Patent Application
20180313908
This disclosure relates to the field of calibrating a magnetic imaging array, specifics y a system and method thereof.
Optically pumped magnetometers (OPMs), also called atomic magnetometers, optical magnetometers, or optical atomic magnetometers, are used in a number of scientific and advanced technology applications including medical imaging. In their simplest form, these sensors contain a light source, a container to hold atoms, and a detector. The light source may be a laser or other optical device used to produce light of a certain wavelength. The container may be a vapor cell or other device used to house atoms. The detector would necessarily be specific to the light output.
Single OPMs or small arrays of OPMs have been used routinely to create magnetic field images or gradient magnetic field images and to localize magnetic sources. In many cases, the sensors or sensor arrays are mounted onto moving platforms and moved in regular patterns over the area of interest. Alternatively, larger arrays allow the sensors to be stationary. In order to localize magnetic sources, the positions of the sensors have to be known. For large area images, the sensor location can be determined with global navigation satellite systems (GNSS), such as but not limited to the global positioning system (GPS). For smaller areas of interest, sensor positions have been determined geometrically or optically. For some OPMs, an additional complication comes from the fact that the position at which the magnetic field is measured, is determined by the position of the light or laser beam, not a physical component of the sensor.
There are several other factors that determine the quality of the image and the source localization apart from the locations of the sensors in the array, such as but not limited to, sensor orientation, sensor gain as a function of frequency, sensor bandwidth, sensor cross-talk, and sensor linearity. All these sensor array parameters are usually calibrated at least once before the measurement.
Optically-pumped magnetometers can be arranged into flexible arrays, which results in the need to determine all sensor positions and orientations every time the array configuration is changed. In addition, the orientation of the sensing axes of each magnetometer in the array is affected by cross-talk from neighboring sensors. Furthermore, parameters such as the gain, the bandwidth, and the linearity could vary with changes in the light power or background magnetic fields. Due to limited bandwidths of optically-pumped magnetometers, the gain also has a frequency dependence within the frequency range of interest, which can change as laser or vapor cell parameters change. These parameters therefore require frequent calibration, in order to create high-resolution images.
In prior art, Kim et al. (K. Kim et al, NeuroImage 89, 143 (2014)) have calibrated the position and orientation of an OPM-based sensor array by applying a set of calibrated linear magnetic field gradients to the array prior to use. The magnetometers were then utilized to determine the magnetic field at the location of the atoms. This allowed deduction of the exact “effective” sensor positions and orientations at the time of calibration but not throughout the duration of the measurement of the target fields or data collection.
Magnetoencephalography (MEG) uses large imaging arrays, of often hundreds of magnetometers, to measure magnetic fields produced by head and brain tissue. These magnetometers were traditionally superconducting quantum interference devices, but recently OPMs have also been employed. Several methods to calibrate the magnetometer positions in the imaging array have been developed. Most of them use a set of dipolar sources in the form of coils whose relative positions and orientations are precisely known prior to data collection. The magnetic field distributions are measured with the magnetometer array and the values are compared to theoretical models. This allowed estimation of the magnetometer positions and orientation with respect to each other and the source array (A. Bruno and P. Costa Ribeiro, Rev. Sci. Instrum., Vol. 62, 1005, 1991; R. Kraus Jr. et al., Biomedizinische Technik 46, 38, 2001; A. Pasquarelli et al., Neurology and Clinical Neurophysiology, 94, 1, 2004; Y. Adachi et al., IEEE Trans. Mag. 50, 5001304, 2014; V. Vivaldi, Biomag 2014, Aug. 24-28, Halifax, Canada). Further, Chella et al. (Chella et al., Phys. Med. Biol. 57, 4855, 2012) used a method to compensate for external interference and sensor artifacts to determine the magnetometer positions.
In several previous MEG applications, OPM positions and orientations have been determined geometrically before or after the measurement. In Boto et al. (E. Boto et al., PLOS ONE 11, e0157655, 2016), a snug-fitting printed headcast was used to tightly constrain the outer dimensions of the sensors and the effective sensor position was calculated from that. In O. Alem et al. (O. Alem et al., Optics Express, 25, 7849 (2017)), a printed helmet allowed for sensor movement in the radial direction only and the radial position was recorded after every measurement. The sensor positions were then inferred from the geometric geometry.
All of these methods calibrate the array anywhere from one to several times and do not accommodate variations of the parameters of the imaging array during the measurement of the target magnetic field and/or data collection. The system and method described herein is broadly applicable to imaging systems with one or more sensors, such as magnetometers, positioned in different locations. The system and method described herein may also be particularly useful in situations in which the exact magnetometer locations or other parameters of the imaging array vary over the course of the measurement. These parameters may include gain, bandwidth, orientation, cross-talk, or linearity.
