In Penn State’s newest and most advanced research building, a new program is taking shape that, if successful, will revolutionize the ways in which we interact with the human brain. Led by Srinivas Tadigadapa, an electrical engineer, and Steve Schiff, a neurosurgeon with a background in physics and control engineering, this ambitious project exemplifies the convergence of research fields that are typically separated by distinct disciplinary boundaries.
In 2013, the Obama White House laid out a grand challenge to “accelerate the development and application of new technologies that will enable researchers to produce dynamic pictures of the brain that show how individual brain cells and complex neural circuits interact at the speed of thought.” Called the BRAIN Initiative, it is a 12-year plan to fund research into understanding the brain on multiple levels, using a variety of new and developing technologies. With these tools, it is hoped that the many diseases and malfunctions that afflict the brain can be controlled or eliminated. Schiff and Tadigadapa recently won one of Penn State’s two exploratory BRAIN awards.
A transdisciplinary team to solve a monumental problem
Steve Schiff has the soothing voice and gentle manner of someone who has spent a large part of his career dealing with children, and frequently, children in pain. As a pediatric neurosurgeon, he has leant his skills and bedside manner to treating diseases of the brain in children, but as a researcher he is adding another skill set, one based on his background in engineering and physics, to develop technology to understand and control diseases of the brain.
Schiff is director of the Penn State Center for Neural Engineering, a lab that takes up an entire floor of the Life Sciences wing of the Millennium Science Complex on Penn State’s University Park campus. A series of card-swipe controlled laboratories make up the 11,000-square-foot Center, with facilities for the construction of custom electronics, live animal imaging, surgery, and advanced computerized microscopy. His Center colleagues include medical doctors, engineers and biomedical engineers, and the graduate students they are training.
In the Materials wing of the building in a basement micro and nanoscale devices laboratory, Tadigadapa’s group is developing microelectromechanical systems (MEMS) that miniaturize device arrays for sensing and actuating, some of which the team hopes will one day be implanted into the human skull in order to explore the brain on a cell-by-cell basis.
Penn State’s Millennium Science Complex was built with the concept of integrating the expertise of materials scientists, electrical and mechanical engineers, and nanotechnologists, who occupy the north wing of the building, with medical and biological researchers, who occupy the building’s west wing.
“This building we are in reflects this interaction, because we are half materials science and half life sciences,” Schiff said. “We will literally build these technologies on one side and walk them up the stairs to our lab where we do experiments on neurons. We will use individual neurons that we will be recording from and stimulating to see how far we can push this technology. We are, to our knowledge, the only center at present that is in a position to manufacture these high density arrays for sensing and stimulation in a nanofabrication facility and then literally transition them to an operating room.”
The team also includes, as a consultant, John Wikswo of Vanderbilt University, who is one of the world’s leading experts on magnetic fields in neurons. “John provides some of the key physics expertise that no one else in the world has,” Schiff said.
In short, Schiff and Tadigadapa, with Wikswo’s help, propose to develop a technology capable of measuring the activity of individual cells of the brain and to stimulate those cells at room temperature with a MEMS device capable of being implanted above the inner table of the skull for long-term human use.
Why stimulate the brain?
For the past 70 to 80 years, scientists have been using electrodes on the surface of the brain to measure electrical currents, and, since the 1950s, to stimulate the brain. In recent times, an approach called deep brain stimulation (DBS) was developed as a means to treat the tremors associated with Parkinson’s disease. Now, DBS is being studied experimentally as a method to treat major depression, along with a variety of other ailments. In DBS, a pair of electrodes is implanted in the brain and a generator is surgically implanted in the patient’s chest wall. A pattern of electrical pulses is used to stimulate portions of the brain. The treatment seems to work for a proportion of patients with major depression, although the exact mechanism is still unknown. Another use of such sensing and stimulation is to run robotics for patients with disabilities. However, bleeding, stroke and infection are potential side effects.
