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 Experiment Videos

Multimodal Feature-Level Fusion CBAM U-Net for Static Plantar Pressure Prediction Using Plantar Geometry and Sparse

Chongguang Wang1, Kerrie Evans1,2,3, Dean Hartley1,2

  • 1School of Mechanical and Mining Engineering, The University of Queensland, St Lucia, QLD 4072, Australia.

Sensors (Basel, Switzerland)
|July 15, 2026
PubMed
Summary

Related Concept Videos

Functional Classification of Joints01:09

Functional Classification of Joints

Functional Classification of Joints
The functional classification of joints is determined by the amount of mobility between the adjacent bones. Joints are functionally classified as a synarthrosis or immobile joint, an amphiarthrosis or slightly moveable joint, or as a diarthrosis, a freely moveable joint. Fibrous and cartilaginous joints can be functionally classified as either synarthroses  or amphiarthroses, whereas all synovial joints are classified as diarthroses.
Synarthrosis
An immobile...

You might also read

Related Articles

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

Sort by
Same author

Defining a High-Quality Living Kidney Donor Evaluation: A Delphi Survey and Consensus Conference Summary Report.

Canadian journal of kidney health and disease·2026
Same author

Reconstruction of gait variables from 3D foot geometry and sparse plantar pressure landmarks using deep learning.

Journal of biomechanics·2026
Same author

GLP-1 receptor agonists and weight-loss strategies for individuals with obesity and hip or knee osteoarthritis: a scoping review.

British journal of sports medicine·2026
Same author

IMU-Based Wearable Insoles in Clinical Settings: Key Parameters Differentiating Clinical and Non-Clinical Populations.

Sensors (Basel, Switzerland)·2026
Same author

Is palpation essential in the digital era of orthotic designing?

Frontiers in bioengineering and biotechnology·2026
Same author

Effect of Ni Addition on the Solidification of Liquid Al and Solid Cu Diffusion Couples.

Materials (Basel, Switzerland)·2025

This study introduces a deep learning model that accurately predicts plantar pressure using foot shape and few sensor points. This approach enables low-cost intelligent insoles for biomechanical analysis and clinical use.

Area of Science:

  • Biomechanics
  • Medical Technology
  • Artificial Intelligence

Background:

  • Accurate plantar pressure mapping is crucial for gait analysis, rehabilitation, and diabetic foot care.
  • Current wearable systems face limitations due to sparse sensor layouts, impacting hardware complexity, power, and comfort.

Purpose of the Study:

  • To develop a multimodal deep learning framework for predicting plantar pressure distribution using plantar geometry and sparse landmark data.
  • To evaluate different network architectures, fusion strategies, and landmark densities for optimal performance.

Main Methods:

  • A convolutional block attention module U-Net architecture was designed for dual-encoder feature fusion with attention refinement.
  • Plantar geometry and sparse landmark data were integrated as distinct modalities.
Keywords:
CBAM U-Netfoot plantar geometrymultimodal fusionplantar pressure predictionsparse-to-dense reconstruction

Related Experiment Videos

  • A controlled-variable experimental design was used to systematically evaluate various configurations.

    • Main Results:

    • Feature-level fusion outperformed data-level fusion and unimodal approaches across all tested landmark densities.
    • The model achieved a normalized root mean square error of 0.087 with 16 landmarks and 0.138 with only two landmarks.
    • Feature-level fusion effectively captured complementary information between plantar geometry and anatomical landmarks, especially with sparse data.

    Conclusions:

    • Multimodal feature-level fusion is an effective strategy for sparse-to-dense plantar pressure reconstruction.
    • The proposed framework supports the development of cost-effective intelligent insole systems for biomechanical monitoring and clinical applications.