Image Algorithm And Application In Optical Coherence Tomography

New medical imaging technologies continue to appear for the ability to improve thediagnosis and treatment of diseases. They can also contribute to a better understanding of thepathogenesis of the disease, and thus promote the development of new treatments. OpticalCoherence Tomography (OCT) is a new kind of medical imaging technology, it can achievein vivo tissue micro-structure "optical biopsy", without the need for moving and handling ofspecimens. The general clinical OCT applications:(1) biopsy of tissue which is impossible tomove. It can be applied to organizations including eyes, artery, or nerve tissue;(2) Standardexamination of the excised tissue and histopathological diagnosis of many diseases, such ascancer.(3) Ability to guide surgery. OCT has the ability to see the subcutaneous tissue, so itcan help the micro-operation surgery such as stenting. OCT can guarantee the accuracy of thesurgery and can develop new minimally invasive surgery. The OCT Doppler imaging,birefringence and spectral characteristics and so can be used in the development of newsurgical instruments.The OCT use low coherent light to obtain depth information of the target.Two-dimensional or three-dimensional imaging is obtained by continuous scans in the X or Ydirection. Like many other imaging modalities, digital signal processing techniques canimprove its image quality. Digital signal processing can suppress noise and to ensure that theimage obtained is consistent with the known attributes of the object. Images that are morecomplex, more detailed can reveal more features to some extent. Since the OCT can be applied to clinical ophthalmology, retinal internal microstructurecan be detected in detailed. OCT is a more rapid and non-invasive examination which isdifferent from the other way, such as Fluorescein Angiography (FA). In many differentclinical studies, retinal thickness detection is an important parameter.The previous segmentation algorithms tend to be concentrated in the use of edge-basedsegmentation techniques. This thesis proposes a region-based retinal image segmentationalgorithm, including extraction of target area, interpolation enlargement, establish orientationmap, optimized merger, stepper control. Open-close reconstruction is used to extract theretinal target area, which can improve the efficiency and accuracy of the segmentationalgorithm; use the B-spline interpolation to enlarge, which can reduce the amount ofsegmentation in the case of retaining the boundary information; calculate the imageorientation map as judgment basis; calculate the characteristics of the different regions andsimilarity, which can guide region merger; stepper control is used to improve merge processoccurs in smaller threshold. After the segmentation of this algorithm, what we can get fromthe retinal images is boundary information and hierarchical information, which may play asignificant role in the diagnosis and treatment of eye diseases.Finally, gastrointestinal endoscopic scan’s prospective application of optical coherencetomography in vitro slice scanning experiments is done, then the effects and causes of itsimaging are analyzed.