Congratulations to the U-Build Bionic Knee for winning the #NCMRR @NICHD_NIH @NIBIB DEBUT Competition🎊😍!!! They build a low-cost powered knee #prosthesis🦿 and made the design #opensource for everyone! youtube.com/watch?v=xS9t9RJg… @UUtah @UtahCoE @UofUNews @UtahMech @UtahBME

6) Next, we trained 2 more models: an IMU-based Model, and a Combined Model trained on both IMU and A-mode ultrasound. The Combined Model had the lowest errors in most cases. Therefore, A-mode ultrasound adds important information about the user’s intent that IMU can’t capture!

5) Here is the kinematic prediction error (difference between predicted and measured kinematics). Top: Average error for each activity. Bottom: Average error for each variable across activities (bottom) Our accuracy is similar to previous work using #EMG and B-Mode ultrasound!

4) We used the ultrasound-based model to predict the kinematics of the prosthesis knee and ankle. Comparing the predicted kinematics (blue line) to the measured kinematics (black), we can evaluate accuracy. Overall, it did pretty well!

2) Seven #transfemoral amputee subjects performed 20 trials of our “ambulation circuit” using their prescribed passive #prosthesis. We recorded lower-limb kinematics from both legs using #xsens #IMU, and we recorded A-mode #ultrasound data from 4 sensors on their residual limb

1) In our newest paper, we used A-Mode ultrasound and machine learning to predict lower-limb kinematics while above-knee amputees performed various activities in our recently published paper in IEEE-TBME doi.org/10.1109/TBME.2023.32… 🧵 #academicTwitter #SciComm #PhDchat

5/6) We also looked at the energy dissipated during sit-down by each joint. On the prosthesis side, the powered prosthesis dissipated more energy than the passive prosthesis at the hip, knee, and ankle for every subject. However, the intact leg didn’t improve much.

4/6) Passive prostheses provide some support during sit-down using passive dampers. We still saw the powered prosthesis providing more torque than passive. Perhaps even more important, the powered leg's peak torque occurs later in the movement when users need the most support!

3/6) We hypothesized that the intact leg’s quadriceps muscle effort would decrease, but we didn’t see a significant difference

2/6 ) Ten above-knee amputees sat down using their passive prostheses and our powered prosthesis. This graph shows weight-bearing symmetry: zero is “perfect” symmetry. The powered leg improved symmetry significantly, and the powered leg improved symmetry for every subject! 🤩

1/6 We just published a new paper, “Can a powered knee-ankle prosthesis improve weight-bearing symmetry during stand-to-sit transitions in individuals with above-knee amputations?” Read on to learn more! 🧵doi.org/10.1186/s12984-023-0… #academicTwitter #SciComm #PhDchat