Application of the no-sampling linear sampling method to breast cancer detection

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Electromagnetics

Background:

• Microwave imaging offers a promising non-ionizing radiation technique for breast cancer detection.
• Accurate modeling of breast tissue is crucial for effective microwave imaging.
• Magnetic Resonance Imaging (MRI) provides high-resolution anatomical data for realistic breast models.

Purpose of the Study:

• To apply and assess a no-sampling linear sampling method for microwave imaging of cancerous breast tissues.
• To evaluate the method's sensitivity to tumor anomalies compared to healthy breast tissue.
• To investigate the method's performance in noisy environments and its robustness to parameter uncertainties.

Main Methods:

• Utilized a no-sampling linear sampling algorithm for microwave imaging.
• Employed realistic female breast models derived from MRI data.
• Formulated the imaging problem using a modified far-field equation sensitive to anomalies.
• Incorporated the healthy breast configuration via a numerically computed Green's function.
• Assessed performance against tumors of varying positions and sizes in simulated noisy environments.
• Conducted a numerical analysis of the method's robustness to uncertainties in model parameters.

Main Results:

• The no-sampling linear sampling method demonstrated sensitivity to anomalies indicative of cancerous tissues.
• The imaging technique successfully inspected tumors of several positions and sizes within the realistic breast models.
• Performance was evaluated under simulated noisy environmental conditions.
• Numerical analysis confirmed the method's robustness even when model parameters for Green's function evaluation were not precisely known.

Conclusions:

• The no-sampling linear sampling method is a viable technique for microwave imaging of breast cancer.
• The approach shows potential for detecting tumors in realistic breast models, even with environmental noise and parameter uncertainties.
• Further validation in clinical settings is warranted to establish its efficacy in real-world breast cancer diagnosis.