Related Experiment Video
Updated: Dec 18, 2025

Author Spotlight: Addressing Technical and Subjective Challenges in Measuring Classroom Attention
Published on: December 15, 2023
Application of Convolutional Neural Network in Identifying Different Levels of Isokinetic Exercise Efforts
1Department of Forensic Medicine, School of Basic Medical Sciences, Fudan University, Shanghai 200032, China.
A convolutional neural network (CNN) accurately identifies moment of force-time diagrams from isokinetic knee exercises, distinguishing maximal from half effort. This technology aids in assessing patient exertion levels during rehabilitation and training.
Area of Science:
- Biomechanics
- Rehabilitation Engineering
- Artificial Intelligence
Background:
- Isokinetic knee exercises are crucial for assessing muscle function and guiding rehabilitation.
- Accurate quantification of effort levels during these exercises is essential for effective training and recovery.
- Distinguishing between maximal and sub-maximal efforts from force-time data presents a challenge.
Purpose of the Study:
- To develop and validate a convolutional neural network (CNN) model for classifying isokinetic knee exercise moment of force-time diagrams.
- To differentiate between maximal and half effort levels based on collected force-time data.
- To assess the accuracy and reliability of the developed CNN model in identifying effort variations.
Main Methods:
- Collected moment of force-time diagrams from 200 healthy volunteers performing isokinetic knee flexion-extension exercises at maximal and half effort.
- Utilized two angular velocities (30°/s and 60°/s) and repeated trials for data acquisition.
- Trained CNN models using a randomly assigned training set (140 subjects) and validated performance on a separate testing set (60 subjects) over three random sampling iterations.
Main Results:
- Generated 2,400 moment of force-time diagrams representing maximal and half effort levels.
- Achieved classification accuracy rates of 91.11%, 90.49%, and 92.08% across three model training iterations.
- Attained an average classification accuracy of 91.23% for differentiating effort levels.
Conclusions:
- The developed CNN models demonstrate high efficacy in distinguishing between maximal and half effort isokinetic knee exercise diagrams.
- This AI-driven approach offers a reliable method for objectively assessing effort exertion during isokinetic testing.
- The findings support the application of CNNs in clinical settings for enhanced monitoring of patient engagement and performance in rehabilitation programs.
More Related Videos
Related Concept Videos
Motor Unit Stimulation
The latent period of contraction marks the onset of excitation-contraction coupling, when the action potential propagates across the sarcolemma, preparing the muscle fibers for contraction. As the fibers enter the contraction phase, the...
Classification of Skeletal Muscle Fibers
Slow-Twitch Muscle Fibers
Slow oxidative, muscle fibers appear red due to large numbers of capillaries and high levels of...
Exercise and Cardiac Output
Sustained exercise increases the muscles' oxygen demand, which can be...
Correlation between ECG and Cardiac Cycle
A cardiac action potential originates in the SA node and spreads throughout the atria and the AV node in approximately 0.03 seconds. This results in the P wave in an ECG and triggers atrial contraction. The action potential is then briefly slowed at the AV node, allowing the atria to contract and fill the ventricles with blood before...
Exercise and Cardiovascular Response
Light to moderate physical activity initiates a series of interconnected responses in the body. The heart rate modestly increases in anticipation of the workout, followed by widespread vasodilation as oxygen consumption by skeletal muscles increases. This results in decreased peripheral resistance, increased capillary blood flow, and accelerated...