“If I put electrodes into the brain, which I have done a great deal in my career, there is a measurable risk of hemorrhage and damage, and there is always a few percent risk of infection,” Schiff said. “If I’m studying a child for epilepsy, I need to take those things out by two weeks, definitely by three weeks, or I have to go in and free them from the scar that’s already formed.
“Imagine you are trying to run an artificial hand,” posited Schiff. “You want to pick up signals from the hand area of the cortex to give you the intention of the individual to move such a hand. We can only do that now by implanting arrays of electrodes into the hand area of the brain itself.”
Because the bone of the skull is a good insulator, electrical signals from neurons in the brain cannot easily pass through it into the outside world. So the skull has to be opened up to place electrodes on the surface or deeper into the brain. This can and often does damage brain tissue and can cause infection in the cerebral fluid, potentially leading to meningitis. Furthermore, electrodes tend to corrode or be scarred over within weeks to a few years, necessitating more surgery.
An engineer steps in
A better choice would be to sense and stimulate the neurons without penetrating the dura, the covering that protects the brain from infection. That’s where Tadigadapa steps in. He is exploring magnetic MEMS sensors and actuators that can record and stimulate the brain with non-contact techniques.
The Penn State team will attempt to stimulate single neurons with a magnetic field that creates an electrical current in the individual brain cell. Stimulus is actually easier than recording, according to Tadigadapa. He and his engineering team are creating tiny coils that can deliver a localized stimulation to cells. Eugene Freeman, a Ph.D. candidate in Tadigadapa’s lab, is working on the coil approach.
“Usually what people do to stimulate neurons is to have very big coils that go outside the head,” said Freeman. “They’re not implantable. They’re about the size of your fist, so you have to go in to a lab for the treatment, which is called transcranial magnetic stimulation (TMS). These large magnets activate a relatively large part of the brain. You can’t get single neuron specificity. We are experimenting and simulating microcoils in different sizes and shapes, the smallest so far being about 500 microns in diameter (about half the size of a grain of salt). We use microglass structures and pattern 3D copper coils on them.”
The ability to do three-dimensional fabrication is a recent development, Freeman said. “We are just looking to see if that might be a better way to do it than just a flat coil.”
Currently, the only magnetic sensors sensitive enough to detect the magnetic waves of the brain either need to be cooled to liquid helium levels in a device called a SQUID or heated to 180 degrees C to vaporize metals in a device called an atomic magnetometer. Neither technique is suitable for outpatient sensing and stimulation. So, recording of neural activity that is not potentially damaging to the brain either doesn’t have cellular resolution, requires shielded rooms, or cannot be performed at safe temperatures. A recently developed third method uses a synthetic diamond material with random nitrogen vacancies to pick up very small magnetic fields, but that requires using microwaves – like the ones in a kitchen microwave – which have a tendency to cook things in the vicinity. Not an optimal solution.
Blocking the Earth’s magnetic field
In order to make an implantable sensor to detect magnetic fields in the brain, something has to be done about Earth’s magnetic field. Tadigadapa proposes to build active and passive circuits on a CMOS microchip that will cancel out magnetic noise using a simple feedback loop, the same technique used in noise cancelling headphones. The microdevice will generate an on-board magnetic field within the MEMS chip that will compensate for the Earth’s magnetic field and other interfering magnetic fields in the nearby environment. Other circuits in the implantable chip will amplify the magnetic signal from the cells.
“The Earth’s magnetic field is around 60 microTesla,” Tadigadapa said, “and the magnetic field of the human brain is around a picoTesla (around 60 million times weaker). So there is need to block Earth’s huge magnetic field. Currently that is done within an isolated room that costs $10 million to build, plus the cost of the SQUID itself and the high cost of liquid helium to cool the device. We hope to be able to do it with an on-chip circuit.”
