Deep learning model for fully automated breast cancer detection system from thermograms

Area of Science:

• Medical Imaging
• Artificial Intelligence
• Oncology

Background:

• Breast cancer is a leading cause of death in women globally.
• Early detection significantly improves patient outcomes.
• Thermography offers a non-invasive method for breast imaging using infrared technology.

Purpose of the Study:

• To develop a fully automatic system for breast cancer detection using thermal imaging.
• To enhance the accuracy and efficiency of breast cancer diagnosis through artificial intelligence.

Main Methods:

• Utilized U-Net network for automatic segmentation of breast tissue from thermal images.
• Developed and trained a two-class deep learning model for classifying normal versus abnormal breast tissue.
• Employed the DMR-IR benchmark database for system evaluation.

Main Results:

• Achieved a high diagnostic accuracy of 99.33%.
• Demonstrated excellent sensitivity (100%) and specificity (98.67%) in detecting breast abnormalities.
• The automated system effectively isolated breast tissue, reducing noise for improved classification.

Conclusions:

• The proposed automated thermography system shows high efficacy for breast cancer detection.
• This AI-driven tool has the potential to assist physicians in clinical settings for earlier and more accurate diagnosis.
• Further integration into clinical practice could improve breast cancer screening protocols.