A New 3D Segmentation Algorithm Based on 3D PCNN for Lung CT Slices

Abstract

Three-dimension (3D) based image data analysis has an important role for significantly improving the detection and diagnosis of lung disease with computed tomography (CT). In this paper, we proposed a new volume-based 3D segmentation algorithm based on the extended 3D pulse coupled neural network (PCNN) model. This algorithm was successfully used to segment the lung field in CT slice with the mean distance, root means square distance and Tanimoto coefficient of 0.0029±0.0005, 0.0715±0.0056, 0.9760±0.0093, respectively. Furthermore, the means running time was only 273s, which was much less than those of 2D PCNN segmentation algorithm and Otsu algorithm. The experimental results demonstrated the extended 3D PCNN segmentation algorithm had the advantage of short execution time with good segmentation accuracy. The results suggest that the proposed 3D PCNN algorithm can be potentially used for lung computer-aided diagnosis.