Automatic detection of stent struts with thick neointimal growth in intravascular optical coherence tomography image sequences

To assist cardiologists investigating neointimal tissue growth on stents during follow-up with optical coherence tomography (OCT), we developed an automatic algorithm to locate deeply buried stent struts and to quantify the restenosis burden. The technique is based on an improved steerable filter for computing the local ridge strength and orientation. It also uses an ellipsoid fitting algorithm and continuity criteria to obtain globally optimal stent localization. The restenosis burden calculations were compared to manual assessment of OCT coronary artery image data obtained from in vivo human clinical studies. Compared to manual assessment by expert readers, the algorithm operated with > 97% accuracy in the measurement of mean and maximum restenosis burden. The results indicated that the technique yielded comparable accuracy in measuring restenosis burden, and significantly reduced user interaction time.