Because of the complexity of tissue features associated with brain tumors, such as varying grades of active tumor, peritumoral edema, and tissue necrosis, characterizing tumors and presurgical planning is challenging with T1-weighted and T2-weighted MR images. Diffusion-MRI derived apparent diffusion coefficient (ADC) has been used to reflect tumor cellularity but significantly confounded by surrounding edema or cerebrospinal fluid (CSF). Surgical resection is the mainstay of glioma treatment but the effectiveness of Diffusion tensor imaging (DTI) tractography was significantly hampered by the complex pathologies in pre-surgical planning. The separation of pathologies and fibers can help address above issues. We have recently developed a new diffusion basis spectrum imaging (DBSI) by modeling diffusion-weighted MR signals as a linear combination of multiple discrete anisotropic (representing multiple crossing fiber tracts) and a spectrum of isotropic (representing cellularity, edema, and necrosis) diffusion tensors^1–4. The purpose of this work is to evaluate the capabilities of assessing complex tumor pathologies and the potentials of DBSI to improve tracking for tumor with confounding peritumoral edema. Patients with brain tumors were diagnosed and imaged at National Yang-Ming University, Taiwan. Diffusion-weighted images (DWIs) were collected with a multi–b-value 99 direction diffusion scheme at 3T (DISCOVERY MR750; GE MEDICAL SYSTEMS) with following parameters: TR = 6 s, TE = 88 ms, maximal b-value = 2000 s/mm², in-plane resolution = 1×1 mm², slice thickness = 2.5 mm, FOV (field of view) = 256 × 256 mm², one average and total acquisition time around 10 minutes. Both DTI and DBSI were computed using in-house software developed by Matlab. DTI and DBSI tractography were performed by DSI-Studio. DTI tracking was done using whole brain seeding with FA threshold 0.1, angular threshold as 60 degree, and step size as 0.5mm. DBSI tracking was also done using whole brain seeding with fiber fraction threshold as 0.1, angular threshold as 60 degree and step size as 0.5mm. Tumor and edema volumes were manually outlined on T2W and FLAIR were used to visually enhance tracking difference between two methods. Complex tissue structures of patient with meningioma were characterized using DBSI (Fig. 1). Complex tissue structures associated with the presence metastatic lung adenocarcinoma and the associated significant brain edema were characterized using DBSI (Fig. 2). The white matter tracts invisible to DTI tractography clearly seen using DBSI with the extent of edema quantified and labeled on individual tracts (Fig. 3). Our results suggest DBSI derived isotropic fractions can quantitatively assess tumor cellularity, edema and necrosis. By separating fibers from confounding pathologies, DBSI was able to detect white matter tracts through peritumoral edema where DTI tractography failed. The quantitative measurements of heterogeneous tumor pathologies and enhanced tracking capability can serve as useful tools to assist tumor treatment and pre-surgery planning.