Collaborative Multiscale and Wavelet-Based Fusion Network for Leakage Area Semantic Segmentation of Ultrawide Field Fluorescein Angiography

Key Points

• The aim is to improve the accuracy and efficiency of semantic segmentation in UWFA images to quantify leakage areas related to diabetic retinopathy.
• Developed a collaborative multiscale and wavelet-based fusion network.
• Applied the method to UWFA images to address irregular morphology.
• Focused on enhancing segmentation accuracy and computational efficiency.
• Achieved improved segmentation accuracy for leakage areas in UWFA images.
• Increased computational efficiency allows for quicker analysis.
• Supports timely clinical quantification, potentially aiding earlier interventions.