Title: Magnetomicrometry: Tissue Length Tracking via Implanted Magnetic Beads
Target tracking is an important problem relevant to a wide range of disciplines and across a wide range of scales. In the medical field, information on tissue location is important for precision surgery, endoscopy, heart valve monitoring, tumor monitoring, muscle and joint tracking, and many other applications. In scientific investigation, target tracking is used in applications such as fluid dynamics, geological surveying, beam bending measurements, and animal population tracking. Target tracking is also becoming increasingly relevant in virtual and augmented reality applications and other human-computer interactions. As a consequence of these widespread applications, advances in target tracking drive cascades of new medical, scientific, and technological capabilities. Particularly, this dissertation advances magnetomicrometry, a technology that tracks visually-obscured magnetic beads implanted within biological tissue to monitor in-vivo tissue length and speed within freely moving animals and humans.
There are many methods for the tracking of visually-obscured objects, but magnetic target tracking has the advantages of being low-cost, compact, portable, passive, and safe. However, current magnet tracking technologies suffer from high latencies, preventing high-speed target tracking from becoming a reality. Determining magnet states (i.e. locations, orientations, and strengths) from an array of magnetic field sensors is not guaranteed to have a closed form solution. Thus, the state of a magnetic target is commonly determined using optimization techniques, which often suffer from large delay or from convergence to local minima.
This dissertation develops the mathematics for an improved method to track one or more magnets at high speed and accuracy. The method is validated with a demonstration of real-time muscle length tracking. We first develop a method to track multiple magnets at high speed in real time using the analytic gradient of the magnetic field prediction error. We then extend this method to compensate for magnetic disturbance fields in real time in a manner that is more portable and less complex than currently published magnetic disturbance field compensation methodologies. We validate our method in a physical system against state-of-the-art motion capture, and we demonstrate increased maximum bandwidths of 336%, 525%, 635%, and 773% for the simultaneous tracking of 1, 2, 3, and 4 magnets, respectively, with tracking accuracy comparable to state-of-the-art magnet tracking techniques.
Using pairs of implanted magnetic beads to wirelessly track the length and speed of muscle, we apply a mechanical frequency sweep to an in-vivo turkey gastrocnemius muscle and compare magnetic-bead-derived real-time muscle length measurements against stereo X-ray videofluoroscopy. We further collect longitudinal data from computational tomography scans to investigate minimum magnetic-bead separation distances in muscle. We find submillimeter agreement between magnetomicrometry and stereo X-ray videofluoroscopy, and a minimum magnetic-bead separation distance of approximately two centimeters.
It is our resolve that magnetomicrometry will lay the groundwork for peripheral nervous system control of wearable robots via real-time tracking of muscle lengths and speeds, as well as for the in vivo tracking of biological tissues to elucidate biomechanical principles of animal and human movement.
Hugh M. Herr, Ph.D. (Professor of Media Arts and Sciences, MIT Media Lab)
Thomas J. Roberts, Ph.D. (Associate Professor, Ecology and Evolutionary Biology, Brown University)
Joseph A. Paradiso, Ph.D. (Professor of Media Arts and Sciences, MIT Media Lab)