Efficient and Automatic Breast Cancer Early Diagnosis System Based on the Hierarchical Extreme Learning Machine.

Area of Science:

• Oncology
• Medical Imaging
• Artificial Intelligence

Background:

• Breast cancer is a leading cause of death in women, necessitating early and accurate diagnosis.
• Current diagnostic methods like MRI and biopsy are complex and time-consuming, especially for large-scale screenings.
• Automated image analysis can alleviate the workload and improve diagnostic efficiency.

Purpose of the Study:

• To develop an efficient and automatic diagnosis system for breast cancer using ultrasound images.
• To implement a hierarchical extreme learning machine (H-ELM) for primary image diagnosis.
• To enable early and precise diagnosis of breast cancer from low-resolution ultrasound images.

Main Methods:

• An automated diagnosis system was developed using the hierarchical extreme learning machine (H-ELM) framework.
• The system processes low-resolution (28x28 pixels) PNG ultrasound images.
• Training and application are designed for compatibility with general medical software.

Main Results:

• The H-ELM system achieved an 86.13% accuracy in classifying breast cancer from the BUSI dataset.
• Performance surpassed conventional deep learning methods on the same dataset.
• Training time was significantly reduced to 5.31 seconds with minimal resource consumption.

Conclusions:

• The developed H-ELM system provides a precise and efficient method for early breast cancer diagnosis.
• The system is effective even with low-resolution ultrasound images and limited data.
• This automated approach can assist specialists in primary examinations and large-scale screenings.