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 Video

Updated: Jun 26, 2026

Single-stage Dynamic Reanimation of the Smile in Irreversible Facial Paralysis by Free Functional Muscle Transfer
19:53

Single-stage Dynamic Reanimation of the Smile in Irreversible Facial Paralysis by Free Functional Muscle Transfer

Published on: March 1, 2015

Objective grading of facial paralysis using Local Binary Patterns in video processing.

Shu He1, John J Soraghan, Brian F O'Reilly

  • 1Department of Electronic&Electrical Engineering at the University of Strathclyde, UK. shu.he@eee.strath.ac.uk

Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference
|January 24, 2009
PubMed
Summary
This summary is machine-generated.

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

A shape optimization method for resistance reduction of local piping components with multiscale validation.

Communications engineering·2026
Same author

Unveiling the Fragrant Secrets of <i>Dendrobium devonianum</i>: Terpenoid Pathways and Floral Scent Dynamics.

Metabolites·2026
Same author

Loss of METTL14 in dopaminergic neurons disrupts ER homeostasis via m6A-dependent regulation of Atp2a3 mRNA: Implications for Parkinson's Disease.

NPJ Parkinson's disease·2026
Same author

A NeuroD1 AAV-Based Gene Therapy for Functional Brain Repair in Alzheimer's Disease-Like Non-Human Primate Model.

Advanced science (Weinheim, Baden-Wurttemberg, Germany)·2026
Same author

Electronic Interface Interaction on Al<sub>2</sub>O<sub>3</sub>/Ag Inverse Catalysts for Enhanced Catalytic Reduction of 4-NP.

The journal of physical chemistry letters·2026
Same author

The axon initial segment-associated microglia regulate neuronal activity and visual perception.

Cell research·2026
Same journal

Analysis of End-Tidal CO2 Variability During Plateau Waves Episodes: An Information Theoretic Approach<sup></sup>.

Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference·2025
Same journal

AI and Tomosynthesis for Breast Cancer Molecular Subtyping: A step toward precision medicine<sup></sup>.

Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference·2025
Same journal

Towards Sustainable Protein Recovery from Biological Waste: Assessing Polyethersulfone-based Microfiltration.

Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference·2025
Same journal

Analysis of the cardiovascular response to standardized polymicrobial peritonitis experimental model.

Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference·2025
Same journal

Automated Wrist Ultrasound Image Bone Enhancement and Segmentation Using Deep Learning.

Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference·2025
Same journal

A Deep Learning approach for Depressive Symptoms assessment in Parkinson's disease patients using facial videos.

Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference·2025
See all related articles

This study introduces a new computer-based method to objectively assess the severity of facial paralysis from video recordings. By analyzing facial movement patterns and symmetry, the system provides a reliable score that matches standard clinical grading scales.

Area of Science:

  • Biomedical engineering and Local Binary Patterns analysis
  • Medical imaging and diagnostic informatics

Background:

Clinical assessment of facial nerve impairment often relies on subjective observation by medical professionals. Such qualitative evaluations frequently suffer from inter-observer variability and lack the precision required for longitudinal tracking. No prior work had resolved the need for automated, objective metrics that capture subtle motion deficits. Researchers have explored various image processing techniques to quantify muscle activity in clinical settings. However, existing approaches often struggle with computational complexity or sensitivity to environmental noise. That uncertainty drove the development of more robust feature extraction frameworks for dynamic facial analysis. This paper addresses these limitations by leveraging temporal-spatial patterns to standardize diagnostic reporting. The current landscape necessitates a shift toward reproducible, machine-driven quantification of neurological motor function.

Purpose Of The Study:

The primary aim of this research is to establish a novel framework for the objective measurement of facial paralysis using biomedical video data. Current clinical practices often lack the precision needed for consistent diagnostic reporting. This study seeks to bridge the gap between subjective human observation and quantitative machine analysis. The authors intend to develop a system that captures both motion information and appearance features from facial regions. They focus on improving algorithmic robustness by minimizing the impact of noise in clinical recordings. A specific goal involves creating a multi-resolution feature vector that remains computationally simple for practical application. The researchers also aim to validate their approach by correlating automated scores with the established House-Brackmann Scale. This work addresses the urgent need for standardized, reproducible metrics in the assessment of neurological motor impairment.

