FEATURE DRIFT RESILIENT TRACKING OF THE CAROTID ARTERY WALL USING UNSCENTED KALMAN FILTERING WITH DATA FUSION

被引：0

作者：

Dorazil, Jan ^{[1
]}

Repp, Rene ^{[1
]}

Kropfreiter, Thomas ^{[1
]}

Prueller, Richard ^{[1
]}

Riha, Kamil ^{[2
]}

Hlawatsch, Franz ^{[1
]}

机构：

[1] TU Wien, Inst Telecommun, Vienna, Austria

[2] Brno Univ Technol, Dept Telecommun, Brno, Czech Republic

来源：

2020 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING | 2020年

关键词：

Atherosclerosis; common carotid artery; B-mode ultrasound; unscented Kalman filter; data fusion; MOTION;

D O I：

10.1109/icassp40776.2020.9054703

中图分类号：

O42 [声学];

学科分类号：

070206 ; 082403 ;

摘要：

An analysis of the motion of the common carotid artery (CCA) provides effective indicators for cardiovascular diseases. Here, we propose a method for tracking CCA wall motion from a B-mode ultrasound video sequence. An unscented Kalman filter based on a suitably devised state-space model fuses measurements produced by an optical flow algorithm and a CCA wall localization algorithm. This approach compensates for feature drift, which is a detrimental effect in optical flow algorithms. The proposed method is demonstrated to outperform a state-of-the-art tracking method based on optical flow.

引用

页码：1095 / 1099

页数：5

共 26 条

[1]

[Anonymous], 2022, Pattern Recognition

[2]

[Anonymous], 2001, Tech. Rep.

[3]

Challa S., 2011, Fundamentals of Object Tracking

[4] ULTRASOUND TISSUE DISPLACEMENT IMAGING WITH APPLICATION TO BREAST-CANCER [J].

CHEN, EJ ;

ADLER, RS ;

CARSON, PL ;

JENKINS, WK ;

OBRIEN, WD .

ULTRASOUND IN MEDICINE AND BIOLOGY, 1995, 21 (09) :1153-1162

[5]

Dorazil J., 2019, P EUSIPCO 2019 COR P, P1

[6]

Dorazil J, 2019, 2019 42ND INTERNATIONAL CONFERENCE ON TELECOMMUNICATIONS AND SIGNAL PROCESSING (TSP), P605, DOI [10.1109/TSP.2019.8769077, 10.1109/tsp.2019.8769077]

[7] Motion Tracking of the Carotid Artery Wall From Ultrasound Image Sequences: a Nonlinear State-Space Approach [J].