Comparison of 2D vs. 3D Unet Organ Segmentation in abdominal 3D CT images

A two-step concept for 3D segmentation on 5 abdominal organs inside volumetric CT images is presented. Firsteach relevant organ’s volume of interest extracted as bounding box. The acts input asecond stage, wherein two compared U-Nets with different architectural dimensions re-construct an organ segmen-tation label mask. In this work, we focus comparing 2D vs. U-Net counterparts. Our initial resultsindicate Dice improvements about 6% at maximum. study to our surprise, liver and kidneys instancewere tackled significantly better using the faster GPU-memory saving U-Nets. For other keyorgans, there were no significant differences, but observe highly advantages interms GPU computational efforts all under study.