Surgery remains dangerous, and accurate knowledge of what is presented to the surgeon can be of great importance. One technique to automate this problem is non-rigid tracking of time-of-flight camera scans. This requires accurate sensors and prior information as well as an accurate non-rigid tracking algorithm. This thesis presents an evaluation of four algorithms for tracking and semantic labeling of deformable tissues for medical applications, as well as additional studies on a stretchable flexible smart skin and dynamic 3D bioprinting. The algorithms were developed and tested for this study, and were evaluated in terms of speed and accuracy. The algorithms tested were affine iterative closest point, nested iterative closest point, affine fast point feature histograms, and nested fast point feature histograms. The algorithms were tested against simulated data as well as direct scans. The nested iterative closest point algorithm provided the best balance of speed and accuracy while providing semantic labeling in both simulation as well as using directly scanned data. This shows that fast point feature histograms are not suitable for nonrigid tracking of geometric feature poor human tissues. Secondary experiments were also performed to show that the graphics processing unit provides enough speed to perform iterative closest point algorithms in real-time and that time of flight depth sensing works through an endoscope. Additional research was conducted on related topics, leading to the development of a novel stretchable flexible smart skin sensor and an active 3D bioprinting system for moving human anatomy.