Robust Bi-CBMSegNet framework for advancing breast mass segmentation in mammography with a dual module...

Area of Science:

• Medical Imaging
• Artificial Intelligence in Healthcare
• Oncology

Background:

• Breast cancer screening via mammography is crucial for early detection and reduced mortality.
• Manual mammogram interpretation faces challenges like misdiagnosis due to similar shapes of masses and glands.
• Existing computer-assisted diagnosis tools for breast mass segmentation show limited accuracy and application.

Purpose of the Study:

• To introduce a novel framework, Bi-Contextual Breast Mass Segmentation Framework (Bi-CBMSegNet), for precise and efficient breast mass segmentation in mammograms.
• To enhance the accuracy of automated breast mass segmentation, aiding in earlier and more reliable breast cancer diagnosis.
• To improve the diagnostic and treatment planning processes in breast cancer care.

Main Methods:

• Developed Bi-Contextual Breast Mass Segmentation Framework (Bi-CBMSegNet) using an encoder-decoder architecture.
• Incorporated Global Feature Enhancement Module (GFEM) for comprehensive contextual feature assimilation.
• Integrated Local Feature Enhancement Module (LFEM) for detailed semantic information and precise delineation.

Main Results:

• Bi-CBMSegNet demonstrated superior computational efficiency compared to existing methods.
• The framework achieved enhanced performance metrics in breast mass segmentation accuracy.
• Rigorous evaluation on two public mammography databases confirmed the model's efficacy.

Conclusions:

• Bi-CBMSegNet offers a significant advancement in automated breast mass segmentation for mammograms.
• The proposed framework can improve the accuracy and efficiency of breast cancer screening.
• This technology holds promise for augmenting diagnostic accuracy and treatment planning in breast cancer care.