Study On The Detection Of Lung Nodules From CT Images Based On Hessian Matrix

Lung cancer is a malignant cancer that threatens human health and life. Early diagnosis and treatment of lung cancer patients can help improve the chances of survival. Images of lung cancer can be seen in the form of lung nodule from X-ray computed tomography. With the development of X-CT technology, it is gradually considered as an effective means of early diagnosis of lung cancer. However, considering the large information and confounding factors (such as blood vessels), doctors have to spend much time to analyze data and easily make some mistakes in the process. Therefore, how to make effective extraction of lung nodule became an important medical research. With the progress of current computer technology and mathematical knowledge, effective enhancement and extraction of lung nodule are possible, which can read the information effectively and improve the diagnosis of lung cancer.This paper mainly uses the correlation matrix to enhance lung nodule and extract different morphology of the lung nodules, which can achieve the purpose of computer-aided detection of lung nodules. The paper firstly introduces the basic composition and characteristics of CT images through the basic principles of CT scan and images reconstruction. In order to reduce the computational complexity of CT images, the paper enhances the CT images after the pre-processing, makes use of open computing to eliminate noise, and takes the weighted sum of multiple Gaussian model to achieve the extraction of the lung parenchyma. To complete the enhancement and extraction of lung nodules and effectively improve the efficiency and accuracy, this paper presents a realization of lung nodules based on Hessian matrix enhancement algorithms and respectively uses two-dimensional filter and three-dimensional detection to extract the lung nodules.Experimental results show that, by this design method, it will be more effective and efficient to enhance and extract lung nodules from CT images and can highlight the imaging of lung nodules. So this method can make the diagnosis of lung nodules more efficient and greatly improve the accuracy, which has great practical value in the clinical application.