Longitudinal Enhancement of the Hyperechoic Regions in Ultrasonography of Muscles Using a Gabor Filter Bank Approach: A Preparation for Semi-Automatic Muscle Fiber Orientation Estimation

• Published in: Ultrasound in medicine & biology Vol. 37; no. 4; pp. 665 - 673
• Main Authors: Zhou, Yongjin, Zheng, Yong-Ping
• Format: Journal Article
• Language: English
• Published: New York, NY Elsevier Inc 01.04.2011; Elsevier
• Subjects: Adult; adults; Algorithms; Biological and medical sciences; e-mail; Gabor filtering; Humans; Image enhancement; Image Enhancement - methods; Image Interpretation, Computer-Assisted - methods; Image segmentation; Investigative techniques, diagnostic techniques (general aspects); Medical sciences; Muscle; muscle fibers; Muscle Fibers, Skeletal - diagnostic imaging; muscles; Osteoarticular system. Muscles; Pattern Recognition, Automated - methods; Radiodiagnosis. Nmr imagery. Nmr spectrometry; Radiology; Reproducibility of Results; Sensitivity and Specificity; skeletal muscle; Soft tissue; Sonogram; Sonomyography; ultrasonography; Ultrasonography - methods; Ultrasound; Image segmentation; Sonogram; Gabor filtering; Muscle; Soft tissue; Sonomyography; Ultrasound; Image enhancement; Sonography; Filtering; Image processing; Muscular fiber; Striated muscle; Biceps muscle; Echography; Upper limb; Fiber orientation; Technique; Semiautomatic method
• ISSN: 0301-5629; 1879-291X; 1879-291X
• DOI: 10.1016/j.ultrasmedbio.2010.12.011

In this study, to complement our previously proposed method for estimating muscle fiber orientation, the Gabor filter bank (GF) technique was applied to sonograms of the biceps and forearm muscles to longitudinally enhance the coherently oriented and hyperechoic perimysiums regions. The method involved three steps: orientation field estimation, frequency map computation and Gabor filtering. The method was evaluated using a simulated image distorted with multiplicative speckle noises where the “muscles” were arranged in a bipennate fashion with an “aponeurosis” located in the middle. After enhancement using the GF approach, most of the original hyperechoic bands in the simulated image could be recovered. The proposed method was also tested using a group of biceps and forearm muscle sonograms collected from healthy adult subjects. Compared with the sonograms without enhancement, the enhanced images led to the detection of more linear patterns including muscle fascicles and smaller angle differences compared with the mean of manual results from two operators, therefore, were better prepared for the automatic estimation of muscle fiber orientation. The proposed method has the potential of assisting in the visualization of strongly oriented patterns in skeletal muscle sonograms as well as in the semi-automatic estimation of muscle fiber orientations. (E-mail: yongjin.zhou@inet.polyu.edu.hk)