The efficacy of a motorized lower-limb prosthetic device: a pilot study

BACKGROUND Performing daily activities is challenging for individuals with a transfemoral amputation. Tech- 26 nological advancements in prosthetic prototypes aim at improving functional independence. A state-of-the-art ac- 27 tive device, the CYBERLEGs-gamma (CLs-ɣ) prosthesis, consisting of powered ankle and knee joints, has been 28 designed and constructed. The control system combines pressure-sensitive insoles and inertial motor units to syn- 29 chronize both joints to work together. To date, the novel device has not been clinically evaluated. Therefore, the 30 objective of this study was to investigate the efficacy of the CLs-ɣ prosthesis during daily activities compared to 31 current passive and quasi-passive devices. 32 METHODS Participants performed a familiarization trial, an experimental trial with the current prosthesis, three 33 adaptation trials and an experimental trial with the CLs-ɣ prosthesis. Participants completed a stair climbing test, 34 a timed-up & go test, a sit to stand test, a two-minute dual task (i.e. the psychomotor vigilance task during treadmill 35 walking) and a six-minute treadmill walk test at normal speed. Nonparametric Wilcoxon-signed rank tests were 36 conducted with critical alpha set at 0.05. length and better simulated able-bodied stair ambulation.

3 Background 54 Performing daily activities can be challenging for individuals with a transfemoral amputation (TFA). For example, 55 a higher metabolic cost and reduced physical performance are shown during walking, ramp ambulation, stair 56 climbing and rising from a chair [1][2][3][4]. Moreover, task characteristics differ. People with a TFA often ambulate 57 stairs step-by-step, whereas step-over-step is common for able-bodied individuals [3]. During walking, an 58 asymmetrical pattern is observed and additional attentional effort is required due to the lack of information coming 59 from the proprioceptive system and loss of motor control [5,6]. These challenges and adaptations lead to the  actuator provides torque by compressing or decompressing series elastic springs, whereas a parallel spring system 77 acts between the shank and the foot to reduce energy consumption by storing potential and kinetic energy during 78 dorsiflexion and releasing it during plantarflexion [12]. At the knee joint, the actuator provides resistance, similar 79 to eccentric muscle power, during the weight acceptance phase and is activated right before heel strike. Afterwards, 80 the controller provides active torque, similar to concentric muscle power, during the stance phase when the knee 81 is extending. The motion is constrained to the sagittal plane, i.e. flexion and extension. The CLs-ɣ prosthesis 82 comprises the control system and electronics in the leg structure and still requires an external power source (battery 83 4 pack) to be placed on the pelvis. Electronics of the system are custom made boards to control not only the 84 prosthesis, but also act as interface between the pressure-sensitive insoles that are instrumented in the shoes of 85 both feet and the inertial motor units that are attached to the trunk and lower limbs [13]. A more detailed 86 mechatronic description is available in literature [14].

102
The aim of this study was to evaluate the CLs-ɣ prosthesis during daily activities. We hypothesized improved 103 physical performance during stair climbing, sit to stand and walking with the CLs-ɣ compared to the current 104 prosthesis. It was also hypothesized that walking with the CLs-ɣ restored a more symmetrical walking pattern (e.g.

105
step length and width, stance and swing phases, and heel pressure), and decreased physical (e.g. metabolic cost, 106 heart rate and rating of perceived exertion) and cognitive (e.g. reaction time) effort compared to the current

141
The main experimental trials with the current and CLs-γ prosthesis consisted of five consecutive tasks with ten 142 minutes of rest between each task: the stair climbing test (SCT), the timed-up & go test (TUG), the sit to stand test 6 (STS), a two-minute dual task (i.e. the psychomotor vigilance task during treadmill walking) and a six-minute 144 treadmill walk test (6MWT) [16, 19-23]. Description of each task can be found in aforementioned references. The

145
walking tests were carried out at normal walking speed. Remark that during the psychomotor vigilance task 146 participants had to respond by pushing a button with the index finger of their dominant hand. Earplugs were 147 required during the task to reduce distraction related to sound. The visual stimulus, displayed as a red dot, was 148 visualized on a computer screen with a random time interval between 1000 and 10000 ms. The interval stimulus-149 response onset was set at 500 ms and the distance to the screen was approximately one meter.  7 midfoot and heel. Cadence, stance and swing phases, step width, step and stride length were continuously 174 determined. Gait variability, expressed as the coefficient of variation, was calculated for step width, and step and 175 stride length from the standard deviation dividing by the mean [26]. Ergospirometrical data were continuously 176 gathered during the 6MWT using a portable system (Cosmed K5 ® , Cosmed, Rome, Italy). Preceding each test, a 177 calibration (volume, ambient air and reference gas) was performed after a system warm-up period of thirty minutes.

