Visualization of bone structures on lumbar ultrasound images

Abstract

Bone visualization in ultrasound images is required in many clinical applications such as Computer-Assisted Orthopedic Surgery (CAOS), epidural, and fracture identification. Presently, X-ray, Computed Tomography (CT), Magnetic Resonance Imaging (MRI), and fluoroscopy imaging modalities are used to visualize bones. All these techniques involve exposure to radiation and are expensive. During childbirth, surgery, and pain relief management, the epidural procedure is involved. Currently, manual palpation is used to insert the needle for epidural anesthetic procedures. Incorrect needle insertion may cause nerve damage, chronic pain, paralysis, and death. To avoid these complications, image based guidance is required for needle insertion. newlineThus, it is necessary to introduce a safer alternative imaging modality than existing ones to identify bone surfaces. Ultrasound imaging is noninvasive, nonionizing, economical, and most reliable. Hence, the ultrasound technique is suggested to identify bone surfaces. Since this technique has noisy and poor image quality, this work focused on developing image processing techniques toward image enhancement and segmentation of bone surfaces. In order to explore the clinical applications of ultrasound imaging, the epidural was selected as the most suitable approach. However, ultrasound provides a low contrast image and contains speckles, which degrade the picture quality. Hence, it is necessary to develop robust and accurate bone segmentation. newline newline