Automatic Lung Regions Extraction Algorithm from 3D CT-Images Based on the Bit-Plane Slicing Technique
Sammouda, Rachid . 2006
This paper describes a method for an automatic extraction of lung regions from 3D-CT images using pure basic image processing techniques. This extraction process should be as much as possible accurate and reliable because its results will be used as a base to develop a Computer Aided Diagnosis (CAD) system for lung cancer. First, each 2D slice is converted to a set of binary images using bit-plane slicing technique instead of the thresholding technique that is used in most of the proposed systems. Bit-plane slicing technique is both faster and data and user independent compared to the thresholding technique. Then a sequence of image processing techniques such as erosion, median filter, and dilation is applied to each bit-plane component of the 2D slice. In our study the third lowest bit-plane shows a higher accuracy in the bellow described algorithm for lung regions extraction. The method has been tested by processing 2668 CT slices from 11 patients, and has been successful in extracting the lung regions in 95% of all cases
Presents contributions to improve a previously published approach for the segmentation of magnetic resonance images of the human brain, based on an unsupervised Hopfield neural network.
The paper presents a method for automatic segmentation of sputum cells color images, to develop an efficient algorithm for lung cancer diagnosis based on a Hopfield neural network.