Related Experiment Video
Updated: Aug 5, 2026
11:49
Single-port Non-liposuction Endoscopic Axillary Lymph Node Dissection in Breast Cancer Surgery
Published on: April 3, 2026
Machine Learning Classification of Axillary Lymph Nodes Using Microwave Signals
Daniela M Godinho1, João M Felício2, Carlos A Fernandes2
1Instituto de Biofísica e Engenharia Biomédica, Faculdade de Ciências, Universidade de Lisboa, Campo Grande, 1749-016 Lisbon, Portugal.
Sensors (Basel, Switzerland)
|July 28, 2026
Summary
This study introduces a novel microwave signal classification method for breast cancer staging. It accurately differentiates healthy and metastasised axillary lymph nodes (ALNs) using signal geometry, achieving up to 95% accuracy.
Area of Science:
- Biomedical Engineering
- Medical Imaging
- Electromagnetics
Background:
- Axillary lymph node (ALN) status is critical for breast cancer staging.
- Conventional imaging modalities have limitations in ALN assessment.
- Microwave imaging (MWI) offers a promising alternative for ALN evaluation.
Purpose of the Study:
- To investigate the direct classification of ALNs and axillary regions using microwave signals, bypassing image reconstruction.
- To assess the feasibility of differentiating healthy and metastasised ALNs based on geometric differences in microwave signals.
- To explore the potential of signal-based classification for breast cancer staging.
Main Methods:
- Generated 80 numerical ALN models based on anatomical descriptions.
- Simulated microwave signals for single and dual ALN scenarios (healthy vs. metastasised).
- Evaluated various signal types, feature extraction methods, and classifiers (kNN, SVM) under realistic conditions, including limited angular views.
Main Results:
- Achieved 95% classification accuracy for single-ALN scenarios using kNN.
- Reached up to 83.3% accuracy for complex two-ALN scenarios using SVM.
- Demonstrated successful differentiation of healthy and metastasised ALNs based on microwave signal characteristics.
Conclusions:
- Microwave signal-based classification shows significant potential for differentiating healthy and metastasised ALNs.
- This approach can support breast cancer staging without requiring image reconstruction.
- Future integration with MWI image interpretation could enhance ALN assessment.
