Angiogenesis is critically important in invasive tumor growth and metastasis. Many recent studies have focused on the microvasculature in tumor. The most often studied tumor tissues are from animals; rarely human tumors are selected. The analysis of microvasculature may provide useful information to understand the microperfusion of blood in tumor. One question is whether integration of histological analysis of the microvasculature in combination with diffusion-weighted MRI (DWI) gives coherent information about microperfusion. In this work, we aim to achieve the quantitative analysis of microvasculature in human tumor and attempt to compare the vessel features with blood perfusion in tumor. We applied a novel automatic procedure to extract blood vessels and reconstruct the 3D vessel structures in lung tumor based on serial histological tumor tissue slides. The digitized histological tissue slide is in 2D with fine resolution down to cellular level. The method was tested on a stack of 288 serial slides of a human lung tumor. We compared the vessel area fraction in two specific lung tumor regions with the perfusion-related parameter from IVIM (intravoxel incoherent motion) of DWI. High perfusion was observed in tumor sample with high vessel area fraction, hence our preliminary results indicate a possible relation between DWI and histological parameters, which shall be consolidated in future studies.