Quantification of thyroid volume using 3-D ultrasound imaging

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Radiology

Background:

• Ultrasound (US) is a widely used, non-invasive diagnostic tool, valued for its speed, cost-effectiveness, and lack of ionizing radiation.
• Current 2-D US methods for thyroid imaging may lack accuracy in volumetric assessment due to the absence of 3-D data.
• Accurate thyroid gland and nodule volume measurement is crucial for diagnosis and treatment monitoring.

Purpose of the Study:

• To develop and evaluate a semi-automatic segmentation technique for analyzing 3-D ultrasound data of the thyroid gland.
• To improve the accuracy of volumetric quantification of thyroid gland lobes and potentially other organs.
• To address the limitations of conventional 2-D US in thyroid imaging.

Main Methods:

• Acquisition and pre-processing of 3-D ultrasound images of the thyroid gland.
• Implementation of a semi-automatic segmentation approach using an extended geodesic active contour level set formulation.
• Validation of the segmentation results against manual segmentations performed by a medical expert on five patient cases.

Main Results:

• The proposed semi-automatic segmentation framework enables accurate volumetric quantification of thyroid gland lobes from 3-D US data.
• The method demonstrated comparable results to expert manual segmentation, highlighting its clinical potential.
• The framework provides a more comprehensive analysis than traditional 2-D ultrasound techniques.

Conclusions:

• The developed semi-automatic 3-D US segmentation method offers a novel and accurate approach for thyroid volumetric quantification.
• This technique may be extendable to other parenchymatous organs, broadening its clinical applicability.
• The study highlights the advantages of 3-D US over 2-D US for precise thyroid gland assessment.