Evaluation of a prosthetic arm for transradial Amputation in Bimanual Activities

Introduction: There have been significant breakthroughs in developing highly functional myoelectric prostheses, yet individuals who have experienced upper-limb loss consistently report low levels of satisfaction when performing daily tasks while using myoelectric prostheses. This research aims to quantify the user’s experience while using a myoelectric prosthesis and evaluate the change in the user’s experience after performing a developed training program for a week. The user’s experience was evaluated by using our experimental protocol. This protocol documents our evaluation procedure and training program designed for myoelectric prosthesis users with transradial amputation. It also describes the outcome measures and the motion capture system and data analysis pipeline

Materials and

Methods: One team member with a transradial amputation was consented to be the study participant. The participant worked with a CPO at Silver Spring Medical Center to identify the reliable flexor and extensor muscle on the residual part of forearm for gripping actuation. A brace, which contains the electrodes and battery, was custom-designed to fit user’s residual forearm and was connected to an i-Limb® hand model (Össur, Reykjavík, Iceland)



The participant brought the prosthesis back for training at home and lab evaluation after the adjustment of electrodes and brace were completed. The participant performed three training tasks to use the arm prosthesis to achieve good object handling inside the participant’s shoulder range of motion. The participant was asked to train two 30-minute sessions a day for a week. The evaluation was performed to collect the participant movement using marker based motion capture, and vertical ground reaction force using pressure mat at the baseline and post training when the participant performed six functional and bimanual tasks (Figure 1).



Marker data was preprocessed to produce joint angles in Nexus software and the foot pressure data was preprocessed to produce vertical ground reaction force and center of force in the Strideway software. A MATLAB data storage script was developed to store raw and processed data. A MATLAB post-processing script was developed to analyze the outcome measures (joint angle, movement smoothness, object control, vertical ground reaction force and center of force trajectory).

Results, Conclusions, and Discussions: We have compared the outcome measures of one trial of targeted box block test (tbbt) between baseline (pre1) and post training (post1). The task completion time at this baseline trial was 55 seconds versus 48 seconds at this post training trial.​The trunk and shoulder angles are compared. Table 1 shows trunk and shoulder range of motion (ROM) in its respective degree of freedom. Four out of six angle ROMs were reduced.



The smoothness of movement for one trial of tbbt task performed during the baseline and post evaluation can be seen in Figure 2, with smoothness on the vertical axis and normalized time on the horizontal axis. Smoothness was calculated as the third derivative of left-hand marker position. The participant's head movement was plotted against the left-hand movement in temporal domain. The phase shift between these movements were observed to determine how well the participant controls the object when using prosthesis as shown in Figure 3.



Results showed some enhancements in outcome measures in a post-training trial compared to a baseline trial. Further analysis is under-way to examine our hypothesis regarding improvements in joint angle, range of motion, movement smoothness, and object control in other performed tasks. 



Comparison shows the shoulder ROM increased in all three DOFs. Shoulder flexion and abduct decreased by 44 deg and 35 deg respectively. That is what we expected. However, the shoulder internal rotation ROM decreased by 71 degrees which is opposite to what we expected. This might be because the participant relied on the pelvis rotation to help transfer blocks as the trunk internal rotation stays same. Further data analysis on pelvis and lower extremity joints and other tasks will help us examine our hypothesis. In general, as the users are more adaptive to the device, they are expected to have less ROM in the proximal part of body and are expected to rely more on their distal part of upper body to perform the tasks. The vertical ground reaction force and center of force trajectories were evaluated for one trail of post tbbt (Figure 4) after the connection issue was addressed.

Acknowledgements (Optional): This research was sponsored by the 2022 Waterman Award granted by Dr. Lara Thompson, under the Center of Biomechanical & Rehabilitation Engineering.​