Level set contouring for breast tumor in sonography.

Area of Science:

• Medical Imaging
• Oncology
• Computer-Aided Diagnosis

Background:

• Sonographic features like echogenicity and texture aid in differentiating benign from malignant breast tumors.
• Accurate tumor contouring is crucial for diagnosis, but can be time-consuming and requires expertise.

Purpose of the Study:

• To develop an efficient and automatic method for precise breast tumor contour detection in ultrasound (US) images.
• To assist physicians, especially those with less experience, in making accurate diagnoses through automated contouring.

Main Methods:

• A sophisticated preprocessing filter was used to reduce noise while preserving tumor shape and contrast.
• An adaptive initial contouring method generated an approximate circular tumor contour.
• Deformation-based level set segmentation was employed for automatic extraction of precise tumor contours.

Main Results:

• The method was evaluated on ultrasound images from 118 patients with breast tumors.
• Computer simulations demonstrated that the automated contours closely matched manually sketched contours.
• The proposed method proved to be robust and fast for breast ultrasound image contouring.

Conclusions:

• The developed automatic contouring method is efficient, stable, and accurate for breast tumors in ultrasound images.
• This approach has the potential to significantly reduce the time physicians spend on manual contour sketching.
• Automated contouring can enhance diagnostic consistency and support less experienced clinicians.