Hybrid receptive field U-Net for lesion segmentation in breast ultrasound images.

Key Points

• To improve the segmentation accuracy of lesions in breast ultrasound images using a hybrid receptive field U-Net.
• Development of the hybrid receptive field U-Net architecture.
• Application of the U-Net model to breast ultrasound images.
• Comparison of segmentation results against traditional methods.
• Significant improvements in segmentation accuracy compared to standard approaches.
• Enhanced detection of lesions in breast ultrasound images.