Automated three-dimensional detection of intracoronary stent struts in optical coherence tomography images

Optical coherence tomography (OCT) is a new intracoronary imaging tool that has been recently introduced and has become the method of choice to investigate new treatment methods for coronary artery disease. Due to the OCT's high image resolution, hundreds of stent struts are visualized per patient and therefore a computer-assisted stent strut detection method could help to improve accuracy by reducing analysis time. An automated strut detection algorithm was developed based on an adapted K-nearest neighbor method. Validation in stent just implanted resulted in a success rate of 77%. In a stent follow-up group (n=14) 6 months after implantation with tissue growth a success rate of 50% was observed. Computer-assisted stent strut detection in OCT images is well feasible in patients directly after implantation; in case of considerable tissue growth it is more challenging.