Robust generation of cropped images in the context of stent production

Context

Coronary heart disease (CHD), which includes myocardial infarction, is currently the most common cause of death in Germany. One of the ways in which they can be treated is by implanting a cardiovascular implant ("stent") in the affected vessel. These (stents) are therefore medical devices that must meet appropriate quality criteria. At the same time, personalization, i.e. the production of a stent adapted to a patient, is the exception. If such personalization takes place, the affected vessel is measured by means of imaging (e.g. CT and MRI) and a stent is simulatively designed for it. Subsequently, the stent is produced, e.g. by means of braiding. The designed stent geometry must be adhered to as precisely as possible, otherwise, in the worst case, one or more thromboses or ruptures of the vessel may occur. Therefore, a visual inspection system is being developed at ITIV that monitors the geometry of a stent during its production so that intervention can be taken in the event of an error. As part of this system, images of the manufactured stent are captured using a camera. The image is then cropped and the stent is measured. During this process, errors can occur in the measurement system due to vibrations and movement of the camera and/or the stent.

Goals

Design and implementation of a robust procedure for image section generation, which compensates for stent displacements in the image.

Requirements

• English business fluent in written and spoken
• Structured and analytical approach
• Programming experience in Python
• Optional: Knowledge in the field of image processing
• Optional: Knowledge in the area of (image) registration
• Optional: Knowledge in the area of machine learning