Research Of Pulmonary Nodules Detection In CT Lung Images

Lung cancer is one of the most common cancers in our life and the rate of incidence and mortality of lung cancer increases rapidly. If the lung cancer can be discovered earlier and treated in time, the patients are able to survive more likely. When a radiologist screens the CT images, however, he will have to read a large number of images, and unfortunately he will overlook some lung nodules. Thus, using a computer-aided diagnostic (CAD) scheme to detect lung nodules is important because it can provides radiologists with the information of the lung nodules in the CT images, helping the radiologists save much time.Through the study of lung nodule detection algorithm, this paper puts forward relevant new algorithms. For the traditional algorithm in solitary pulmonary nodules detection could not effectively eliminate false nodules. The methods we proposed is the Dot-Filter and Region growing for the detection of solitary pulmonary nodules.Firstly, Two-dimensional dot-filter is constructed by using Two-dimensional Hessian matrix, and is used to detect the solitary pulmonary nodules. For the CT values of the vessels and airway do not conform to Gauss distribution, lots of normality treated as nodules will emerge after the way used above. Finally, we use Region growing and Tree-dimensional Hessian matrix to eliminate the false-positive. For the adhesion nodules, a rapid extraction algorithm is proposed. A suspected pulmonary nodule detection method was proposed based on dot-filter and extracting center line algorithm. In this paper, we focus on the distinguishing the adhesion pulmonary nodules attached to vessels in two-dimensional (2D) lung computed tomography (CT) images. Firstly, the dot-filter based on Hessian matrix was constructed to enhance the circular area of the pulmonary CT images, which enhanced the circular suspected pulmonary nodule and suppresses the line-like areas. Secondly, to detect the non-distinguishable attached pulmonary nodules by the dot-filter, an algorithm based on extracting center line was developed to enhance the circle area formed by the end or head of the vessels including the intersection of the lines. Through experiments it proves that this algorithm has high extraction speed and high accuracy.