hamghalam/convolutional-3d-to-2d-image

A novel pixel-wise segmentation framework through a convolutional 3D to 2D MR patch conversion model to predict class labels of the central pixel in the input sliding patches.

Convolutional-3D-to-2D-image conversion

A novel pixel-wise segmentation framework through a convolutional 3D to 2D MR patch conversion model to predict class labels of the central pixel in the input sliding patches.

Segmentation results are overlaid on FLAIR axial and saggital slices on BraTS 2019 Validation Data. The yellow label is edema, blue color means enhancing tumor, and the green one presents the necrotic and non-enhancing tumor core. Each column displays one slice of different Subject IDs of BraTS 2019 (https://www.cbica.upenn.edu/BraTS18/lboardValidation.html) .