You are viewing the Ward Rounds Online Archive. To view the current issue, click here.


Mr. Roboto

By Devon McPhee


If you were a study participant in the lab of Jules Dewald, PT, PhD, chair of the Department of Physical Therapy and Human Movement Sciences (NUPTHMS), you might find yourself with an arm floating in virtual anti-gravity, or pushing through simulated viscose matter, or repeating an exercise that replicates the action of lifting a gallon of milk, all while attached to a high-definition electroencephalograph (EEG) recorder. By studying these movements and mapping the accompanying brain patterns, Dewald and his team hope to unlock the basic science behind movement disorders — specifically those found in stroke and cerebral palsy patients — and develop more targeted therapies to treat them. Reaching this goal, Dewald says, requires an innovative use of robotics in human movement science.

The ACT-2D measures the effect of abnormal shoulder/elbow joint torque coupling on the expression of spasticity in stroke survivors.

“Robotics has been used in research since the early ’90s, but we are among the first to use it to study the science behind movement disorders,” explains Dewald, associate professor of physical therapy and human movement sciences, biomedical engineering, and physical medicine and rehabilitation at Northwestern University. “Once we understand the science, we can use these robots for therapy.”

Dewald’s science-based approach to developing therapies for movement disorders diverges from the way rehabilitative treatments have traditionally been developed.

“The way physical therapy currently works is through qualitative evaluation — trial and error,” he says. “Usually, therapies are developed before we understand the underlying mechanisms of the movement disorder. Our approach is to first understand the mechanisms, and then develop therapies based on that understanding.”

Getting from point A (figuring out the science) to point B (more targeted therapies) requires a hefty investment of time, since researchers must break a movement into distinct pieces and understand each one to get at the whole.

“It’s literally like taking Humpty Dumpty apart and putting him back together again,” he explains.

The slow pace of understanding the mechanisms underlying movement impairment following brain injury is likely one of the main reasons this approach to rehabilitation science remains novel, Dewald says, but he has faith that once clinicians see the results from science-based therapies, they will become converts.

“The best way to convince rehabilitation clinicians about the merits of reductionist scientific inquiry is to demonstrate the results you can obtain,” he says.

Jules Dewald, PT, PhD, chairs the Department of Physical Therapy and Human Movement Sciences at Northwestern University.

A team of researchers aids Dewald in his pursuits, including two assistant research professors of physical therapy and human movement sciences, Ana Maria Acosta, PhD, and Jun Yao, PhD; an instructor in physical therapy and human movement sciences, Michael Ellis, MPT, DPT; a research associate in biomechatronical engineering from the Netherlands, Arno Stienen, PhD; five DPT/PhD in biomedical engineering students, Theresa Sukal Moulton, DPT, Christa Nelson, Laura Miller, Rachel Hawe, and Lindsay Garmirian; three PhD students, Daniel Krainak, MS, Jacob McPherson, MS, and Natalia Sanchez; and three research physical therapists, Carolina Carmona, DPT, Donna Hurley, DPT, and Justin Drogos, DPT.

Taking a peek inside the Dewald lab, one could easily draw comparisons to a strength training room at the gym — except the machines here build knowledge instead of muscle. Three large machines have customized robotic arms that measure movement or create “haptic” or virtual environments for study subjects to operate in, and serve as important tools for the research being conducted.

In May, Dewald added a fourth robot to his collection, which he calls the “Rolls Royce” of upper extremity robotics. The new device, developed and built with a quarter-million dollar grant from the National Institutes of Health (NIH), will allow for a much larger work space on which to conduct studies and has the added ability to perturb the arm — mimicking the action of an arm getting pushed out of the way or the jerking of a steering wheel — to add to the complexity and real-world application of the lab’s studies. The device also gives participants the ability not only to move their arms forward but also up and down.

A Sequence of Studies

Laterality Index values for both control (pink) and stroke (green) groups across movement tasks. Positive values indicate more contralateral hemisphere activity, while negative values indicate more ipsilateral hemisphere activity.

For the past 20 years, Dewald has studied one of the most stereotypical movement disorders following a stroke, abnormal movement coordination of the shoulder and elbow. For a stroke patient with this disability, everyday tasks such as putting on a coat or turning a car radio dial become difficult because the more she lifts the affected arm at the shoulder, the more her elbow flexes.

To begin decoding this impairment, Dewald developed two studies that quantified the mechanics of the shoulder-elbow system in stroke patients.

