With recent advances in machine learning, semantic segmentation algorithms are becoming increasingly general-purpose and translatable to unseen tasks. Many key algorithmic advances in the field of medical imaging are commonly validated on a small number of tasks, limiting our understanding of the generalizability of the proposed contributions. A model which works out-of-the-box on many tasks, in the spirit of AutoML (Automated Machine Learning), would have a tremendous impact on healthcare. The field of medical imaging is also missing a fully open source and comprehensive benchmark for general-purpose algorithmic validation and testing covering a large span of challenges, such as: small data, unbalanced labels, large-ranging object scales, multi-class labels, and multimodal imaging, etc. To address these problems, in this project, as part of the MSD challenge, we propose a generic machine learning algorithm which we applied on two organs: liver and tumors, spleen. We propose an unsupervised generic model by implementing U-net CNN architecture with Generalized Dice Coefficient as loss function and also as a metric. The MSD dataset consists of dozens of medical examinations in 3D (per organ), we’ll transform the 3-dimensional data into 2-d cuts as an input of our U-net. Experimental results show that our generic model based on U-net and Generalized Dice Coefficient algorithm leads to high segmentation accuracy for each organ (liver and tumors, spleen), separately, without human interaction, with a relatively short run time compared to traditional segmentation methods.