A method for vector displacement estimation with ultrasound imaging and its application for thyroid nodular disease

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Diagnostic Ultrasound

Background:

• Ultrasound elastography aids thyroid cancer diagnosis but requires advanced motion estimation due to complex tissue deformation.
• Accurate sub-pixel motion analysis is crucial for reliable elastographic measurements.

Purpose of the Study:

• To develop and evaluate a novel motion estimation method for ultrasound elastography.
• To compare the performance of the proposed method against classical block matching techniques.

Main Methods:

• Developed the Bilinear Deformable Block Matching (BDBM) method using an 8-parameter bilinear deformation model.
• Implemented an iterative multi-scale process and a triangle scan-based motion initialization technique.
• Tested BDBM on simulated ultrasound data, a dedicated elastography phantom, and clinical thyroid data acquired with freehand compression.

Main Results:

• The iterative approach in BDBM reduced displacement estimation error by a factor of 1.4 within two iterations.
• BDBM achieved a 4-fold reduction in absolute displacement error on simulated images compared to classical block matching.
• Local cross-correlation analysis confirmed BDBM's superior performance over block matching on experimental thyroid and phantom data.

Conclusions:

• The developed Bilinear Deformable Block Matching method enhances motion estimation accuracy in ultrasound elastography.
• BDBM shows significant potential for improving the diagnostic capabilities of ultrasound elastography, particularly for thyroid cancer detection.