Skip to main content
. 2024 Nov 21;15(12):6905–6921. doi: 10.1364/BOE.541655

Fig. 3.

Fig. 3.

Automatic 3D segmentation results of freestanding individuals from the test dataset using our proposed semi-supervised learning framework. (a1)-(c1) demonstrate that raw 3D OCT data consist of massive B-scans. (a2)-(c2) show the manually annotated 3D results, (a3)-(c3) are the corresponding 3D segmentation results obtained by our method.