Applied Analysis Seminar - Abstracts


Talk

  • Thursday, 18 November 2021, 13:00-14:00 in HG01.060


  • Gijs Hendriks (Radboud UMC)
    Developments in ultrasound imaging in breast cancer detection

    Abstract

    Ultrasound imaging is a common imaging method to detect, classify and monitor breast cancers. 2D hand held ultrasound imaging and Doppler are often used to locally evaluate palpable breast lesions or lesions that were detected by other imaging modalities. For breast cancer screening, automated breast volume scanners (ABVS) are introduced in which a large ultrasound transducer (12cm) translates automatically over the breast while collecting images. Those scanners are especially suitable for screening since ultrasound images are obtained automatically of the entire breast and the Radiologist can evaluate the image afterwards. Although the sensitivity of ABVS is high, the specificity is limited resulting in high recall rate and unnecessary biopsies. Another limitation is the long scan time to obtain one volume (90 seconds) resulting in breathing artifacts. In this talk, I want to present our work in solving those two limitations. We developed 3D elastography for the ABVS to improve the specificity by obtaining mechanical properties of breast lesions next to the regular ultrasound images. In ultrafast imaging, the frame rate can be increased to reduce scan times by modifying the acquisition scheme. We developed and improved ultrasound reconstruction methods in ultrafast imaging to preserve and even improve image quality compared to conventional ultrasound imaging.


    (back to Applied Analysis Seminar homepage)

    Department of Mathematics
    Radboud University Nijmegen

    Heyendaalseweg 135
    6525 AJ Nijmegen