CylinGCN: Cylindrical structures segmentation in 3D biomedical optical imaging by a contour-based graph convolutional network

Image reconstruction of multispectral sparse sampling photoacoustic tomography based on deep algorithm unrolling

“…In recent years, deep learning has gained prominence in both industrial and medical imaging applications [18] , [19] , [20] . For PAT image reconstruction, Waibel et al [21] first proposes to utilize the U-Net framework to realize direct image reconstruction, which harnesses the time series data to obtain initial pressure distribution without pre-processing.…”