PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.
Advances in three-dimensional (3D) microscopy are providing never-before-seen images of coronary microvasculature organization. However, it remains inaccessible to researchers due to difficult sample preparation and image analysis. We present a deep learning network that can segment the coronary microvasculature in 3D microscopy without vessel staining. The network is based on 3D U-net and accepts DAPI (nuclei) and autofluorescence (tissue structure) volumes as inputs. The network detects vessels with high accuracy when compared to the ground truth obtained from isolectin staining. Contrast-free segmentation of vessels simplifies sample preparation, frees fluorescent channels during imaging and opens the door toward user-friendly 3D microscopy.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.
The alert did not successfully save. Please try again later.
Maryse Lapierre-Landry, Yehe Liu, Mahdi Bayat, David L. Wilson, Michael W. Jenkins M.D., "Contrast-free segmentation of blood vessels using deep learning," Proc. SPIE PC12355, Diagnostic and Therapeutic Applications of Light in Cardiology 2023, PC123550E (17 March 2023); https://doi.org/10.1117/12.2650480