Signal processing methods and systems for biomagnetic field imaging
A computer system receives a plurality of signals corresponding to first time-series magnetic data generated from a plurality of unshielded magnetometers proximate to the human subject. The first time-series magnetic data corresponds to magnetic fields generated from the human subject. The plurality of signals includes contributions from a biomagnetic field from at least a portion of the subject's organ and a background magnetic field. The computer system synchronizes the first time-series magnetic data to a common clock to generate synchronized time-series magnetic data. The computer system applies one or more filters to the synchronized time-series magnetic data to obtain filtered data. The computer system applies one or more noise reduction techniques to the filtered data to generate updated time-series magnetic data.
1. A method for determining magnetic fields from an organ of a human subject, comprising:
at a computer system having one or more processors and memory:
receiving a plurality of signals corresponding to first time-series magnetic data generated from a plurality of unshielded magnetometers proximate to the human subject, the first time-series magnetic data corresponding to magnetic fields generated from the human subject, wherein the plurality of signals include contributions from a biomagnetic field from at least a portion of the subject's organ and a background magnetic field;
synchronizing the first time-series magnetic data according to a unique electromagnetic signature signal that is present in each of the plurality of signals, to generate synchronized time-series magnetic data;
applying one or more filters to the synchronized time-series magnetic data to obtain filtered data; and
applying one or more noise reduction techniques to the filtered data to generate updated time-series magnetic data.
2. The method of claim 1 , wherein applying one or more filters to the synchronized time-series magnetic data comprises applying a notch filter having a frequency of an electrical line noise.
3. The method of any of claim 1 , wherein applying one or more filters to the synchronized time-series magnetic data comprises applying a bandpass filter.
4. The method of claim 3 , wherein the bandpass filter includes a frequency range of 0.5 Hz to 40 Hz.
5. The method of claim 1 , wherein:
the plurality of magnetometers comprises an m×n array of magnetometers arranged in a stack of p planes; and
m is a number of magnetometers in a length direction of the array, n is a number of magnetometers in a width direction of the array, and p is a number of planes, in the stack of planes, arranged in a height direction of the array.
6. The method of claim 5 , wherein:
the stack of p planes includes a first plane and a second plane that is adjacent to the first plane; and
applying one or more noise reduction techniques to the filtered data includes:
obtaining positional information corresponding to each magnetometer in the array of magnetometers; and
for a respective pair of magnetometers, in the array of magnetometers, consisting of a respective first magnetometer positioned in the first plane and a respective second magnetometer positioned in the second plane, wherein the respective first magnetometer and the second respective magnetometer are aligned with respect to each other in the length direction and the width direction:
calculating a difference between a first filtered signal corresponding to the respective first magnetometer and a second filtered signal corresponding to the second respective magnetometer.
7. The method of claim 6 , wherein calculating the difference between the first filtered signal corresponding to the respective first magnetometer and the second filtered signal corresponding to the second respective magnetometer further includes:
determining a correlation between the first filtered signal and the filtered second signal using linear regression; and
calculating the difference according to the determined correlation.
8. The method of claim 1 , wherein applying one or more noise reduction techniques to the filtered data includes applying principal component analysis (PCA) to extract, from the filtered data, a subset of variables from all variables of the filtered data.
9. The method of claim 8 , wherein applying PCA includes:
determining a plurality of principal components (PCs) corresponding to the filtered data; and
assigning a value of zero to a subset of the plurality of PCs, thereby removing one or more noise contributions from the filtered data.
10. The method of claim 1 , wherein applying one or more noise reduction techniques to the filtered data includes applying signal source separation (SSS) to the filtered data.
11. The method of claim 1 , wherein:
the plurality of magnetometers comprises an array of magnetometers;
the synchronized time-series magnetic data comprises magnetic data that are recorded over a plurality of events; and
the method further comprises, for a respective magnetometer of the array of magnetometers:
aligning a respective subset of the synchronized time-series magnetic data corresponding to the respective magnetometer over the plurality of events based on a trigger signal to generate a respective subset of aligned signals; and
combining the respective aligned subset of the synchronized time-series magnetic data over the plurality of events.
