Neurorehabilitation

Current Projects

Haptic TheraDrive

TheraDrive is a new low-cost, high-force haptic robot with a single degree of freedom. Our lab is currently using this system for two purposes: first, to test the viability and usability of the new robot system; and second, to test if low-functioning patients benefit, and if so how much, from using force-feedback therapy as opposed to devices with zero impedance.  Exercises are performed by study subjects and an adaptive controller monitors patient performance to ensure that exercises are difficult but doable, which is important for maintaining patient motivation.  Currently, we hypothesize that not only will the new system be viable, but that it will provide better robot-assisted therapy to a large variety of patients, especially low-functioning stroke survivors with hemiplegia.

Acknowledgements

  • Grant: NIH R42HD104325. El Comite’ Technico y de Administracion del Fondo Mixto CONACYT-Goberno del Estado de Chihuahua CHIH 2009-CO2-127781
  • American Heart Association Grant #0635450Z
  • Advancing a Healthier Wisconsin Grant #5520015
  • MCW Research Affairs Committee Grant #3303017
  • Recupero Robotics, LLC
  • Therapists
  • Engineers
  • Trainers

Leveraging Robot-Based Haptic Dyads to Improve Community-Based Stroke Rehabilitation

Stroke and age-related disabilities affect millions of older adults worldwide, with the greatest burden falling on low- and middle-income countries. This project explores how affordable robotic therapy and haptic interaction—the use of touch, vibration, and force feedback—can support rehabilitation in community settings.

 

 

We are investigating how pairs of individuals, including healthy older adults and stroke survivors, learn motor skills when connected through a haptic link in multiplayer robot-based games. This approach not only enables implicit communication but may also increase motivation and social engagement during therapy.

 

Our work focuses on three goals:

  1. Studying motor learning – How do motor and cognitive impairments affect learning when practicing with a haptic partner?
  2. Modeling communication – Using computational tools to understand how people with impairments share information through sensory feedback.
  3. Designing adaptive robots – Building controllers that balance ability levels so partners can support each other effectively.

 

By answering fundamental questions about sensorimotor communication, this project aims to guide the design of the next generation of accessible, community-based rehabilitation robots.

 

Acknowledgements

  • Grants: This research is supported by supported by the Eunice Kennedy Shriver National Institute Of Child Health & Human Development of the National Institutes of Health under Award Numbers F31HD116597 and R42HD104325.

Leveraging CT-Derived Lesion Metrics to Stratify Motor and Cognitive Recovery After Stroke

The goal of this project is to enhance the precision and accessibility of post-stroke rehabilitation by developing imaging-based tools that can predict and classify motor and cognitive impairments — even in settings where MRI is unavailable. The project bridges neuroimaging, computational modeling, and robotic rehabilitation to establish CT as a viable, scalable alternative for brain lesion analysis. Stroke is a leading cause of disability worldwide, but MRI—the gold standard for lesion analysis—is often inaccessible in low- and middle-income countries. This project addresses this gap by evaluating how CT-based lesion features can be used to:

  • Characterize brain injury with comparable accuracy to MRI.
  • Predict recovery profiles for both motor and cognitive domains.
  • Integrate lesion-based metrics with robotic and behavioral assessments to improve clinical decision-making.

Acknowledgments

  • Grants: NIH R42HD104325 and NIH R42HD104325-03S1. NIH-funded T32 grant (Training in Musculoskeletal Research): T32AR007132-48.
  • The Penn Fontaine Society
  • Department of Bioengineering, University of Pennsylvania, School of Engineering and Applied Science
  • Department of Physical Medicine and Rehabilitation, University of Pennsylvania, School of Medicine

Neural and Motor Functional Changes in HIV and Stroke Before and After Robot-Assisted Neurorehabilitation

The purpose of this study is to understand the role that HIV has on the brain's function and provide insight into the relationship between HIV and stroke. We are interested in knowing how the presence of HIV affects robot-assisted rehabilitation after stroke and its association with clinical improvement. We will use fMRI and Theradrive to assess differences between patients with just stroke and patients with both stroke and HIV.
 

