Delta Radiomics Based on Longitudinal Dual-modal Ultrasound Can Early Predict Response to Neoadjuvant Chemotherapy in...

Area of Science:

• Medical Imaging
• Oncology
• Radiomics

Background:

• Neoadjuvant chemotherapy (NAC) is crucial for breast cancer treatment.
• Early prediction of pathological response to NAC is vital for treatment optimization.
• Current monitoring methods may lack sufficient sensitivity for early response assessment.

Purpose of the Study:

• To develop and validate a monitoring model using radiomics analysis.
• To integrate longitudinal B-mode ultrasound (BUS) and shear wave elastography (SWE) data.
• To achieve early prediction of pathological response to NAC in breast cancer patients.

Main Methods:

• Prospective study including 112 breast cancer patients receiving NAC.
• Acquisition of BUS and SWE data pre-treatment and after 2 and 4 cycles of NAC.
• Extraction of radiomics features and calculation of delta radiomics (△R) changes; support vector machine classification.

Main Results:

• △R derived from SWE showed superior performance (AUC 0.88-0.90) compared to BUS (AUC 0.83-0.84) for NAC response prediction.
• The longitudinal dual-modal ultrasound radiomics (LDUR) model demonstrated excellent discrimination (AUC 0.97), sensitivity (95.52%), and specificity (91.11%).
• No significant performance difference was found between early (C2) and later (C4) assessments.

Conclusions:

• The LDUR model offers a highly effective tool for early prediction of pathological response to NAC.
• This dual-modal ultrasound radiomics approach enhances monitoring capabilities during breast cancer chemotherapy.
• The findings support the clinical utility of LDUR for personalized treatment strategies in breast cancer.