12. The method of claim 1 , wherein:
the plurality of magnetometers comprises an array of magnetometers; and
the method further comprises:
obtaining positional information corresponding to each magnetometer in the array of magnetometers; and
correlating a respective signal of the plurality of signals to the obtained positional information.
13. The method of claim 1 , wherein:
the plurality of magnetometers comprises an array of magnetometers;
the updated time-series magnetic data includes a plurality of updated signals corresponding to respective magnetometers in the array of magnetometers; and
the method further comprises:
obtaining positional information corresponding to each magnetometer in the plurality of magnetometers;
correlating a respective updated signal of the plurality of updated signals to a respective magnetometer based on the obtained positional information;
generating a magnetic field map that spatially correlates a magnetic field distribution of the organ of the human subject with the positional information of each magnetometer in the array of magnetometers; and
causing display of the magnetic field map on a display device.
14. The method of claim 1 , wherein the plurality of magnetometers includes at least one optically pumped magnetometer (OPM).
15. The method of claim 1 , wherein the plurality of magnetometers includes at least one diamond nitrogen vacancy (NV) center magnetometer.
16. The method of claim 1 , wherein the plurality of magnetometers includes at least one fluxgate magnetometer.
17. The method of claim 1 , wherein:
the plurality of magnetometers comprises a plurality of vector magnetometers; and
the method further comprises:
after receiving the plurality of signals, corresponding to vector field signals from the plurality of vector magnetometers, computing a dot product of vector field components of the plurality of vector magnetometers to derive the first time-series magnetic data.
18. A computer system for determining magnetic fields from an organ of a human subject, comprising:
one or more processors; and
memory; and
one or more programs stored in the memory for execution by the one or more processors, the one or more programs including instructions for:
receiving a plurality of signals corresponding to first time-series magnetic data generated from a plurality of unshielded magnetometers proximate to the human subject, the first time-series magnetic data corresponding to magnetic fields generated from the human subject, wherein the plurality of signals include contributions from a biomagnetic field from at least a portion of the subject's organ and a background magnetic field;
synchronizing the first time-series magnetic data according to a unique electromagnetic signature signal that is present in each of the plurality of signals, to generate synchronized time-series magnetic data;
applying one or more filters to the synchronized time-series magnetic data to obtain filtered data; and
applying one or more noise reduction techniques to the filtered data to generate updated time-series magnetic data.
19. The computer system of claim 18 , wherein the instructions for applying one or more filters to the synchronized time-series magnetic data include instructions for:
applying a notch filter having a frequency of an electrical line noise.
20. The computer system of claim 18 , wherein each magnetometer in the plurality of magnetometers comprises:
an optically pumped magnetometer (OPM);
a diamond nitrogen vacancy (NV) center magnetometer; or
a fluxgate magnetometer.
21. The computer system of claim 18 , wherein:
the plurality of magnetometers comprises an m×n array of magnetometers arranged in a stack of p planes; and
m is a number of magnetometers in a length direction of the array, n is a number of magnetometers in a width direction of the array, and p is a number of planes, in the stack of planes, arranged in a height direction of the array.
22. The computer system of claim 21 , wherein the stack of p planes includes a first plane and a second plane that is adjacent to the first plane.
23. A non-transitory 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 operations for determining magnetic fields from an organ of a human subject, the operations comprising:
receiving a plurality of signals corresponding to first time-series magnetic data generated from a plurality of unshielded magnetometers proximate to the human subject, the first time-series magnetic data corresponding to magnetic fields generated from the human subject, wherein the plurality of signals include contributions from a biomagnetic field from at least a portion of the subject's organ and a background magnetic field;
synchronizing the first time-series magnetic data according to a unique electromagnetic signature signal that is present in each of the plurality of signals, to generate synchronized time-series magnetic data;
applying one or more filters to the synchronized time-series magnetic data to obtain filtered data; and
applying one or more noise reduction techniques to the filtered data to generate updated time-series magnetic data.