A Simple Ultrasound Based Classification Algorithm Allows Differentiation of Benign from Malignant Breast Lesions by...

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Oncology

Background:

• Distinguishing benign from malignant breast lesions is crucial for appropriate patient management.
• Ultrasound (US) elastography and Doppler US offer quantitative parameters for lesion characterization.
• Developing objective criteria can improve diagnostic accuracy and reduce unnecessary invasive procedures.

Purpose of the Study:

• To develop a simple classification algorithm using quantitative ultrasound parameters.
• To differentiate benign from malignant breast lesions.
• To aid in the decision-making process for breast biopsies.

Main Methods:

• Prospective study of 124 patients with biopsy-proven breast lesions.
• Quantitative analysis using B-mode US, Color/Power Doppler US, and Acoustic Radiation Force Impulse (ARFI) elastography.
• Classification algorithm developed using exhaustive chi-squared automatic interaction detection.

Main Results:

• The classification algorithm incorporated maximum shear wave velocity (SWVmax) and resistive index (RI).
• The algorithm achieved an Area Under the Curve (AUC) of 0.887, with 98.46% sensitivity and 61.02% specificity.
• Application of the algorithm could have avoided 61% of false-positive biopsies.

Conclusions:

• A classification algorithm combining SWVmax and RI demonstrates high diagnostic performance.
• This objective approach accurately differentiates benign from malignant breast lesions.
• The algorithm has the potential to significantly reduce unnecessary breast biopsies.