Interpage Links
Research
Postdoctoral Research
Stroke survivors are more impaired further down their limbs, often suffering from the inability to extend the fingers. While the proximate reasons for impaired finger extension is currently under debate, the ulimate reason is due to the stroke. The goal of this project is to develop a novel rehabilitation intervention at the source of the problem, the brain. We intend on developing a method of targeted plasticity in the brain and correlate these changes with motor function performance of the hand.
Ph.D. Research
Many stroke survivors suffer from disabilities, including gait. One common disability is known as stiff-legged gait, or stiff-knee gait. The lack of sufficient knee flexion forces the person to compensate by swinging (circumducting) the affected leg in order to clear the foot. While this compensation allows functionality, stiff-legged gait is energy-inefficient, slow and aesthetically unpleasing. A number of studies have examined stiff-legged gait and have come up with many reasons why this behavior occurs. My research involves testing some of these hypotheses using a lightweight, active knee orthosis.
Publications:
Masters Research
Although stroke survivors can improve functionality in their chronic stage (> 6 months post-stroke), insurance typically covers therapy from 2 to 6 months. While survivors can train by themselves at home, the patient lacks the interaction and motivation provided by a physical therapist. Robotic therapy has been introduced in the past decade as a possible low-cost solution for home rehabilitation.
Starting in 2004, I developed a novel method of joint actuation using cables. With the intended application of upper-extremity robotic rehabilitation at home, this method is potentially more lightweight, inexpensive and safer than other methods available. The concept, known as MARIONET (Moment arm Adjustment for Remote Induction Of Net Effective Torque) uses both moment arm manipulation and cable tension to produce torque. I developed and tested a single-joint proof-of-concept for the elbow, finding it suitable for further development into a multi-joint version.
MARIONET Theory:
Publications:
- Sulzer, J S, M A Peshkin, J L Patton. Pulling Your Strings: Cable Moment Arm Manipulation as a Modality of Joint Actuation. IEEE Robotics and Automation Magazine vol. 15, no. 3. pp. 70-78.
- James S. Sulzer. "Moment Arm Manipulation of a Cable-Driven Joint". (MS thesis, Northwestern University, 2006).
- Sulzer J, M A Peshkin and J L Patton. Catastrophe and Stability Analysis of a Cable-driven Joint. IEEE International Conference of Engineering in Medicine and Biology Society. New York, NY. August, 2006.
- Sulzer JS, MA Peshkin, JL Patton. MARIONET: An exotendon-driven Series Elastic Actuator for exerting joint torque. Proc. Int. Conf. on Rehabilitation Robotics (ICORR). Chicago, IL June 2005.
Other Research
When we explore our environment, we integrate sensory information to obtain the full experience. For instance, when running a hand across a pillow, we use both visual information (from the eyes) and haptic information (from proprioceptors in the arm and fingers) to obtain information about the softness of the pillow. But how much of that experience is due to vision? How much is due to haptic feedback?
Using a position-dependent, curved virtual surface rendered on a two degrees-of-freedom planar manipulandum, I asked users to make multiple reaches from one point to another 10 cm away. I varied both haptic stiffness (stiffness the user felt) and visual stiffness (stiffness the user could see) and observed the resulting trajectories. From this short pilot experiment, I found that vision plays a strong role in haptic feedback and that humans may be more sensitive to making a surface look more compliant than stiffer. However, a larger scale experiment needs to be conducted before developing a quantitative model.
Publications:
Undergraduate Research
Testing for possible head trauma in a given situation is a time consuming, costly procedure. Some experiments require a dummy head (headform), others even require a cadaver skull. It would be very advantageous if this process could be simulated on a computer instead. My undergraduate thesis involved developing a model using a program typically for simulating car crashes known as MADYMO (R), and applying it to head trauma. I found that the multibody collsion model employed in MADYMO was too simple to effectively model the viscoelastic headform.
Publications: