Enhanced microvascular imaging through deep learning-driven OCTA reconstruction with squeeze-and-excitation block...

Area of Science:

• Biomedical Engineering
• Medical Imaging
• Cardiovascular Research

Background:

• Skin microvasculature imaging is crucial for health monitoring but faces challenges.
• Existing optical coherence tomography angiography (OCTA) methods require many scans, increasing acquisition time and motion artifacts.

Purpose of the Study:

• To develop a novel deep learning approach for enhanced OCTA processing.
• To improve the accuracy and reduce the acquisition time for skin microvasculature imaging.

Main Methods:

• Integration of a convolutional neural network with a squeeze-and-excitation block into OCTA processing.
• Utilizing local information efficiently for microvascular analysis.
• Dynamically recalibrating features to enhance stability and accuracy.

Main Results:

• The proposed deep learning method significantly reduces measurement time for OCTA.
• Enhanced accuracy in skin microvasculature imaging is achieved.
• Improved stability and reduced motion artifacts compared to traditional methods.

Conclusions:

• Deep learning, particularly with squeeze-and-excitation blocks, offers a powerful solution for OCTA challenges.
• The novel method advances microvascular imaging for better cardiovascular health and thermoregulation assessment.
• This approach paves the way for more efficient and reliable clinical applications of OCTA.