Patent No. 5526811 Apparatus and process for determining the sources of biomagnetic activity

 

Patent No. 5526811

Apparatus and process for determining the sources of biomagnetic activity (Lypchuk, Jun 18, 1996)

Abstract

An apparatus and process for making biomagnetic measurements of a biological organism permits the internal sources of the activity to be identified. An array of dipole sources is identified by providing a plurality of biomagnetic sensors disposed at locations external to the biological organism, measuring a measured biomagnetic response at each of the sensors, and amplifying and filtering the measured biomagnetic response. A solution of dipole sources within the biological organism is determined by forward calculating a computed biomagnetic response at each of the sensors resulting from the biomagnetic activity of a plurality of dipole sources, each of which dipole sources contributes a normalized total signal strength at the sensors, and solving for the strengths of each of the dipole sources by a minimum norm estimation procedure. Convergence on the solution is aided by iteratively removing from the determination those apparent sources that contribute only a small portion of the signal strength, and then resolving the resulting relationship.

Notes:

BACKGROUND OF THE INVENTION

This invention relates to measuring magnetic fields produced by living organisms, and, more particularly, to determining the sources of such biomagnetic activity from external measurements.

Living organisms produce magnetic fields that can be measured noninvasively with sensors positioned outside of the organism. These magnetic fields arise from electrical activity within the organism. Measurement of the magnetic fields can lead to an understanding of the electrical activity. For example, the measurement of magnetic fields produced by the brain can lead to an understanding of the mechanisms of perception and sensory response, as well as to the normal functioning of the brain and conditions that lead to illnesses and abnormalities.

Biomagnetometers are devices that measure the small magnetic fields produced by a living organism. The biomagnetometer typically includes a number of sensors arranged in an array external to the organism, which measure the magnetic field at a number of locations. Each sensor typically has a magnetic field pickup coil that may be a magnetometer or a gradiometer. When a small magnetic flux change penetrates the pickup coil, a small electrical current flows in the coil. This small current is detected by a sensitive detector of electrical currents, preferably a Superconducting Quantum Interference Device, known by the acronym "SQUID". The output of the various SQUIDs, after amplifying, filtering, and signal conditioning, is provided to a computer which stores and analyzes the data. The SQUIDs operate only at superconducting temperatures, and to attain the best system performance the pickup coils and SQUIDs are usually placed into a cryogenically cooled dewar. Because the biomagnetic fields produced by the body are so small compared to the magnetic fields of the earth and many types of electrical apparatus, it is common to place the subject of the biomagnetic study into a magnetically shielded room that excludes external magnetic and radio frequency fields.

The course of development of biomagnetometers has led to ever-increasing numbers of sensors (pickup coils and SQUIDs) in each unit. As of this writing, biomagnetometers with 37 sensors are available commercially, and even larger numbers of sensors are likely in future units. One impetus to increasing the number of sensors is that the larger amount of information available from the array of sensors offers the promise of magnetically imaging the source region during operation of the magnetic fields. That is, with a sufficient amount of information it becomes possible to form pictures or maps of activity in organs such as the human brain and heart.

The fundamental problem in analyzing sources of magnetic activity in the brain within the context of operation of a biomagnetometer is that the number of potential sources is greater than the number of sensors. Exact solutions are therefore not possible. Additionally, the magnetic field produced within the organism passes through several media (e.g., tissue, bone, air) before being measured, thereby complicating the analysis. In order to produce optimal solutions within these constraints, various analytical techniques to solving this "inverse problem of biomagnetism" have been utilized.

One common analytical approach has been to assume that the externally measured magnetic field is produced by a single dipole source or some small number of dipole sources, and to calculate the location, orientation, and strength of the source(s) from the external measurements. This approach is questionable because of its oversimplification of the physiology of the organism, but has been useful in a number of contexts. More rigorous approaches involve mathematical analyses of the data gathered by the sensors, such as the lead field synthesis technique described in U.S. Pat. No. 4,977,896. The minimum norm estimation technique also shows promise, because it provides a solution optimized according to physical principles. However, it may not yield fully acceptable results in some circumstances.

There is therefore a need for an improved biomagnetometer and approach to making biomagnetic measurements wherein the magnetic field sources within the subject organism are imaged based upon the external measurements of magnetic field. The present invention fulfills this need, and further provides related advantages.

