Breast tumor parameter estimation and interactive 3D thermal tomography using discrete thermal sensor data.

Area of Science:

• Biomedical Engineering
• Medical Imaging
• Oncology

Background:

• Breast cancer diagnosis relies on various imaging modalities.
• Accurate tumor localization and characterization are crucial for effective treatment planning.
• Non-invasive, cost-effective screening tools are needed to improve early detection rates.

Purpose of the Study:

• To develop a low-cost wearable device for breast thermography.
• To create a methodology for estimating tumor parameters (diameter, blood perfusion, metabolic heat generation, location) from thermal data.
• To validate the system's efficacy in phantom and clinical settings for breast cancer detection.

Main Methods:

• Utilized a wearable device with discrete thermal sensors to map breast skin surface temperature.
• Developed an interactive 3D thermal tomography system for visualizing breast thermal anatomy.
• Employed Finite Element Method (FEM) and an evolution-based inverse method for parameter estimation.
• Validated the methodology using phantom experiments and clinical trials on 60 subjects.

Main Results:

• Phantom experiments showed <10% error for heat generation and <15% error for location.
• Clinical trials demonstrated distinguishable differences in blood perfusion and metabolic heat generation between cancerous and non-cancerous breasts.
• Tumor diameter and location estimations showed good agreement with clinical reports.
• Achieved a sensitivity of 82.78% and specificity of 87.09%.

Conclusions:

• The proposed breast tumor parameter estimation methodology with 3D thermal tomography is an effective screening tool for breast cancer detection.
• The system provides clinicians with valuable information on tumor location, including depth.
• This low-cost, non-invasive approach can aid in early breast cancer diagnosis and management.