One study asked participants to reach for a target, first with the arm supported by an air-bearing device, then without support. Dewald and his colleagues at the Rehabilitation Institute of Chicago found that individuals with severe limitations could reach the target fairly accurately when supported but not when unsupported. This occurs, Dewald says, because a support provides a surface for the arm to push down on, eliminating the need for the shoulder to lift up and allowing the elbow to extend as normal.

The second study involved evaluating the isometric movement of the shoulder and elbow using a load cell, a mechanical device that measures the forces acting on it. Participants’ forearms were placed on a metal plate attached to the load cell. Hands, wrists, and half of their forearms were set in a cast to restrict movement and to hold the arm in a set position. The subjects then lifted up their arm or flexed their elbow, and the load cell measured what was happening. At the same time, patients were attached to a high-resolution EEG device (using 160 electrodes), recording which parts of the brain were activated.

Data from the second study helped Dewald determine two things about stroke patients. First, coupled with information from the first study, it presented a quantifiable definition of how the shoulder and elbow worked together. Second, by pinpointing the areas in the brain that were activated during shoulder and elbow activity, Dewald and his team could then compare the patterns of stroke and non-stroke subjects. They found that non-stroke patients had more succinct areas of activation, while stroke patients had more overlap. In other words, stroke patients were asking their brains to do the same job with fewer neurons.

Developing a Treatment

Therapeutic benefits of the research were observed in a small study of seven individuals with chronic moderate to severe stroke. First, the researchers measured the work area (ability to extend in a circular motion) of participants using a range of limb weights generated by the ACT-3D haptic robot. They found that as the limb became heavier, the participant’s work space became smaller. Next, subjects conducted reaching exercises three times a week over the course of eight weeks. Throughout the study, if a subject was able to consistently reach at least 90 percent toward the target, the weight of the limb was increased by 25 percent.

After eight weeks, each participant’s work area was once again measured. All patients saw, at minimum, a 20 percent increase. Dewald attributes this improvement to the plasticity of the brain: it had learned how to use its remaining neural pathways more efficiently, as opposed to creating new ones in that short time span.

“It’s like when the express lanes on the highway open up,” Dewald explains. “No new road is built, but you make better use of what exists.”

Brain images taken during the study bolster Dewald’s hypothesis. Those taken at the beginning of the session showed that participants used both ipsilateral and contralateral sides of the brain as they reached with the affected limb. After eight weeks, activity had shifted primarily to the contralateral side which, in non-stroke individuals, controls the entire movement.

Current Projects

With the basic mechanics laid out, the team has now begun the process of unraveling how a stroke patient’s shoulder, elbow, and brain operate in real-world situations. Combining robotics with virtual reality, a new study explores reaching movements in a variety of environments, including across a hard surface, through sticky matter, and in anti-gravity.

Going one body part further with stroke research, the lab has also started looking at hand disabilities. The complexity of the hand makes it one of the most highly affected areas post-stroke, Dewald says, because a big part of the cortical loss occurs to parts of the brain that control it, which also makes it one of the most difficult systems to decode. Current work has shown promise in the use of neuromuscular electrical stimulation of wrist and hand muscles to help subjects regain the ability to open and close their hand, as well as to grasp objects.

The lab began its initial work on movement disorders in children with cerebral palsy about five years ago. That effort has accelerated over the past two years thanks to an R01 grant from the NIH. These funds will support a study that hopes to determine how the muscles in children with cerebral palsy work together and a second study that will challenge the mobility of subjects with reaching exercises.

Finally, an extension of the lab’s international reach is set to occur this year when research associate Arno Stienen, PhD, returns to the Netherlands and the University of Twente. Stienen will retain an adjunct assistant professorship at Northwestern, and facilitate a fluid collaboration between the two schools.

“Under our collaboration, we will be taking the [robotics] wish list from this department and seeing what we think is possible to build in the Netherlands,” Arno explains.

Research and Technology

Dewald attributes his lab’s success to the strong support he has received across the university, saying that Feinberg’s focus on systems research and close collaboration with the Robert R. McCormick School of Engineering and Applied Science was what initially drew him to the school in 1988. Today, as chair of NUPTHMS, Dewald has made it a point to incorporate research and general engineering knowledge into the DPT curriculum.

The department launched a new curriculum in 2009 that includes a technology-related course during the final trimester of the third year, and all DPT students spend at least a year and a half conducting research. Additionally, instead of separating clinical and basic science courses, the new curriculum gives students exposure to the two in parallel, making the relevance of basic scientific knowledge to clinical cases immediately apparent and allowing DPT students to better retain this information for their future careers, Dewald says.