Past Projects

The Activities of Daily Living Exercise Robot (ADLER):

The rationale for the ADLER environment was born out of existing occupational therapy paradigms which support using purposeful tasks that mimic real activities of daily living (ADLs) to improve the generalization or carryover of practiced functional movements to unsupervised environments.  ADLER uses a HapticMaster 3 degrees of freedom, force-controlled haptic interface from FCS Moog Robotics. The main purpose of the ADLER is to help rehabilitate stroke patients who have suffered from an upper body hemiparesis. This rehabilitation process involves executing ADL tasks, wherein ADLER assists impaired arm movement and administers customized forces along programmed trajectories.  For example, a large number of stroke patients have difficulty performing ADL tasks (i.e. drinking water or combing  hair in a smooth and coordinated manner) due to motor impairment and deficits in temporal and spatial coordination.  The custom trajectories of the ADLER seeks to take advantage of brain plasticity by restoring and/or rerouting neural pathways in the brain to help regain lost mobility.

Acknowledgments

  • Grants: Advancing a Healthier Wisconsin #5520015, Medical College of Wisconsin Research Affairs Grant #3303017, NIH K25 Grant #1K25NS058577 – 04
  • Department of Physical Medicine and Rehabilitation, Medical College of Wisconsin
  • Department of Physical Medicine and Rehabilitation, University of Pennsylvania, School of Medicine
  • Dr. Katherine Kuchenbecker and Max Mintz, GRASP REU and GRASP LAB, University of Pennsylvania, School of Engineering and Applied Science
  • Company partners: Eyal Halm and Ben van Leeuwen from FCS MOOG

The Bilateral Assessment System (BiAS):

The BiAS is a low-cost, portable system which consists of two small passive robots.  Our lab is currently using this system to gain insight into impaired arm motor control on bilateral and unilateral functional activities across several patient cohorts:  stroke survivors, upper extremity (UE) amputees, spinal cord injury paraplegics, adults with cerebral palsy, and non-UE-injured control subjects.   Upon comparison of the subgroups data, a study of the respective kinematics will be performed.  By conducting this study, we aim to better  understand how changes to brain integration, neural connectivity, and muscle activity affect activities of daily living (i.e. drinking from a cup), with the ultimate goal of building better rehabilitation devices based on an increased understanding of how a particular injury affects common movements.

Acknowledgements

  • Grants:  Advancing a Healthier Wisconsin #5520015, Medical College of Wisconsin Research Affairs Grant #3303017, NIH K25 Grant #1K25NS058577 – 01A1
  • Rehabilitation Robotics Research and Design Lab, Zablocki VA Medical Center (Kimberly Wisneski, Dominic Nathan, Michael Pollemann)
  • University of Reading for development of low-cost, 3 DOF robots.
  • Brook Slavens, Terri Walton, Dr. John McGuire, Dr. Gerald Harris
  • NSF Louis Stokes Alliance for Minority Participation

Bi-ADLER:

Bi-ADLER is a therapy robot designed to help treat neurological disorders such as stroke and cerebral palsy. It is specifically designed for patients with upper extremity impairments, wherein one arm is impaired and the other one has higher degree of functionality.  In the sense that it has an active arm like ADLER, it is an extension upon this system; however, BiADLER also has a second arm – a passive one – as the name of the system suggests.  The active arm is mainly for supporting the impaired arm and the passive arm is controlled by the patient’s good arm.  Each of the arms are machined to have 6 degrees of freedom to take into consideration both position and orientation.

The passive arm is equipped with position sensors alone and used to measure the motion and orientation of the good arm. This information can then be used to control the impaired arm using Bi-ADLER’s active arm, which is equipped with both position sensors and controllers (hence the term active).  Various types of bilateral ADLS, for example symmetric, asymmetric or a mirrored movement can be performed using the system.  Other activities of daily living, such drinking from a cup, reaching, etc., can be performed using only the active arm to which the control algorithms are pre-fed; the BiAdler team is also looking into adaptive control algorithms for performing these activities.   Ultimately, we aim to improve patients’ ability to complete both unilateral or bilateral coordinated reaching and grasp tasks.

Acknowledgements

  • Research supported by RERC Technologies for Children with Orthopedic Disabilities (TECP4POD): US Department of Education, NIDRR H133E100007
  • National Institutes of Health Career Award K25NS058577-1A04.
  • Department of Physical Medicine and Rehabilitation, Medical College of Wisconsin (MCW)
  • Department of Physical Medicine and Rehabilitation, University of Pennsylvania (School of Medicine)
  • Clement J. Zablocki VA for their infrastructure support of the Rehabilitation Robotics Research and Design Lab.
  • Penn Partners and Pennsylvania Institute of Rehabilitation
  • Our colleagues from Shriners Hospital of Chicago for their support on the design. Specifically, we thank Dr. Gerald Harris, Adam Graf, Joseph Krzak and Jasmine Gilliam.
  • The BiADLER/ADLER Research Team:  Anushree Singh, Eli Soffer, Roshan Rai, Addwitey Chrungoo