Abstract:Tracking skeletal muscle architecture using B-mode ultrasound is a widely used method in the field of human movement science and biomechanics. Sequential methods based on optical flow algorithms allow for smooth and coherent muscle tracking but are known to drift over time. Non-sequential feature detection methods on the other hand, do not suffer from drift, but are limited to tracking only lower-dimensional features. They are also known to be sensitive to image noise, and therefore often result in highly irregular tracking patterns. Building on the complimentary nature of both approaches, we present a novel fully automated hybrid muscle tracking approach that combines a sequential feature-point tracking method and a non-sequential method based on Hough transform. Tibialis anterior fascicle pennation angle and length, and central aponeurosis displacement, were measured in five healthy individuals during isometric contractions at five different ankle angles. Our hybrid method was demonstrated to significantly (p < 0.001) reduce drift compared to two sequential methods, and curve irregularity was significantly (p < 0.001) decreased compared to a non-sequential method. These findings suggest that the proposed hybrid approach can uniquely mitigate drift and irregularity limitation of individual methods used for tracking skeletal muscle architecture. Fully automated muscle tracking allows for convenient analysis of large datasets, whereas automatic drift correction opens the door for tracking muscle architecture in long ultrasound recordings during common movements, such as walking, running, and jumping without the need for manual intervention.

An iterative approach to Hough transform without re-voting

Skeleton-driven 2D Distance Field Metamorphosis Using Intrinsic Shape Parameters

Constrained Fitting for 2D Profile-Based Reverse Modeling

Human Pose Tracking Algorithm Based on Skeleton-Texture Model

Application of Homotopy Method in Reverse Engineering Constraints Solution

A functional skeleton transfer

Technical Section: 2D shape morphing via automatic feature matching and hierarchical interpolation

Curve Skeleton Extraction From 3D Point Clouds Through Hybrid Feature Point Shifting and Clustering

A Robust Multi‐View System for High‐Fidelity Human Body Shape Reconstruction

Generalizing the Hough transform to detect arbitrary shapes

A 3D 2-subiteration thinning algorithm for human pose estimation

A Hybrid Method for Ultrasound-Based Tracking of Skeletal Muscle Architecture

Skeletal Representations and Applications

Fast and robust curve skeletonization for real-world elongated objects

Mass-Driven Topology-Aware Curve Skeleton Extraction from Incomplete Point Clouds

Mobility Fitting Using 4d Ransac

Personal Virtual Human Modeling from Images

Deformable modeling using a 3D boundary representation with quadratic constraints on the branching structure of the Blum skeleton

A 3D Skeleton Points-Based Hierarchical Body Modeling Approach for Intelligent Online Clothing Fitting Systems

Shape Reconstruction for Continuum Robot Based on Pythagorean Hodograph–Bézier Curve With IMU and Vision Sensors

Human motion tracking based on shape analysis