Briefly describing the invention, the system and method includes at least one or multiples of non-target magnetic field producing sources, generating well-defined magnetic field distributions that vary over the spatial area covered by the at least one, but potentially hundreds of magnetometers of an imaging array. The positions and orientations of the non-target sources with respect to each other are known before data collection begins. One, or several parameters of the imaging array can be calibrated by simultaneously measuring the non-target and target magnetic fields. In other words, once data collection begins, signals are applied to the non-target sources in such a way that the magnetometers can identify the source it originates from. This is accomplished by, but may not be limited to, the signals from non-target sources being sinusoidal modulations at defined frequencies, where every source has its own frequency, or dipole frequencies or by temporally switching the sources on and off in a deterministic temporal pattern so that only one source operates at a time. The magnetometers are then measuring the strength of the signals of the non-target magnetic field as well as the target magnetic field. The strength of the non-target magnetic field is used to deduce the array parameter of interest and calibrate this parameter periodically.
In the invention, at least one non-target magnetic source, which produces a well-defined magnetic field pattern is operated. A magnetic imaging array, consisting of at least one magnetometer, is used to measure both the target and the non-target source. The information obtained from the measurements of the non-target magnetic sources is used to obtain information about the imaging array itself. This information is used to calibrate at least one parameter of the magnetic imaging array. Such parameters include, but are not limited to, magnetometer position, magnetometer orientation, magnetometer gain, linearity, and cross-talk between magnetometers.
The invention is a system for continuously calibrating a magnetic imaging array, the system comprising: at least one non-target magnetic source capable of creating a known magnetic field pattern; and an imaging array comprising at least one magnetometer, wherein the magnetometer is capable of simultaneously measuring the magnetic fields of the at least one non-target magnetic source and a target magnetic source; and a device that uses the magnetic field measurement from the at least one non-target magnetic source to generate at least one calibration parameter of the imaging array. A further embodiment is a method for continuously calibrating a magnetic imaging array, the method comprising the steps of: using at least one non-target magnetic source to create a known magnetic field pattern; measuring the known magnetic field pattern along with a target magnetic field to create a magnetic field measurement of the known field along with the target magnetic field measurement; and using the magnetic field measurement of the known field to produce a calibration parameter of the imaging array.
Before explaining the disclosed embodiments of the present invention in detail, it is to be understood that the invention is not limited in its application to the details of the particular arrangement shown, since the invention is capable of other embodiments. Also, the terminology used herein is for the purpose of description and not of limitation.
In order to distinguish between the non-target and target sources and not limit the measurements, the sources may generate fields within a narrow frequency band, where each non-target source could have its own frequency band outside the target measurement band of interest (frequency multiplexing). Alternatively, all sources could use the same frequency band and the sources are emitting successively, where only one non-target source is emitting at any given time (time multiplexing).
The non-target and target magnetic field information sensed by each of the magnetometers 3 is measured simultaneously and the target magnetic field information can be used to continuously or periodically calibrate parameters of the imaging array. These parameters include, but are not limited to magnetometer positions, orientations, cross-talk, gain, linearity, and bandwidth.
As an example, a simple magnetic imaging array has been constructed out of three optically-pumped magnetometers (OPMs) 50 as shown in
The above example described how the present invention was used to continuously calibrate the positions, orientations and the overall gain of the imaging array. In the current example the bandwidth and related frequency dependence of the gain were continuously calibrated. Since in most OPMs, the bandwidth depends on laser parameters as well as DC background fields, it is prone to drift and requires frequent recalibration. In this example, a current dipole was continuously driven with the sum of several sinusoidal modulations at 100 Hz, 200 Hz, 300 Hz and 400 Hz of the same amplitude. The magnetic field was recorded continuously. With a bandwidth of the OPM around 150 Hz, the peaks in the power spectrum corresponding to these modulation fields were clearly seen to decrease with higher frequency. Notch filters were applied around the modulation peaks to the time series in order to minimize the effect of the modulation on the data. The amplitudes of the modulation fields at the different frequencies were then used to calculate correction factors that take the frequency-dependent gain and related bandwidth and phase shifts into account.
Although the present invention has been described with reference to the disclosed embodiments, numerous modifications and variations can be made and still the result will come within the scope of the invention. No limitation with respect to the specific embodiments disclosed herein is intended or should be inferred. Each apparatus embodiment described herein has numerous equivalents.
The following application is an application for patent under 35 USC 111 (a). This invention was made with government support under HD074495 awarded by the National Institutes of Health. The government has certain rights in the invention.
