Digital Eversion Techniques For Human Hollow Structures

Nowadays, medical imaging and visualization techniques are increasingly applied tothe medical education and clinical diagnosis. As for human hollow structures, the commonlyused method for 3D visualization is the technique of virtual endoscopy. However, the virtualendoscopic method has the defect of limited view angle, which makes the visualization anddiagnosis become time-consuming, tedious and prone to errors. The existing improvementsfor virtual endoscopy, such as virtual unfolding and ?attening, virtual cutting open, still havesome disadvantages.In this paper, a totally new concept for visualization of human hollow structures ispresented, which is called"digital eversion". The main idea of digital eversion is to convertthe inner surface of a hollow structure to the outer surface via computer. The advantagesof this method include the direct visualization of a larger portion of the inner surface andthe close correlation to the important features of the global anatomy, without the need fortedious and time-consuming navigation.Two eversion algorithms are proposed based on partial electrical field model. Thus,each point on the inner surface of a hollow structure can be everted outside along the elec-trical force lines generated either from charged centerline points or from sampling points onthe neighboring surface. A real CT dataset of human colon is tested to implement the digitaleversion experiments. The experimental results show the feasibility of the digital eversionapproach.Finally, some other methods which can improve the effect of visualizing the human hol-low structures are also originally proposed. Those methods include color mapping, virtualinsufflation, global virtual cutting and a method of fast volume rendering based on segmen-tation and dilation.