Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Exercise-stimulated primary cilia on preosteoclasts promote periosteal-bone formation.

Experimental & molecular medicine·2026
Same author

Critical role of low cloud feedback in irreversible sea level rise.

Nature communications·2026
Same author

Paretic limb biomechanical response to hip exoskeleton limb assistance strategies during walking for individuals post-stroke.

Journal of biomechanics·2026
Same author

Contemporary United States PRRSV 1-4-4 L1C.5 isolate causes severe disease comparable to historic highly pathogenic PRRSV.

Virology·2025
Same author

Deep domain adaptation eliminates costly data required for task-agnostic wearable robotic control.

Science robotics·2025
Same author

Toward personalizing prosthesis prescription: A take-home study of three microprocessor-controlled prosthetic knees: A randomized crossover study.

PM & R : the journal of injury, function, and rehabilitation·2025

Related Experiment Video

Updated: May 20, 2025

A Structured Rehabilitation Protocol for Improved Multifunctional Prosthetic Control: A Case Study
06:58

A Structured Rehabilitation Protocol for Improved Multifunctional Prosthetic Control: A Case Study

Published on: November 6, 2015

9.4K

Mode-Unified Intent Estimation of a Robotic Prosthesis using Deep-Learning.

Hanjun Kim1, Dawit Lee2, Jairo Y Maldonado-Contreras1,3

  • 1Hanjun Kim is with the Woodruff School of Mechanical Engineering, Georgia Tech, Atlanta, GA 30332-0405 USA.

IEEE Robotics and Automation Letters
|March 24, 2025
PubMed
Summary

This study introduces a unified approach for robotic prostheses to recognize user intent by continuously estimating terrain slopes, improving accuracy over traditional discrete mode classifiers for transfemoral amputees.

Keywords:
Prosthetics and exoskeletonsdeep learningintention recognitionmode unificationslope estimation

More Related Videos

Engineering Platform and Experimental Protocol for Design and Evaluation of a Neurally-controlled Powered Transfemoral Prosthesis
11:16

Engineering Platform and Experimental Protocol for Design and Evaluation of a Neurally-controlled Powered Transfemoral Prosthesis

Published on: July 22, 2014

16.2K
Characterization of the Sense of Agency over the Actions of Neural-machine Interface-operated Prostheses
05:21

Characterization of the Sense of Agency over the Actions of Neural-machine Interface-operated Prostheses

Published on: January 7, 2019

7.8K

Related Experiment Videos

Last Updated: May 20, 2025

A Structured Rehabilitation Protocol for Improved Multifunctional Prosthetic Control: A Case Study
06:58

A Structured Rehabilitation Protocol for Improved Multifunctional Prosthetic Control: A Case Study

Published on: November 6, 2015

9.4K
Engineering Platform and Experimental Protocol for Design and Evaluation of a Neurally-controlled Powered Transfemoral Prosthesis
11:16

Engineering Platform and Experimental Protocol for Design and Evaluation of a Neurally-controlled Powered Transfemoral Prosthesis

Published on: July 22, 2014

16.2K
Characterization of the Sense of Agency over the Actions of Neural-machine Interface-operated Prostheses
05:21

Characterization of the Sense of Agency over the Actions of Neural-machine Interface-operated Prostheses

Published on: January 7, 2019

7.8K

Area of Science:

  • Robotics
  • Biomechanics
  • Prosthetics

Background:

  • Traditional robotic prostheses use discrete modes (level, ramp, stairs) for ambulation, which is insufficient for continuous human movement.
  • Existing mode classifiers struggle with accurate intent recognition due to the continuous nature of terrain variations.

Purpose of the Study:

  • To develop and validate a mode-unified intent recognition strategy for transfemoral amputations.
  • To enable continuous slope estimation for more accurate prosthetic limb control across diverse terrains.

Main Methods:

  • Utilized deep temporal convolutional networks trained on locomotion data from 16 individuals with transfemoral amputation.
  • Developed a mode-unified slope estimator and compared its performance against traditional mode classifiers using leave-one-subject-out validation.
  • Evaluated the system's ability to replicate able-bodied knee kinematics.

Main Results:

  • The mode-unified slope estimator achieved a lower Mean Absolute Error (MAE) of 1.68 ± 0.60 degrees compared to the mode classifier's 1.94 ± 0.97 degrees (p<0.05).
  • The proposed system significantly improved knee kinematics replication, with MAE of 5.13 ± 2.00 degrees for knee clearance and 6.74 ± 2.97 degrees for knee contact angle during stair ascent.
  • These results were significantly better than the traditional classifier's MAE of 12.10 ± 5.20 degrees and 13.80 ± 3.28 degrees (p<0.01).

Conclusions:

  • A mode-unified approach enables continuous terrain adjustment in robotic prostheses without relying on discrete mode classification.
  • This strategy enhances prosthetic limb control and improves the naturalness of gait for individuals with transfemoral amputations.
  • The findings suggest a more intuitive and adaptive control system for lower-limb prosthetics.