Near Field Breast Tumor Detection Using Ultra-Narrow Band Probe with Machine Learning Techniques.

Area of Science:

• Biomedical Engineering
• Electromagnetics
• Medical Imaging

Background:

• Early detection of breast tumors is crucial for effective treatment.
• Current screening methods have limitations in accessibility and cost.
• Novel sensing modalities are needed for improved breast cancer diagnostics.

Purpose of the Study:

• To propose and validate a near-field microwave sensing modality for detecting breast tumors.
• To assess the sensitivity of a resonant probe to electromagnetic property changes in breast tissue.
• To develop a classification algorithm for tumor identification based on sensor response.

Main Methods:

• Utilizing a single near-field resonant probe with an ultra-narrow frequency response.
• Simultaneously sensing electromagnetic properties of both breasts to leverage symmetry.
• Employing Principle Component Analysis (PCA) to differentiate probe responses between breasts.
• Validating the concept using numerical simulations with realistic breast phantoms and experimental tests with metallic spheres in meat phantoms.

Main Results:

• The near-field microwave sensor demonstrated high sensitivity to changes in breast tissue's electromagnetic properties.
• Numerical simulations and experimental results successfully detected embedded tumors (metallic spheres) in phantoms.
• PCA effectively highlighted differences in probe responses, indicating the presence of tumors.
• The technique showed promise in distinguishing between healthy and tumor-affected tissue.

Conclusions:

• The proposed near-field microwave sensing modality is a viable method for breast tumor detection.
• This technique offers potential for inexpensive, portable, and regular breast cancer screening.
• Further development could lead to a non-invasive and accessible diagnostic tool for early tumor detection.