Most individuals with lower limb amputations use passive prostheses, which do not provide energy to assist with activities such as stair or ramp ascent or sit-to-stand transitions. This limits mobility, in particular for those with above-knee amputations. Powered leg prostheses could improve the mobility and community participation of such individuals; however, these devices are heavy, and current control systems require the user to manually transition between different ambulation activities, which is cumbersome. With prior R01 funding, we developed an adaptive, hierarchical pattern recognition control system that uses data from sensors on the prosthesis, and incorporates electromyographic (EMG) signals from the user, depending on their reliability, to determine user intent and enable safe prosthesis control with automatic, seamless transitions between ambulation activities. A mobile application allows for rapid tuning of the prosthesis and enables the user to choose between manual or automatic transitions. With other funding, we developed a novel prosthetic leg that can operate in passive mode ?during level-ground walking or in active mode?during activities such as stair climbing or sit-to-stand transitions. This approach enables smaller, lighter motors, transmissions, and batteries, making our Hybrid Leg significantly lighter and quieter than other powered devices. Our long-term objective is to develop clinically viable technologies to improve the quality of life for lower limb amputees. A lightweight powered prosthesis with a safe, intuitive control system may increase mobility?facilitating employment, leisure, and community participation activities?and reduce the physical and psychological consequences of low activity. We will compare the Hybrid Leg with subjects' passive devices in both in-lab and home environments.
In Aim 1, we will transition our adaptive control system to the Hybrid Leg, train users to walk with this device while the experimenter manually transitions the device between activity modes, and collect sensor data and EMG signals to create a user-specific pattern recognition control system. We will then determine the classification accuracy of this system.
Aims 2 and 3 together constitute a randomized clinical trial, with AB-BA design, comparing the Hybrid leg with subjects' own passive devices.
In Aim 2, we will provide advanced community-mobility training for either the subject's passive leg or the Hybrid leg, in random order, to meet both subject-specific and general activity goals necessary for community ambulation, and complete a full biomechanical assessment of ambulation activities such as stair or ramp ascent/descent and sit-to-stand transitions with that leg.
In Aim 3, subjects will use the same leg for 4 weeks in their home and community, where activity and community participation will be monitored by a custom smartphone?based app. We will compare the number of steps taken and number of transitions between activities for each device. We expect that the control system will be safe, with a low classification error rate and without errors that may cause a fall. In addition, we hypothesize that, using the Hybrid leg, subjects will ambulate more and transition between activities more frequently, with biomechanics more similar to those of non-amputees.
Compared to passive devices, powered prosthetic legs may enable individuals with lower limb amputation to walk more smoothly using less energy, ambulate more efficiently up stairs and ramps, and perform sit-to-stand transitions more easily. However, such devices are heavy, and control systems for these devices require manual transitions between activities. We will compare biomechanics and community activity of subjects using their commercially available, passive devices and a novel, lightweight hybrid prosthetic leg that can function as a powered or passive device and is controlled by an adaptive pattern recognition control system (developed under our previous R01 award) that enables safe, automatic, and seamless transitions between ambulation modes.
| Spanias, J A; Simon, A M; Finucane, S B et al. (2018) Online adaptive neural control of a robotic lower limb prosthesis. J Neural Eng 15:016015 |
| Simon, Ann M; Ingraham, Kimberly A; Spanias, John A et al. (2017) Delaying Ambulation Mode Transition Decisions Improves Accuracy of a Flexible Control System for Powered Knee-Ankle Prosthesis. IEEE Trans Neural Syst Rehabil Eng 25:1164-1171 |
| Kuiken, Todd A; Barlow, Ann K; Hargrove, Levi et al. (2017) Targeted Muscle Reinnervation for the Upper and Lower Extremity. Tech Orthop 32:109-116 |
| Woodward, Richard B; Spanias, John A; Hargrove, Levi J (2016) User intent prediction with a scaled conjugate gradient trained artificial neural network for lower limb amputees using a powered prosthesis. Conf Proc IEEE Eng Med Biol Soc 2016:6405-6408 |
| Simon, Ann M; Spanias, John A; Ingraham, Kimberly A et al. (2016) Delaying ambulation mode transitions in a powered knee-ankle prosthesis. Conf Proc IEEE Eng Med Biol Soc 2016:5079-5082 |
| Spanias, John A; Simon, Ann M; Perreault, Eric J et al. (2016) Preliminary results for an adaptive pattern recognition system for novel users using a powered lower limb prosthesis. Conf Proc IEEE Eng Med Biol Soc 2016:5083-5086 |