SUMMARY OF THE INVENTION

The present invention provides an apparatus and process for making external, noninvasive biomagnetic measurements of a biological organism, and relating those measurements to the internal magnetic field sources of the subject that produce the measured results. The technique can utilize the numbers of sensors on available apparatus, and can also be adapted for use with even larger arrays of sensors as they become available. The approach of the invention is particularly useful where the biomagnetic sources lie at varying depths within the body.

In accordance with the invention, a process for making biomagnetic measurements of a biological organism comprises the steps of providing a plurality of biomagnetic sensors disposed at locations external to the biological organism, measuring a measured biomagnetic response signal at each of the sensors, and amplifying and filtering the measured biomagnetic response signals. In array of dipole sources within the biological organism that produces the externally measured biomagnetic response is determined by forward calculating a computed biomagnetic response at each of the sensors resulting from the biomagnetic activity of a plurality of dipole sources, each of which dipole sources contributes a normalized total signal strength at the sensors, and solving for the biomagnetic signal strengths of each of the dipole sources by a minimum norm estimation procedure using the computed biomagnetic response and the amplified and filtered biomagnetic response signal.

Further in accordance with the invention, an apparatus for making biomagnetic measurements of a biological organism comprises a plurality of biomagnetic sensors, means for measuring a measured biomagnetic response of a biological organism by each of the sensors, and means for amplifying and filtering the measured biomagnetic response. There is, additionally, means for determining an array of dipole sources within the biological organism by forward calculating a computed biomagnetic response at each of the sensors resulting from the biomagnetic activity of a plurality of dipole sources, each of which dipole sources contributes a normalized total signal strength at the sensors, and solving for the strengths of each of the dipole sources by a minimum norm estimation procedure utilizing the computed biomagnetic response and the amplified and filtered measured biomagnetic response.

In the preferred implementation, the forward calculation is performed by choosing and fixing an array of candidate dipole sources within the biological organism as the linear relationship

In this relationship, y.sub.j is the contribution of the jth dipole source to the total biomagnetic signal; b.sub.i is the measured biomagnetic response of the ith sensor; A.sub.ij =f.sub.i (r.sub.j,q.sub.j), f.sub.i is the effect on the ith sensor of the jth dipole source; r.sub.j is the source position of the jth dipole source; and q.sub.j is the dipole moment of the jth dipole source. A key feature is that

for all of the dipole sources. Here,

where n is the number of sensors. This solution, y.sub.j, indicates the relative contributions of the biomagnetic signals of the dipole sources.

These relationships are utilized to obtain the minimum norm solution for the y.sub.j values. Once these values are known the dipole strengths are determined through the relation

Here, as above

The strength of the jth dipole is obtained by multiplication as

When this process is implemented, it is often found that many of the dipole sources appear to make minimal contribution to the measured magnetic field. Such dipole sources producing low signal strength are likely either physically present but related to noise or processes not of interest, or are artifacts of the process. To improve the precision of the analysis so that attention may be focused on the dipole sources of physiological interest as related to particular processes within the subject, an iterative approach may be followed. In a second round of the determination process, those dipoles which contribute less than a preselected amount to the signal strength are removed from the determination, while the other dipoles determined in the substep of solving are retained. Then the substep of solving for the signal strengths of the remaining dipole sources by a minimum norm estimation procedure is repeated, on the smaller universe of dipole sources under evaluation. After several rounds of calculation, it is usually observed that the determination converges to a set of dipole sources. Simulations under controlled conditions have demonstrated that this final set of dipole sources resulting from the process closely approximates the actual set of dipole sources that produce the externally measured magnetic fields. It is particularly significant that the approach does not favor near-surface sources over deep sources.

The present approach is an important advance in the art of biomagnetometers, and of making biomagnetic measurements and imaging the biomagnetic sources within a subject. In many instances, the biomagnetometer has limited usefulness when the readouts of the sensors are provided as raw data. The analysis of the present approach, when used in conjunction with the biomagnetometer sensors and their information, permits the biomagnetometer to "image" magnetic sources in the body. Other features and advantages of the invention will be apparent from the following more detailed description of the preferred embodiments, taken in conjunction with the accompanying drawings, which illustrate, by way of example, the principles of the invention.

-------------------------------------------------

The present approach provides an apparatus and process for determining the location and strength of magnetic field sources within living subjects, without bias toward shallow sources. Although particular embodiments of the invention have been described in detail for purposes of illustration, various modifications may be made without departing from the spirit and scope of the invention. Accordingly, the invention is not to be limited except as by the appended claims.