Keywords:
computer visionmedical diagnosticsfeature extractionsupport vector machines

Frequently Asked Questions

The researchers utilize a multi-resolution extension of uniform Local Binary Patterns to capture both micro-patterns and large-scale patterns. This approach creates a feature vector that improves algorithmic robustness while minimizing noise, unlike simpler methods that often fail to distinguish between subtle movement deficits and background interference.

The Resistor-Average Distance (RAD) serves as the primary metric for evaluating symmetry. This tool compares the extracted features from the left and right sides of the face, whereas traditional clinical assessments rely on visual inspection to estimate the degree of asymmetry between facial halves.

A Support Vector Machine (SVM) is necessary to map the extracted feature vectors to the House-Brackmann (H-B) Scale. This classification step allows the system to provide a quantitative evaluation, contrasting with manual grading which requires subjective interpretation of clinical symptoms by a physician.

More Related Videos

Dynamic Digital Biomarkers of Motor and Cognitive Function in Parkinson's Disease
10:28

Dynamic Digital Biomarkers of Motor and Cognitive Function in Parkinson's Disease

Published on: July 24, 2019

Related Experiment Videos

Last Updated: Jun 26, 2026

Single-stage Dynamic Reanimation of the Smile in Irreversible Facial Paralysis by Free Functional Muscle Transfer
19:53

Single-stage Dynamic Reanimation of the Smile in Irreversible Facial Paralysis by Free Functional Muscle Transfer

Published on: March 1, 2015

Dynamic Digital Biomarkers of Motor and Cognitive Function in Parkinson's Disease
10:28

Dynamic Digital Biomarkers of Motor and Cognitive Function in Parkinson's Disease

Published on: July 24, 2019

Main Methods:

The research team developed a computational framework to process clinical video recordings of patients. They utilized temporal-spatial domain analysis to isolate motion information across horizontal and vertical axes. Apex frames were specifically targeted to extract appearance features from distinct facial regions. The investigators applied block schemes to enhance the spatial and temporal resolution of these extracted data points. A multi-resolution extension of uniform patterns was implemented to combine micro-scale and large-scale features into a single vector. Symmetry was evaluated by calculating the Resistor-Average Distance between features derived from the left and right sides. A Support Vector Machine classifier was then trained to map these vectors to the House-Brackmann Scale. Validation involved testing the entire pipeline on a dataset consisting of 197 individual subject videos.

Main Results:

The proposed framework successfully achieved objective measurement of facial paralysis across a dataset of 197 subject videos. The integration of multi-resolution uniform patterns significantly increased algorithmic robustness against noise. By combining micro-patterns and large-scale patterns, the system maintained computational simplicity while improving feature representation. Symmetry analysis using the Resistor-Average Distance provided a consistent metric for quantifying motor deficits. The Support Vector Machine effectively translated these features into standardized House-Brackmann Scale scores. Experimental validation confirmed the high accuracy of the automated grading process compared to manual assessments. The system demonstrated efficiency in processing dynamic facial movements during the testing phase. These results indicate that the framework is suitable for reliable clinical evaluation of facial nerve function.

Conclusions:

The authors demonstrate that their framework provides a reliable, objective assessment of facial paralysis severity. Their approach effectively correlates automated measurements with established clinical grading standards. This synthesis suggests that integrating temporal-spatial features enhances the accuracy of diagnostic tools. The researchers highlight that their multi-resolution strategy successfully balances computational efficiency with algorithmic robustness. These findings imply that objective metrics can reduce reliance on subjective human observation in clinical practice. The study confirms that symmetry analysis via Resistor-Average Distance offers a viable pathway for quantifying motor deficits. By utilizing Support Vector Machines, the system achieves consistent classification performance across a large dataset. Future clinical workflows may benefit from these automated techniques to improve patient monitoring and treatment evaluation.

The framework processes video data by extracting motion information in both horizontal and vertical directions. These temporal-spatial features are enhanced through block schemes, providing a more detailed representation of muscle activation than static image analysis, which ignores the dynamic nature of facial expressions.

The system measures the degree of facial paralysis by calculating the distance between LBP features on opposite sides of the face. This phenomenon of asymmetry serves as the basis for the final score, whereas healthy individuals exhibit minimal distance values compared to those with significant nerve damage.

The authors propose that their method enhances diagnostic efficiency by providing a standardized, objective score. They suggest that this framework could reduce inter-observer variability, unlike current clinical practices that are limited by the subjective nature of human-based grading systems.