtered data by assigning zero values ​​to a subset of the plurality of PCs.

76. The method according to any one of claims 67 to 75, wherein applying one or more noise reduction techniques to the filtered data includes applying signal source separation (SSS) to the filtered data.

77. The plurality of magnetometers include an array of magnetometers, The aforementioned synchronous time-series magnetic data includes magnetic data recorded across multiple events. The aforementioned method further, For each magnetometer in the aforementioned magnetometer array, Based on the trigger signal, the respective subsets of the synchronous time-series magnetic data corresponding to each of the magnetometers across the multiple events are aligned to generate the respective subsets of the aligned signals. Combining the aligned subsets of the synchronous time-series magnetic data across the aforementioned multiple events, The method according to any one of claims 67 to 76, including

78. The plurality of magnetometers include an array of magnetometers, The aforementioned method further, To acquire positional information corresponding to each magnetometer in the array of magnetometers, Correlating each of the aforementioned multiple signals with the acquired position information, The method according to any one of claims 67 to 77, including

79. The plurality of magnetometers include an array of magnetometers, The updated time-series magnetic data includes a plurality of updated signals corresponding to each magnetometer in the array of magnetometers, The aforementioned method further, To acquire positional information corresponding to each of the aforementioned multiple magnetometers, Based on the acquired location information, the updated signals of each of the multiple updated signals are correlated to their respective magnetometers. To generate a magnetic field map by spatially correlating the magnetic field distribution of the organs of the human subject with the positional information of each magnetometer in the magnetometer array, Displaying the aforementioned magnetic field map on a display device, The method according to any one of claims 67 to 78, including

80. The method according to any one of claims 67 to 79, wherein the plurality of magnetometers include at least one optical pumping magnetometer (OPM).

81. The method according to any one of claims 67 to 80, wherein the plurality of magnetometers include at least one diamond nitrogen vacancy (NV) center magnetometer.

82. The method according to any one of claims 67 to 79, wherein each of the plurality of magnetometers includes an OPM.

83. The method according to any one of claims 67 to 79, wherein each of the plurality of magnetometers includes a diamond nitrogen vacancy (NV) center magnetometer.

84. The method according to any one of claims 67 to 79, wherein each of the plurality of magnetometers includes a fluxgate magnetometer.

85. The plurality of magnetometers include a plurality of vector magnetometers, The above method further, The process includes receiving the plurality of signals corresponding to the vector magnetic field signals from the plurality of vector magnetometers, and then calculating the dot product of the vector magnetic field components of the plurality of vector magnetometers to derive the first time-series magnetic data. The method according to any one of claims 67 to 79.

86. A computer system for determining magnetic fields from the organs of human subjects, One or more processors, Memory and One or more programs stored in the memory for execution by the one or more processors, Equipped with, A computer system comprising one or more programs, each including instructions for performing the method according to any one of claims 67 to 85.

87. A computer-readable storage medium storing computer-executable instructions that, when executed by one or more processors of a computer system, cause the computer system to perform the method according to any one of claims 67 to 85.

Images