Reviewed by Lexie CornerMay 22 2024
An international group of researchers from the UT Southwestern Medical Center developed a model that combined traditional MRI with artificial intelligence (AI) to help detect breast cancer metastases more accurately. Scientists also suggest that this model potentially eliminates the need for needle or surgical biopsies. This study was recently published in the journal Radiology: Imaging Cancer.
Basak Dogan, M.D., is a Professor of Radiology, Director of Breast Imaging Research, and a member of the Harold C. Simmons Comprehensive Cancer Center at UT Southwestern. Image Credit: UT Southwestern Medical Center
The non-invasive approach detects axillary metastasis (the presence of cancer cells in the lymph nodes beneath the arms) using conventional Magnetic Resonance Imaging (MRI) in conjunction with machine learning AI.
Most breast cancer deaths are due to metastatic disease, and the first site is usually an axillary lymph node. Determining nodal status is critical in guiding treatment decisions, but traditional imaging techniques alone do not have enough sensitivity to rule out axillary metastasis. That often requires patients to undergo invasive procedures that involve radioisotope and dye injection followed by surgery to remove and test whether the axillary nodes harbor cancer cells.
Basak Dogan M.D., Professor and Study Leader, Department of Radiology, University of Texas Southwestern Medical Center
Dogan is also a Director of Breast Imaging Research and a member of the Harold C. Simmons Comprehensive Cancer Center at UT Southwestern.
This study demonstrated that, compared to MRI or ultrasound, the AI model was noticeably more accurate in identifying individuals with axillary metastases. In clinical practice, the AI model successfully identified 95 % of patients with axillary metastasis while avoiding 51 % of benign (noncancerous) or needless surgical sentinel node biopsies.
That is an important advancement because surgical biopsies have side effects and risks. Despite having a low probability of a positive result confirming the presence of cancer cells, improving our ability to rule out axillary metastasis during a routine MRI–using this model–can reduce that risk while enhancing clinical outcomes.
Basak Dogan M.D., Professor and Study Leader, Department of Radiology, University of Texas Southwestern Medical Center
Around 350 newly diagnosed breast cancer patients from UT Southwestern and the Moody Center for Breast Health, which is situated on Parkland Health's main campus in Dallas, had their dynamic contrast-enhanced breast MRI tests used in the retrospective analysis, and everyone knew about the nodal condition.
The AI model was trained to recognize axillary metastasis using machine learning techniques, using photos and a variety of clinical indicators.
For many patients, the model can also relieve the anxiety and financial burden of unnecessary testing when it is utilized in conjunction with routine imaging scans.
Patients with benign findings from traditional MRI exams or needle biopsies are often subjected to sentinel lymph node biopsy because those tests can miss a significant proportion of metastasis. Our research demonstrates that it is possible to identify–with a high degree of accuracy–patients who are nonmetastatic, which benefits the patient and also allows the physician to tailor treatment.
Basak Dogan M.D., Professor and Study Leader, Department of Radiology, University of Texas Southwestern Medical Center
The study expands on earlier UT Southwestern research on breast cancer imaging and the creation of tools for metastatic prediction.
Dr. Dogan said, “Our study is a testament to UT Southwestern’s commitment to impactful research that addresses real-world healthcare challenges, the development and validation of AI models for medical imaging holds great promise in helping us in the fight against breast and other cancers, and this new tool is a significant step forward.”
To support their conclusions, researchers aim to incorporate a wider range of data and refine the picture analysis procedure.
Co-authors of the research, first author Dogan Polat, M.D., a second-year resident in Radiology; Albert Montillo, Ph.D., Assistant Professor in the Lyda Hill Department of Bioinformatics and Biomedical Engineering; Keith Hulsey, Ph.D., Instructor of Radiology; and Liqiang Wang, Ph.D., Faculty Associate, and Son Nguyen, Ph.D., postdoctoral researcher, in the Lyda Hill Department of Bioinformatics.
Dr. Dogan is also a Eugene P. Frenkel, M.D. Scholar in Clinical Medicine.
This research was funded by the Simmons Cancer Center, the National Institutes of Health (NIH), the National Institute of General Medical Sciences, the NIH National Institute of Aging, the NIH National Cancer Institute, the King Foundation, and the Lyda Hill Foundation.
Journal Reference:
Polat, D. S., et al. (2024) Machine Learning Prediction of Lymph Node Metastasis in Breast Cancer: Performance of a Multi-institutional MRI-based 4D Convolutional Neural Network. Radiology. Imaging Cancer. doi.org/10.1148/rycan.230107.