Two new degree offerings (both launched during the 2009-10 academic year) also focus on the combination of movement science and engineering or research. Students in the dual-degree DPT/PhD in engineering program enroll in one of three engineering tracks — biomedical, mechanical, or electrical engineering and computer science — as well as in the DPT program. These students take DPT courses with those in the traditional program, further exposing all students to engineering and science principles.

The existing PhD in neuroscience now offers students the option to specialize in movement and rehabilitation science. The NUIN-MRS program prepares graduates for work in neurobiology-related rehabilitation research, specifically in the study of the brain and the neurobiological mechanisms underlying movement disorders. The degree will generate systems neuro-scientists with strong backgrounds in technology and data analysis for the rehabilitation sciences, Dewald says.

Overall, the department chair says he still expects the majority of NUPTHMS graduates to become clinicians, adding that he believes Feinberg DPT students’ early exposure to research and technology during medical school — something that few other institutions offer — will advance the treatment of movement disorders.

“Our students will open doors and be our ambassadors to get newly developed, device-based therapies into clinical practice,” he says.

Online Extra

The Dewald Lab and the Cerebral Palsy Registry Help Facilitate Research

Unlike Australia, Sweden or Denmark, the United States does not have a national cerebral palsy (CP) registry to track the prevalence of the condition in the country and to provide its scientists with a central resource for potential research study subjects. In an effort to fill this gap, Northwestern University Feinberg School of Medicine partnered with the Rehabilitation Institute of Chicago and the University of Chicago to create the Cerebral Palsy Research Registry (CPRR).

Run primarily through the Department of Physical Therapy and Human Movement Sciences (NUPTHMS), and specifically the lab of Jules Dewald, DPT, PhD, the CPRR collects the basic medical history, clinical measurements, and contact information of individuals with cerebral palsy who are interested in participating in studies. Now four years old, the registry lists 325 participants from 20 different states who range in age from one to 47 (though anyone up to 90 years old can join). Researchers from any institution who have an approved study can request a search of the database for appropriate subjects and then have those individuals notified about the study. Those not affiliated with the CPRR must pay a $10-per-subject access fee. To date, seven studies have recruited using the registry.

Theresa Sukal Moulton, DPT, a student in Feinberg’s DPT/PhD Program, has been a part of the CPRR since its inception. She helped develop and obtain funding for the registry, is involved with its maintenance, and has even used it to recruit participants for her own studies of hemiplegia, a specific type of cerebral palsy that manifests itself as motor impairment on one side of the body. The registry has proved invaluable to her research, Moulton says, as patient recruitment remains one of her most difficult challenges.

“Our experiments are four hours long, and I am working with children who for the most part are in mainstream schools where they are involved with a lot of activities,” she says. “We’re constantly dealing with time and location constraints. We need subjects we can rely on and the individuals in the registry understand the commitment required.”

The registry was made possible through the Staubitz Charitable Trust, along with private contributions from Art and Linda Staubitz and their daughter, Melissa Siebert. The family donates to the CPRR because Siebert’s daughter, Caroline, has cerebral palsy.

“Following my daughter’s diagnosis, I was surprised at how little federally funded research was going on in regards to cerebral palsy’s causes, potential treatments, and a cure,” Siebert said. “The registry gives researchers access to data for private studies and, hopefully, will help increase federal funding for cerebral palsy research.”

To supplement the seed funding provided by the Staubitz Trust, which will likely run out by the end of next year, the CPRR team is currently soliciting support from foundations and private donations, as well as grants and federal funds, to allow the registry to continue to operate.

While the focus of the CPRR remains facilitating research, its mission has expanded to include data collection in an effort to understand the longitudinal impact of CP and possibly the incidence and prevalence of individuals with cerebral palsy in the United States.

“Eventually, we hope this information will help the cerebral palsy community lobby for more funding for research, and result in more resources for individuals with CP,” Moulton says.

To help build the database, the lab — primarily CPRR coordinator Donna Hurley, PT, DPT — has begun actively recruiting at rehabilitation centers, parent groups, and outpatient clinics around the Chicagoland area, as well as promoting the CPRR website. It has also started discussions with other universities that have expressed interest in joining the project.

Still, it will take many years for the CPRR to reach statistically significant numbers, Moulton says, as the database will need to triple in size to achieve a valid population sample for Chicagoland alone.

“We know that our goal [of understanding the CP population in the United States] is a long way off,” Moulton says. “But we hope that by starting small and collaborating with other institutions around the country, we can eventually gain the critical mass needed to begin answering some of those questions.”