LG177: Ultrasound Image Enhancement for Real-time Automatic Detection of the Lumber Anatomy in Sono-Guided Epidural Needle Insertions

Date: 
Friday, January 31, 2014 - 12:00
Speaker: 
Sima Sobhie, MS
Abstract: 
In many difficult surgical operations in which the stomach or the area underneath it is cut open either partially or completely an epidural anesthesia is performed to reduce the patients’ pain after surgery. Proper performance of the epidural needle insertion is substantially important since an epidural is injected in the epidural space whose interior border is the Dura mater which supports the spinal cord. The conventional approach for epidural anesthesia is blind as the anesthesiologist proceeds only by his sense of resistance-loss. Recently, using ultrasound images as a visual guidance is proposed in order to aid anesthesiologists. However, since sterile conditions should be preserved and anesthesiologists are inexperienced in ultrasound interpretation, ultrasound image processing is required for automatic detection of the lumbar anatomy and insertion depth measurement. The objective of this research is to enhance lumbar ultrasound images and detect the ligamentum flavum at the end of a laminar bone in order to offer real-time help during epidural anesthesia. This research proposes a novel approach for ultrasound despeckling based on a single image. The proposed method applies complex wavelets to the anisotropic diffusion noise reduction method. Furthermore, real-time anatomy detection is performed by applying segmentation methods. The proposed algorithm and the SRAD algorithm are performed on synthesis and real ultrasound images. The results of automatic detection are compared with the measurement performed by the radiologist using a 3.5 MHz curvilinear probe and the real needle insertion depth measured by the anesthesiologist during the epidural injection of 10 patients at the Shohada Tajrish hospital. Results indicate that ultrasound images provide anesthesiologists with valuable information about the place and depth of the needle insertion and are proven to be highly beneficial during the first stage of the needle insertion. However, at the final stage when the needle approaches the epidural space, anesthesiologists should rely on the loss-of-resistance method due to inevitable errors of the sono-guided method.

Sima Sobhiyeh is currently Research Assistant at the AMIR Lab. in the Biomedical Engineering department at the Amirkabir University of Technology. She received her B.Sc. and M.Sc. degrees in Digital Electronics from the Amirkabir University of Technology in 2011 and 2013, respectively. Her main research interest is in the field of Medical imaging and Image Processing.