An Innovative Thermal Imaging Prototype for Precise Breast Cancer Detection: Integrating Compression Techniques and...

Area of Science:

• Biomedical Engineering
• Medical Imaging
• Oncology

Background:

• Early breast cancer detection significantly improves patient survival rates.
• Current diagnostic methods may involve radiation exposure.
• Need for advanced, non-invasive imaging techniques.

Purpose of the Study:

• To develop and validate an innovative, radiation-free thermal imaging prototype for early breast cancer detection.
• To enhance thermal image quality and classification accuracy using compression and illumination techniques.
• To assess the prototype's efficacy in distinguishing cancerous from normal breast tissue.

Main Methods:

• Design and 3D modeling of a thermal imaging prototype incorporating compression mechanisms and controlled illumination.
• Acquisition of thermal images from experimental datasets including cancerous and normal samples.
• Application of image preprocessing and segmentation techniques.
• Classification of thermal images using various machine learning models, including logistic regression.

Main Results:

• The prototype integrates essential components like pressure sensors, motors, and a thermal camera.
• Image quality and classification accuracy were improved through integrated compression and illumination.
• The logistic regression model achieved high performance: 0.976 accuracy, 0.977 F1 score, 1.000 precision, and 0.995 recall.
• Demonstrated high accuracy in classifying thermal abnormalities indicative of breast cancer.

Conclusions:

• The developed thermal imaging prototype is an effective tool for initial breast cancer investigations.
• The system offers a radiation-free and precise alternative for early cancer diagnosis.
• Potential to significantly advance early-stage breast cancer detection and improve patient outcomes.