Estudi i desenvolupament d'un nou sistema d'inspecció de stents basat en tècniques de metrologia òptica

Stent quality control is a critical process. Coronary stents have to be inspected 100% so no defective stent is implanted into a human body. We have developed an optical non-contact measurement instrument that provides stent inspection in an automated and high-speed approach, delivering the user all...

ver descrição completa

Detalhes bibliográficos
Autor: Bermúdez Porras, Carlos
Formato: tesis doctoral
Fecha de publicación:2017
País:España
Recursos:Universitat Politècnica de Catalunya (UPC)
Repositorio:UPCommons. Portal del coneixement obert de la UPC
Idioma:catalán
OAI Identifier:oai:upcommons.upc.edu:2117/108237
Acesso em linha:https://hdl.handle.net/2117/108237
https://dx.doi.org/10.5821/dissertation-2117-108237
Access Level:acceso abierto
Palavra-chave:Microscopia
Metrologia òptica
Visió per computador
Inspecció de stents
Inspecció visual
Microscopy
0ptical metrology
Machine vision
Stent inspection
Visual inspection
Àrees temàtiques de la UPC::Ciències de la salut
Descrição
Resumo:Stent quality control is a critical process. Coronary stents have to be inspected 100% so no defective stent is implanted into a human body. We have developed an optical non-contact measurement instrument that provides stent inspection in an automated and high-speed approach, delivering the user all the necessary information to assess whether a stent is valid or needs to be rejected. As regards to the instrument design, we have developed and built an optical sensor head that obtains high resolution colour images of the stent surfaces at high speed. Acquired images contain not only contour information of the struts, but also depict the edges roundnesses caused by the surface treatment. This has been achieved by the integration of a triple illumination system in combination with a high precision rotational stage and an area seen camera, which also acquires in line-scan mode. As a result, a patent filed by Sensofar Medical, S.L. protects this invention, which is also being commercialized by the same company. Regarding to the rotational system, different driving configurations have been analysed, and the motion errors that they produce to the images have been studied. Taking this into account, manufacturing tolerances have been established accordingly to guarantee high quality image acquisition. As regards to image acquisition and analysis, we have studied and characterized the difference between the image obtained with a conventional, bright field microscope and the proposed system. Moreover, we have introduced a new method to measure sidewalls critical dimensions, consisting of taking the same, line-scanned images at a certain observation angle. Together with a correction model, 1-micron precision measurements are obtained. To be able to process the measurements in an automated approach, we have developed critical dimension measurement algorithms for the four stent surfaces: abluminals (inner and outer) and luminal (sidewalls). Additionally, we have studied, developed and implemented defect detection algorithms controlled by a sensitivity parameter adjusted by the user. Defects are also classified in a supervised classifier approach by means of a previous training with a defects database. The instrument has also been enabled with tridimensional measurement capabilities, based on coherence scanning interferometry to obtain the topography and roughness of any surface. This technique is also used to characterize defects in 30 in order to reduce the stent rejection rate and thus increasing the process efficiency, as the height of the candidates which maybe are not defects can be also assessed. Besides, this technique can also measure the transparent film thicknesses of drug eluting stents. In this case, due to the cylindrical shape of the samples, measured thicknesses are altered by the local slope at the measuring point, so we have proposed and evaluated two correction models that provide the thickness 30 map corrected automatically. All the capabilities described provide a number of advantages in the stent production and quality control stages. Specifically, objectivity and traceability of these processes are significantly increased, making stents every day more safe devices, reducing the restenosis rate, cardiac accidents or illnesses resulting from an implant malfunction.