A preliminary study of a hybrid transformer with enhanced local feature fusion for nuchal translucency ultrasound image segmentation

Key Points

• To evaluate the performance of a hybrid transformer model, TransNTSeg, for nuchal translucency image segmentation.
• Developed a hybrid transformer model for segmentation tasks.
• Conducted tests in a controlled setting to assess accuracy.
• Focused on local feature fusion methods to enhance segmentation quality.
• TransNTSeg showed competitive performance compared to existing methods.
• Indicated potential for accurate nuchal translucency segmentation in ultrasound images.

Abstract