Search

Tendons are connective tissues that aids in body movements and appropriate functioning of musculoskeletal system. A tendon connects a bone to a muscle and acts as a mechanical link that transmits the contractile force generated by a muscle to bone aiding in locomotion. The tendons belong to the class of soft tissues, which usually undergo deformation under applied forces and re-forms, when forces are removed. Ultrasonography (a.k.a. Ultrasound) is one of the most commonly used imaging modality in clinical practice for detecting and analysing musculoskeletal injuries. The reasons for its wide usage in medical diagnosis are that it is non-invasive, cheap, portable and it doesn't involve harmful ionizing radiation for imaging. However, ultrasound images suffer from drawbacks of poor image quality due to image corruption by speckle noise and low contrast between different anatomical structures in an image. These drawbacks make visual interpretation of ultrasound image and delineation of different anatomical structures from ultrasound image, a more difficult and challenging task. Manual delineation of tendon and ligament from ultrasound images suffer from the following drawbacks: (i) It is a time-consuming task (ii) Subjective in nature (inter-subject and intra-subject variability). These limitations of manual delineation necessitate the development of an automated delineation tool so that the delineation becomes more accurate and less time consuming. This paper provides current state-of-art techniques for segmenting a tendon from an ultrasound images. Comprehensive explanation of each methodology is provided in this paper along with advantages and limitations. Then appropriate methodology for tendon segmentation is identified and implemented. Accuracy of the segmentation is calculated by validating it with the manually segmented ultrasound images from experts.