Meeting Banner
Abstract #1151

Relationship Analysis of Axial and Radial Diffusivities May Be Helpful in Discriminating Tumor-Infiltrating Edema from Pure Vasogenic Edema

Zhigang Min1, 2, Chen Niu1, Ming Zhang1

1Radiology, First Affiliated Hospital of Xi'an Jiaotong University, Xi'an, Shaanxi, China; 2Radiology, Yi'xing Hospital Affiliated to Jiangsu University, Yi'xing, Jiangsu, China

Due to the infiltration of tumor cells, the axons and myelin were destroyed. tumor-infiltrating edema has greater radial diffusivities than pure vasogenic edema. But the direct comparisons of ˡ will be influenced by the degree of edema and can not reveal the diffusion characteristics induced by damaged axons and myelin. We compared the relationships of ˡ and ˡ in pure vasogenic edema and tumor-infiltrating edema and found the significant different between the two types of edema. RCˡ-ˡ showed more effective than other metrics in discriminating tumor-infiltrated edema from pure vasogenic edema.

Keywords

accompanied accuracy affected affiliated anisotropy audience available axial axons blue brain characteristic characteristics characterize characterizing china coefficients commonly conducted conflicting considered contain core covered curve curves damage damaged degree destruction differential diffusion diffusivity direct discriminating distinguishing drawn edema effective eigenvalue eigenvalues electric employed enrolled every except extracellular faster fractional gaps general generate generated gives greater help helpful hospital inconsistent independent induced infiltrated infiltrating infiltration influenced inspiration integrity intensity interpolated johns linear major malignant maps math matrix maturation medical metastasis metastatic metrics might miller ming minor neurosurgeons nonspecific operating outside parallel pathological patients pixel plots principal processed pure radial radiology receiver recorded reduction reflect regression relationship relationships represents resolution respectively reveal samples scatter severe significantly slice software solitary statistically structures studies system systems target tensor thirty tumor types useful version volume white