Name
Title | ||
|---|---|---|
Dynamic measurement of pennation angle of gastrocnemius muscles obtained from ultrasound images based on gradient Radon transform |
Abstract | ||
|---|---|---|
In quantitative studies, pennation angle is one of the essential parameters on skeletal muscles and could be used for prosthesis control. Ultrasound imaging is widely applied in the mechanical analysis for measuring the changes in the pennation angle of human muscles. At present, many innovative methods are presented to promote the rapid development of muscle analysis. However, these methods always pose a few limitations, such as measurement errors and time-curve roughness. In this study, we developed a novel method using a weighted-average strategy from a vast amount of muscle fascicles based on gradient Radon transform. Also, this study aimed to realize a fully automatic and precise measurement of the pennation angle during muscle contraction in ultrasound images and evaluate the performance of the proposed method by comparing it with the manual method, muscle torque, and other state-of-the-art methods. The coefficient of multiple correlations (CMC) and the linear regression with a Bland–Altman analysis were implemented to evaluate the reliability. Pearson correlation analysis and polynomial regression analysis were applied to describe the association between the pennation angle and the corresponding muscle torque. Meanwhile, a new criterion, roughness, has been defined to assess the smoothness of the time-pennation angle curve during muscle contraction. The experimental results indicated that the time-angle curves calculated by the proposed method are highly correlated with the torque of muscle contraction (Pearson correlation coefficient = 0.91 ± 0.04, R = 0.87 ± 0.14), in closer agreement with manual results (CMC = 0.83 ± 0.15) and smoother (RN = 0.26 ± 0.10) than other state-of-the-art methods. In conclusion, the proposed novel method can solve the limitations including measurement errors and poor smoothness as reported in the earlier methods and provide an accurate, low jitter and efficient approach for clinical diagnosis and rehabilitation assessment. |
Year | DOI | Venue |
|---|---|---|
2020 | 10.1016/j.bspc.2019.101604 | Biomedical Signal Processing and Control |
Keywords | Field | DocType |
Ultrasound image,Skeletal muscle,Pennation angle,Gradient Radon Transform,Prosthesis control | Biomedical engineering,Muscle contraction,Pearson product-moment correlation coefficient,Torque,Pattern recognition,Polynomial regression,Artificial intelligence,Radon transform,Observational error,Mathematics,Ultrasound,Linear regression | Journal |
Volume | ISSN | Citations |
55 | 1746-8094 | 1 |
PageRank | References | Authors |
0.41 | 0 | 6 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Chenglang Yuan | 1 | 6 | 1.53 |
| Zengtong Chen | 2 | 1 | 0.41 |
| Mingyu Wang | 3 | 135 | 24.90 |
| Jianing Zhang | 4 | 1 | 0.41 |
| Sun Kun | 5 | 559 | 52.07 |
| Yongjin Zhou | 6 | 17 | 9.89 |