Diffusion Tensor Images of Metastatic Brain Tumors:the Impact of Peritumoral Edema on White Matter Indices

• Main Authors: Li-en Lin, 林立恩
• Other Authors: Ja-chih Fu
• Format: Others
• Language: zh-TW
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/88209165178862397903

碩士 === 國立雲林科技大學 === 工業工程與管理研究所碩士班 === 100 === In recent year, Magnetic Resonance Imaging (MRI) and Diffusion Tensor Imaging (DTI) has become an essential portrait medical treatment, MRI can make extraordinary portrait performances in brain tumor and adjacency edema on medical treatment, DTI can not only observe diversification between brain tumor and white matter fiber bundles but evaluate relationship between tumor and brain tissue. Sometimes brain tumor may undeveloped in T1-weighted MRI, this study will evaluate diversification between brain tumor and other brain tissue by applying region of interest and combining measure index of magnetic resonance diffusion. This thesis will reconstruct 3D visualization for brain tumor and area surrounding edema; then we conduct image fusion after reconstruction. This study will calculate the distance between area surrounding edema and white matter by applying morphological image processing, and then we sort out region of interest, using fractional anisotropy and mean diffusivity to be measurement indicators. This study of brain magnetic resonance imaging include four metastatic cancer patients and two normal subjects, research area cover edema side around tumor of cancer patients and white matter of normal subjects, we develop region of interest by applying morphological image processing. Each step represents that we start three-dimensional dilation to expanse four pixels, in our experiment results; we can find FA rising and MD declining in region of interest of D4、D8 in cancer patients, FA and MD emerge smooth in white matter area beside edema area. FA is the lowest and MD is the highest in D4 in five steps in brain tumor patients experiment results. It demonstrates that the area around tumor will decline direction and rise average diffusion coefficient. This thesis focus on region of interest developed and compares FA and MD with brain tissue in difference area between cancer patients and normal subjects. In our experiments, we can see that results show significant difference in FA and MD in four areas under 90% confidence levels. The values of FA and MD in side of the brain edema and white matter regions outside the area and edema are significant difference（p-value＜0.1）. The values of FA and MD in side of the brain edema region corresponds to the region and against the side are significant difference（p-value＜0.1）. The values of FA and MD in brain edema side against the side of the white matter and the corresponding region are significant difference（p-value＜0.1）. The values of FA and MD in side with the opposition side of the brain tumor area of common concern are significant difference（p-value＜0.1）.We can find the diversification between tumor and white matter area around tumor; we hope this study can offer some references for medical image.