178
The setting mixing chamber was used, and data was continuously transferred to the program Omnia (Cosmed).

179
The device was mounted on the back of the participant with a harness. The net metabolic cost of transport      Figure 3A). The increased stride length was due to a greater step length of the amputated leg 215 (22 ± 20 %, p = 0.035; Table 2, Figure 3B). No significant difference in step length of the non-amputated leg was 216 reported ( Figure 3B).
Step width did not significantly change while walking with the current and CLs-γ prosthesis.

217
Coefficients of variation did not differ between both prostheses for stride length, step width, and step length of the 218 amputated and non-amputated leg. The percentage duration of stance phases of the amputated and non-amputated 219 leg did not vary between walking with the current and CLs-γ prosthesis. Additionally, swing phase did not differ 220 between both prostheses. Maximum heel pressure of the amputated and non-amputated leg did not change while 221 walking with the current compared to the CLs-γ prosthesis.

264
Task performance 265 Figure 5 shows that the normal walking speed was significantly lower with the CLs-γ compared to the current 266 prosthesis (42 ± 21 % decrease, p = 0.018; Table 2). Cadence significantly decreased 12 ± 8 % while walking with 267 the CLs-γ compared to the current prosthesis (p = 0.012; Table 2). It took participants significantly longer to 268 respond to the stimulus on the psychomotor vigilance task while walking with the CLs-γ compared to the current 269 prosthesis (19 ± 12 % increase, p = 0.012; Table 2). Participants significantly needed more time to complete the 270 SCT and TUG with the CLs-γ compared to the current prosthesis (125 ± 75 % increase, p = 0.012 and 33 ± 21 % 271 increase, p = 0.012, respectively; Table 2). Number of stands during the STS did not differ between both

279
The aim of this study was to investigate the effectiveness of a novel prosthesis, consisting of powered ankle and 280 knee joints, during daily activities. The main finding was that the CLs-γ prosthesis reduced walking symmetry, 281 and physical and cognitive effort during daily activities compared to current devices. Worth mentioning is that 282 although performance outcome measures were not improved, participants wearing the novel prosthesis were able 283 to conduct the step-over-step during stair climbing instead of the step-by-step strategy with their current prosthesis.

304
Walking with the CLs-γ prosthesis significantly increased oxygen consumption. A weight reduction of the next 305 prototype would most likely decrease the net metabolic cost since it directly influences oxygen uptake. Second, 306 the control system and interface need further development, which apparently affected gait and walking efficiency.

307
This was recently observed in a study wherein the authors outlined that the controller parameters (passive and 308 active loops) could influence the metabolic cost [34].

334
swing phase and step length of the non-amputated side and heart rate during STS) moderate effect. In a next step,

335
it is advised to include an adaptation period of three months at a higher technology readiness level (≥ 6, i.e. 'model 336 or prototype demonstration in a relevant environment') to map long-term adaptation of the novel prosthesis [11].

339
The CYBERLEGs-gamma prosthesis consists of powered ankle and knee joints meant to replace a human ankle-

340
knee system for various tasks including walking, sit to stand, and stair climbing. A higher physical and cognitive 341 effort were required during walking with the novel prosthesis compared to current devices. Overall, our hypotheses 342 were rejected, meaning that technological advancements in lower limb prosthetics are required to enable 343 synchronized and powered ankle and knee joints. Although performance outcome measures were not improved, 13 participants wearing the novel prosthesis better simulated able-bodied stair ambulation. All participants were able 345 to conduct stairs with the step-over-step instead of the step-by-step strategy.