REBRO dataset: micro-ultrasound prostate imaging segmentation dataset

REBRO dataset - micro-ultrasound prostate imaging segementation dataset. Contribuion of "Improving Peripheral Prostate UltrasoundSegmentation via Edge-Aware Hybrid ViT-U-NetClass-Seg Model" by Luka Šiktar, Janko Jurdana, Luka Matijević, Matija Markulin,Domagoj Vlah, Toni Zekulić, Jerko Anđelić, Tomislav Kuliš, AnaKlobučar Barišić, Filip Šuligoj, Marko Švaco

The dataset is split into training, validation and test, containing the .nifti files with .dcm (DICOM) images. Each micro-ultrasound image have the ground-truth segmentation by the expert annotator and pre-annotation created by the baseline MicroSegNet model, as non-expert annotator. The created dataset extends the MicroSegNet baseline dataset by adding uniform prostate scans (300 or 100 images per scan) of a whole prostate (from left to right apex.)