Displacement Field Estimation for Echocardiography Strain Imaging Using B-Spline Based Elastic Image Registration—Synthetic Data Study

• Published in: Recent Global Research and Education: Technological Challenges Vol. 519; pp. 309 - 315
• Main Authors: Wilczewska, Aleksandra, Cygan, Szymon, Żmigrodzki, Jakub
• Format: Book Chapter
• Language: English
• Published: Switzerland Springer International Publishing AG 2016; Springer International Publishing
• Series: Advances in Intelligent Systems and Computing
• Subjects: Artificial intelligence; Cybernetics & systems theory; Elastic image registration; Strain imaging
• ISBN: 9783319464893; 3319464892
• ISSN: 2194-5357; 2194-5365
• DOI: 10.1007/978-3-319-46490-9_42

Strain imaging in echocardiography allows identification of myocardium dysfunctions. This paper describes the use of own implementation of elastic image registration, to calculate displacement field in two-dimensional echocardiographic data. Performance of the algorithm was examined on synthetic ultrasonic data. A series of tests examining the influence of algorithm parameters on the outcome was conducted. Displacement fields were compared with reference data from the finite element model used for generation of the synthetic data. Quality of image registration was evaluated using two error measures: mean absolute error and median absolute error. The displacement field errors obtained in the direction transverse to the ultrasound wave had an order of magnitude 10−5 m and errors in the direction along ultrasound wave: 10−6 m, which is close to the accuracy of two state-of-the-art methods for displacements estimation tested on the same input data.