DEVELOPMENT OF A SMART KNEE BRACE

K. Rudolph,J. Sun,D. S. Reisman,K. Hauck,Joseph White
DOI: https://doi.org/10.1097/01.npt.0000281337.40599.54
2006-01-01
Journal of Neurologic Physical Therapy
Abstract:Purpose/Hypothesis: Patients with hemiparesis lack knee control during walking and may compensate with knee hyperextension. We present the development of a computer controlled Smart Knee Brace (SKB) designed to dynamically control the knee during stance, while allowing unrestricted knee motion during swing. Brace design is presented along with a novel gait event detection algorithm (GED) that detects stance and swing phases of the gait cycle using only kinematic data. Number of Subjects: Data from 2 patients with hemiparesis and 1 neurologically intact individual are presented. Materials/Methods: The SKB was designed on the chassis of Donjoy Armor sports brace. Sensors on the thigh and shank segments of the brace include gyroscopes and biaxial accelerometers to detect rotations and accelerations, respectively. A potentiometer measures knee angle and compression closing foot switches detect stance and swing. Adjustable flexion and extension locks limit the knee from moving beyond preset limits set by a therapist. A damper that uses magnetorheologic (MR) fluid will control the knee during the loading response phase of gait until the preset locking angles are reached. To verify the brace measurements, data were collected from one control and 1 person with hemiparesis using the SKB measurements that were collected simultaneously with a Vicon Motion Analysis System with a video collection rate of 120 Hz and an analog rate of 600 Hz. Knee position from the SKB was compared to that measured by the Vicon Motion Capture system and calculated with Visual3D software. Data from 10 consecutive strides were averaged and compared at 4 points in the gait cycle. Initial contact (IC) and toe off (TO) events were compared between the foot switches and the GED algorithm. Results: The errors between the GED algorithm compared to the foot switch signals were 0.50% gait cycle (%GC) at IC and 9.13% GC at TO in the neurologically intact subject. In the subject with hemiparesis, error in detecting IC was −1.98% GC and 6.17%GS for TO. The flexion lock engaged 10.3 msec after the signal was sent from the computer. In the neurologically intact subject, knee motion measured by the SKB was quite similar to that collected by the motion capture system. The difference in brace angle and actual knee angle ranged from 0.2–3.16 degrees during mid-stance and swing respectively. Conclusions: The SKB is capable of measuring knee motion and detecting gait events well. Errors in knee motion are higher during extreme flexion or extension ranges, most likely due to tissue compression. Errors in GED are lower at IC than TO. Clinical Relevance: Recent research shows that plastic changes can occur in the nervous system with massive practice of the appropriate movement patterns. The SKB is designed to allow patients to experience typical movement patterns during the loading phase of the gait cycle even when volitional knee control is limited. The SKB will provide physical therapists an important tool to enhance gait re-training in patients with hemiparesis.
What problem does this paper attempt to address?