Automated joint skull-stripping and segmentation with Multi-Task U-Net in large mouse brain MRI databases

Riccardo de Feo, Artem Shatillo, Alejandra Sierra, Juan Miguel Valverde, Olli Gröhn, Federico Giove, Jussi Tohka

Abstract: Highlights•We present convolutional neural network (MU-Net) for segmentation of mouse brain MRI.•MU-Net yielded an excellent performance when tested on 1782 mouse MRI volumes.•MU-Net generalized to different mouse genotypes and gender.•MU-Net is extremely fast, requiring less than a second to segment a single image.