“We have a number of designs that Srinivas is going to be placing on these chips. Unlike in the past, these things can be implanted in the body. They are ambient, which means they assume the temperature of their surroundings, rather than freezing them by being dropped into liquid helium. And when implanted, they will allow us to take this technology out of the laboratory for the use of people in an ambulatory setting,” Schiff said.
“It’s a good technical challenge,” Tadigadapa remarked with patent understatement.
Who benefits from this technology?
“Many of our colleagues in surgery and neurology are very excited that we can do this. At Hershey (Penn State Milton S. Hershey Medical Center), we are working with patients in the ALS clinic. We’re keenly interested in improving our ability to let patients communicate as they lose their strength. We are also interested in their ability to control their environment robotically. That’s more of an issue of sensing,” Schiff said.
“If I have someone like one of our young veterans who has lost an arm, you want him or her to be able to run a prosthetic device for 50 years. You need to be able to maintain the device, not damage them any further (through repeated brain surgeries), and importantly, you need a way to give sensory feedback to the brain. This is a way to interact with the brain and to give it signals reflective of what a prosthetic is sensing as it touches something.
“I’ve worked in epilepsy for most of my career,” Schiff continued. “This is a way to potentially not only sense from a part of the brain that makes seizures, but to modulate the activity to prevent seizures. We’ve shown that we can stimulate parts of the brain to improve depression and Obsessive-compulsive disorder. One of the major thrusts of the BRAIN Initiative is to get a better understanding of the major cognitive disorders.”
The economic impact of very common cognitive disorders, such as depression, which affects 17 percent of the population of the U.S., is tremendous. The technology could also address the communication needs of people with spinal cord injuries, ALS or strokes, and of course, would benefit people who have lost limbs.
When will we have this technology?
Schiff and Tadigadapa hope to make rapid progress. Their recent work on magnetoflexoelastic resonators has demonstrated sensitivity in the tens of nanoTesla and on magnetoelectric magnetometers has achieved ~300 picoTesla sensitivity, all at room temperature. The resources of the Penn State Nanofabrication Laboratory will make it possible to make very thin, high-density arrays of sensors and stimulators. Initial testing will be conducted on rat brain slices in vitro to understand the strength of signals and the magnetic field strength required for stimulation. There are billions of neurons in the brain, not to mention many other types of cells that are also important. They will need to develop mathematical filters that can distinguish one cell from another. They will also learn whether it is necessary to use single cells or whether they can work with a few cells. A lot of basic science will result from the first couple of years of work.
“The first chips will probably be on the scale of a centimeter for the part of the chip that just does sensing,” Schiff said. “For implantation the scale we are targeting is 100 micrometers. We won’t be submillimeter for the array the first two years. This is a new technology grant where we’ve shown we have enough technical advances now that with this support we can go to the next stage of iterating electrical engineering and nanofabrication of devices with experiments.”
He expects that the first devices will be placed against the skull, rather than inside, due to the amount of electronics and chip packaging necessary. A wireless transmitter and receiver is also needed. When the electronics packaging issue is solved, Schiff expects to have a biocompatible array that can be placed in a small trough inside the skull where it can be maintained or replaced as needed.
“I think we are looking at a five-year time horizon to the point where we could seriously have the technology ready for application and potential translation testing,” Schiff said.
Andrew Whalen, a Ph.D. candidate in mechanical engineering, and post-doc Herve Kadji are working with Schiff to prepare brain slice experiments using the micro-magnetic coils that Eugene Freeman has designed to stimulate neuronal tissue. Through these experiments they hope to understand the fundamental mechanisms of magnetic stimulation on neuronal excitability.
“Magnetic stimulation has some nice advantages over traditional electric stimulation, which has many complications due to the electro-chemical interface between electrode and tissue,” Whalen explained. “Magnetic stimulation in theory should bypass these issues with electrical resistance interfering with stimulation transmission and the formation of scar tissue in long term stimulation implants.”