The accurate and automatic segmentation of tissue regions in cervigram images can aid in the identification and classification of precancerous regions. We implement and analyze four GPU (Graphics Processing Unit) based clustering algorithms: K-means, mean shift, deterministic annealing, and spatially coherent deterministic annealing. From our results, we propose a novel parallel algorithm using the CUDA programming language for digital cervigram segmentation and clustering. The first step of our fully automatic method is to compute the number of modes in the feature space of a color cervigram image using the mean shift clustering algorithm. Next, we use the number of modes in a novel spatially coherent deterministic annealing optimization technique to produce an approximate optimal solution for the clustering problem. Our GPU based methods perform approximately 38x (deterministic annealing), 134x (mean shift), and 276x (spatially coherent deterministic annealing) faster than an equivalent CPU solution. Our implementation decreases the computational time of an annealing method on a 1280x872 pixel image from 5 hours 3 minutes to 72.12 seconds, enabling the use of this optimization method in clinical settings and on large cervigram datasets.

We present a novel parallel segmentation algorithm based on cellular automata for 3D medical image volumes. The proposed method is able to segment medical volumes quickly and accurately using any number of label classifications. Furthermore, we show that the inherent parallelism of cellular automata reinforces our adaptation of this method to the GPU (Graphics Processing Unit). The results demonstrate that our method is able to compute a complete segmentation nearly 45x faster than conventional CPU methods. Additionally, we have utilized multiple image features including probabilistic density models to increase the final segmentation accuracy. By using the GPU, our parallelized cellular automata is able to give users feedback at interactive rates. From our quantitative and qualitative results, we show that the proposed method enables fast and reliable medical volume segmentation.