Related Experiment Video
Updated: Sep 19, 2025
15:48
Tracking the Mammary Architectural Features and Detecting Breast Cancer with Magnetic Resonance Diffusion Tensor Imaging
Published on: December 15, 2014
22.6K
Sizing It Up: Concordance between Breast Imaging and Pathologically Determined Tumor Measurement
Abigail E Daly1, Kyle J Anderman1, Logan R Holt1
1Breast Section, Division of Gastrointestinal and Oncologic Surgery, Massachusetts General Hospital, MGH Center for Breast Cancer, Boston, MA, USA.
Annals of Surgical Oncology
|June 16, 2025
Summary
Magnetic resonance imaging (MRI) offers the most accurate breast tumor size estimation, outperforming mammography and ultrasound. A combined MRI and ultrasound approach may further improve preoperative size assessment for better treatment planning.
Area of Science:
- Radiology
- Oncology
- Medical Imaging
Background:
- Accurate preoperative breast tumor size estimation is crucial for effective treatment planning.
- Discrepancies in size measurements can significantly alter treatment strategies.
- This study assesses the accuracy of mammography, ultrasound, and MRI against pathological findings.
Purpose of the Study:
- To evaluate the accuracy of tumor size estimation by mammography, ultrasound, and MRI compared to pathology.
- To identify factors influencing the performance of these imaging modalities.
- To develop an optimized formula for breast tumor size estimation.
Main Methods:
- Retrospective analysis of 460 breast cancer patients (2019-2024).
- Concordance defined as measurements within ±20% of pathological size.
- Statistical analyses included t-tests, chi-squared, Lin's CCC, logistic regression, and a Python-based weighted formula optimization.
Main Results:
- MRI showed the highest concordance (62%), followed by mammography (57%) and ultrasound (53%).
- An optimized weighted formula (0.66 × MRI size + 0.35 × US size) achieved 65.2% concordance.
- MRI demonstrated superior T-stage classification accuracy (89%) and performed better in larger, high-grade tumors, but less accurately in dense breasts and lobular histology.
Conclusions:
- MRI is the most accurate imaging modality for breast tumor size and T-stage estimation.
- Imaging accuracy is influenced by tumor characteristics (size, grade, histology) and breast density.
- A combined MRI and ultrasound approach shows promise for enhancing preoperative size estimation.
