At present, coronary stent implantation has become an important treatment for coronary atherosclerotic heart disease. It is difficult to identify and detect drug-eluting coronary stent fracture (CSF). Intravascular optical coherence tomography (IVOCT) has unique advantages in CSF identification and detection due to its extremely high imaging resolution. Given the above situation, we propose an accurate reconstruction method of coronary stents in IVOCT. On the basis of the fact that the metal stents have imaging shadows, after the vessel boundaries are segmented, the intensity values of a stent and its shadows at a certain depth are accumulated to generate a one-dimensional array. Then, the array is arranged according to the pullback order to produce a stent reconstruction image. Compared with other stent reconstruction methods (such as three-dimensional imaging, longitudinal image cutting), the proposed method can not only maintain the overall structure of the stents but also avoid the requirement of spatial imagination on the operators. In conclusion, the method proposed in this paper can quantitatively identify and analyze CSF, providing different IVOCT imaging parameters for different stent structures.