Advanced Ultrasound in Diagnosis and Therapy ›› 2020, Vol. 4 ›› Issue (3): 183-188.doi: 10.37015/AUDT.2020.190024

• Original Research • Previous Articles     Next Articles

Quantitative Analysis of Textural Features Extracted from Sonograms of Biceps under Different Physiological States

Lanting Jia, MDa, Jiaqi Zhao, MDa,*(), Qi Xu, PhDb, Qian Pan, MDa, Jianquan Zhang, MDa   

  1. a Department of Ultrasound, Changzheng Hospital, Second Military Medical University, Shanghai, China
    b Department of Computer Science, Institute of Information Engineering, Shanghai Maritime University, Shanghai, China
  • Received:2019-09-19 Online:2020-09-30 Published:2020-08-21
  • Contact: Jiaqi Zhao, MD, E-mail:qiqiblue67@163.com

Abstract:

Objective: To quantitatively analyze the difference of texture features of skeletal muscle in high-frequency ultrasound images under different physiological states using the multiscale decomposition method of ultrasound echo intensity interface reflections.
Methods: High frequency ultrasound images of the biceps brachii in different physiological states were collected from 20 healthy volunteers. In offline state, eight texture parameters including mean of texture gray scale (Mean), standard variance (SDev) of gray scale, number of blobs (NOB) of texture density, irregularity (IRGL) of texture primitive shape, mean size of blobs (SOB) of texture primitive, homogeneity of distribution (HOD) of texture uniformity, directionality of texture distribution (DOD), and periodicity of texture distribution (POD) were extracted by MATLAB software and compared and analyzed statistically.
Results: With the use of high frequency ultrasound, all healthy volunteers' biceps brachii showed isoechoic muscle bundles, organized arrangement of muscle fibers, and distinct and intact texture of structure. In different physiological states of biceps brachii of the same gender group, the NOB and the Mean showed statistically differences (P < 0.05). In the relaxation state of biceps brachii between different gender groups, the average SOB and the DOD showed statistically differences (P < 0.05). In the load state of biceps brachii between different genders, the NOB and the Mean showed statistically differences (P < 0.05).
Conclusions: The ultrasonic image changes of muscle fibers under different physiological states can be identified by quantitative texture characteristic parameters, providing more information for clinical computer-aided diagnosis of skeletal muscle injury.

Key words: Muscle injury; Ultrasonography; Computer-aided diagnosis; Biceps Texture; Quantitative analysis