Multi-modality morphological correlation of axillary lymph nodes

Area of Science:

• Medical Imaging
• Oncology
• Radiology

Background:

• Accurate staging of breast cancer is crucial for effective treatment planning.
• Axillary lymph node (ALN) status is a key prognostic factor.
• Multi-modality imaging offers potential for improved non-invasive staging.

Purpose of the Study:

• To develop a computerized method for correlating axillary lymph nodes (ALNs) between ultrasound (US) and breast MRI (BMRI).
• To establish a preoperative, non-invasive method for breast cancer staging using multi-modality imaging.
• To improve the accuracy of ALN evaluation in breast cancer patients.

Main Methods:

• Utilized ultrasound (US) and breast MRI (BMRI) scans from 20 breast cancer patients.
• Employed an ellipse fitting algorithm to segment ALN cortex and hilum for measurement of axes and maximum cortical thickness (MCT).
• Computed Euclidean distances between US ALNs and candidate BMRI ALNs to determine the most likely correlation.

Main Results:

• The computerized method successfully identified the same ALN as expert radiologists in 81.25% of cases (13 out of 16 ALNs).
• Maximum cortical thickness (MCT) measurements showed small standard deviations and were comparable between US and MRI.
• No significant differences were found in long axis, short axis, or MCT measurements between US and MRI-T2 weighted sequences.

Conclusions:

• A feasibility study demonstrated that computerized measurements can correlate ALNs across different imaging modalities (US and BMRI).
• This correlation enables the combined use of imaging parameters for ALN evaluation in breast cancer.
• The multi-modality approach promises a more robust non-invasive method for axillary staging.