A New Deep-Learning-Based Model for Breast Cancer Diagnosis from Medical Images.

Area of Science:

• Oncology
• Medical Imaging
• Artificial Intelligence

Background:

• Breast cancer is a leading cause of death among women globally, making early detection critical for improved patient outcomes.
• Machine learning (ML) and deep learning (DL) show promise for enhancing breast cancer screening accuracy, particularly in resource-limited settings.
• DL models typically require large datasets, which are often scarce in medical imaging, hindering their performance.

Purpose of the Study:

• To develop a novel deep learning model for improved breast cancer classification and detection.
• To address the challenge of limited data in medical imaging for deep learning models.
• To enhance diagnostic accuracy and reduce the workload on healthcare professionals.

Main Methods:

• A new deep learning model was proposed, integrating concepts from GoogLeNet and residual blocks.
• The model incorporates granular computing, shortcut connections, learnable activation functions, and an attention mechanism.
• Performance was evaluated by comparing the proposed model against state-of-the-art deep models using two case studies.

Main Results:

• The proposed deep learning model achieved high accuracy rates in breast cancer classification.
• Specifically, the model attained 93% accuracy on ultrasound images and 95% accuracy on breast histopathology images.
• The integration of granular computing and attention mechanisms contributed to improved diagnostic precision.

Conclusions:

• The developed deep learning model offers a promising approach for accurate and efficient breast cancer detection.
• The model's ability to perform well with limited data and its high accuracy suggest its potential clinical utility.
• This research contributes to advancing AI applications in medical imaging for cancer screening.