A multimodal MRI radiomics and deep learning model outperformed traditional models in predicting 5- and 7-year survival for breast cancer patients receiving neoadjuvant chemotherapy.
Key Details
- 1Study involved 216 women with breast cancer post-neoadjuvant chemotherapy.
- 2Model integrated MRI radiomics, pathology, and clinical data using deep learning.
- 3The deep feature-based patho-radiomic model achieved AUCs of 0.89 (training) and 0.82 (test) for 5-year survival, and 0.91 (training) and 0.87 (test) for 7-year survival.
- 4Clinical-only models showed lower AUCs (0.4–0.53 for 5-years; 0.45–0.53 for 7-years).
- 5Traditional clinical and molecular markers (ER, HER2, TNBC) did not significantly predict survival in this cohort.
- 6Authors advocate for prospective studies to guide clinical decisions using the model.
Why It Matters
This research demonstrates the power of combining imaging, pathology, and clinical data through AI to provide more accurate long-term prognostic tools in oncology, surpassing traditional clinical or single-modality models. Improved risk stratification could optimize therapeutic decision-making and patient counseling in breast cancer care.
Source
AuntMinnie
Related News
•Radiology Business
Most Radiology AI Users Lack Clear Evidence of Financial ROI
Survey finds over 75% of radiology organizations using AI lack clear, quantified ROI data.
•AuntMinnie
Paul Chang Discusses Foundation Models and Agentic AI at RSNA 2025
Dr. Paul Chang shares his insights on the role of foundation models and agentic AI in radiology at RSNA 2025.
•AuntMinnie
Toronto Study: LLMs Must Cite Sources for Radiology Decision Support
University of Toronto researchers found that large language models (LLMs) such as DeepSeek V3 and GPT-4o offer promising support for radiology decision-making in pancreatic cancer when their recommendations cite guideline sources.