A contour extraction method using active contour model on ultrasonic images.

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Surgical Navigation

Background:

• Accurate 3D models are crucial for breast tumor excision, aiding in precise localization and excision area determination.
• Speckle noise in ultrasonic images degrades resolution, hindering the creation of recognizable 3D models from 2D tomograms.
• Effective speckle reduction and region of interest (ROI) contour extraction are necessary for robust 3D reconstruction from ultrasound data.

Purpose of the Study:

• To develop and evaluate a method for precise contour extraction of the region of interest (ROI) in ultrasonic images.
• To address the challenge of speckle noise in ultrasound images for improved 3D model reconstruction.

Main Methods:

• An active contour model incorporating gradient vector flow was utilized for contour extraction.
• Speckle reduction techniques were applied to ultrasonic images prior to contour extraction.

Main Results:

• The study successfully extracted the contour of the lesion area from pre-processed ultrasonic images with reduced speckle noise.
• The employed active contour model effectively delineated the ROI boundaries.

Conclusions:

• The proposed method demonstrates efficacy in extracting lesion contours from speckle-reduced ultrasonic images.
• This technique contributes to improved 3D model reconstruction, supporting enhanced surgical navigation during breast tumor excision.