Cross-modal deep learning model for predicting pathologic complete response to neoadjuvant chemotherapy in breast...

Area of Science:

• Oncology
• Medical Imaging
• Artificial Intelligence

Background:

• Pathological complete response (pCR) is a key indicator for neoadjuvant chemotherapy (NAC) effectiveness in breast cancer.
• Accurate early prediction of pCR is crucial for tailoring treatment decisions and improving patient outcomes.
• Artificial intelligence (AI) offers promising avenues for enhancing the precision and timeliness of pCR prediction.

Purpose of the Study:

• To develop and validate a novel cross-modal, multi-pathway AI model for early prediction of pCR in breast cancer.
• To integrate multi-temporal ultrasound (US) imaging and digital pathology data for enhanced predictive accuracy.
• To establish a foundation for personalized breast cancer treatment strategies based on predicted NAC response.

Main Methods:

• A cross-modal, multi-pathway automated prediction model was designed.
• The model integrates temporal and spatial information from digital pathology images and multi-temporal ultrasound (US) images.
• The model was evaluated for its efficacy in predicting pCR status during NAC.

Main Results:

• The proposed AI model demonstrated exceptional predictive efficacy for pCR.
• The fusion of digital pathology and multi-temporal US images significantly improved early prediction accuracy.
• The model effectively utilizes both temporal dynamics and spatial features for prediction.

Conclusions:

• The developed AI model shows significant potential as an auxiliary tool for early NAC response prediction in breast cancer.
• This approach supports the development of personalized treatment paradigms tailored to individual patient responses.
• Integrating diverse imaging modalities with AI can advance precision medicine in breast cancer care.