King's College London

Research portal

A Tensor-based Catheter and Wire Detection and Tracking Framework and Its Clinical Applications

Research output: Contribution to journalArticlepeer-review

Yingliang Ma, Diwei Zhou, Lei Ye, James Housden, Ansab Fazili, Kawal Rhode

Original languageEnglish
Number of pages10
JournalIEEE Transactions on Biomedical Engineering
Accepted/In press2 Aug 2021

Documents

  • IEEEBE_TensorCathTracking_v8_Final

    IEEEBE_TensorCathTracking_v8_Final.pdf, 5.18 MB, application/pdf

    Uploaded date:03 Aug 2021

    Version:Accepted author manuscript

    Licence:CC BY

King's Authors

Abstract

Catheters and wires are used extensively in cardiac catheterization procedures. Detecting their positions in fluoroscopic X-ray images is important for several clinical applications such as motion compensation, co-registration between 2D and 3D imaging modalities and X-ray dose control using collimation. Detecting the complete length of a catheter or wire object is a challenging task as detection might be distracted by nearby interventional instruments or the target object might overlap with other wire-like objects. The majority of existing detection algorithms rely on manual initialization or pre-defined models. In this paper, a novel and fully automatic detection framework for catheters and wires is developed. The framework is based on path reconstruction from image tensors, which are Eigen direction vectors generated from a multiscale vessel enhancement filter. A catheter or a wire object is detected as the smooth path along those Eigen direction vectors. Furthermore, a real-time tracking method based on a template generated from the detection method was developed. The proposed framework was tested on a total of 8,364 images which are from 98 image sequences acquired from 39 clinical cases. Detection errors for catheters and guidewires are 0.56 ± 0.28 mm and 0.66 ± 0.32 mm, respectively, and success rates of 91.4% and 86.3% were achieved. Finally, the proposed framework was tested and validated in two clinical applications. For motion compensation using catheter tracking, the 2D target registration errors (TRE) at the pulmonary veins were calculated to validate the method and a TRE of 1.8 mm ± 0.9 mm was achieved. For co-registration between 2D X-ray images and 3D models from MRI images, a TRE of 2.3 ± 0.9 mm was achieved.

Download statistics

No data available

View graph of relations

© 2020 King's College London | Strand | London WC2R 2LS | England | United Kingdom | Tel +44 (0)20 7836 5454