Login / Signup

Accurate Needle Localization Using Two-Dimensional Power Doppler and B-Mode Ultrasound Image Analyses: A Feasibility Study.

Mohammad I DaoudAhmad ShtaiyatAdnan R ZayadeenRami Alazrai
Published in: Sensors (Basel, Switzerland) (2018)
Curvilinear ultrasound transducers are commonly used in various needle insertion interventions, but localizing the needle in curvilinear ultrasound images is usually challenging. In this paper, a new method is proposed to localize the needle in curvilinear ultrasound images by exciting the needle using a piezoelectric buzzer and imaging the excited needle using a curvilinear ultrasound transducer to acquire a power Doppler image and a B-mode image. The needle-induced Doppler responses that appear in the power Doppler image are analyzed to estimate the needle axis initially and identify the candidate regions that are expected to include the needle. The candidate needle regions in the B-mode image are analyzed to improve the localization of the needle axis. The needle tip is determined by analyzing the intensity variations of the power Doppler and B-mode images around the needle axis. The proposed method is employed to localize different needles that are inserted in three ex vivo animal tissue types at various insertion angles, and the results demonstrate the capability of the method to achieve automatic, reliable and accurate needle localization. Furthermore, the proposed method outperformed two existing needle localization methods.
Keyphrases
  • ultrasound guided
  • deep learning
  • magnetic resonance imaging
  • high resolution
  • physical activity
  • endothelial cells
  • quantum dots