Font Size: a A A

Research And Application Of Registration And Segmentation Techniques For Optical Coherence Tomography

Posted on:2018-02-18Degree:MasterType:Thesis
Country:ChinaCandidate:P X ZangFull Text:PDF
GTID:2358330518468448Subject:Engineering
Abstract/Summary:PDF Full Text Request
Optical coherence tomography(OCT)is an imaging modality for high-resolution,depthresolved cross-sectional,and 3-dimensional(3D)imaging of biological tissue.The OCT system based on the low coherence of the broadband light source to detect the intensity of backscattered light and reflected light from biological tissue.And combine these detected information with the scan time delay to calculate the 2D tomographic image and 3D image of the biological tissue.This technique has been widely used in clinical treatment of many diseases since it was invented.Among these clinical applications,ocular imaging in particular has found widespread clinical use.In the last decade,due to the development of light source and detection techniques,Fourier-domain OCT have demonstrated superior performance in terms of sensitivity and imaging speed over those of traditional time-domain OCT systems.The high-speed of Fourier-domain OCT has made it easier to image not only structure,but also blood flow.This functional extension was first demonstrated by Doppler OCT that images blood flow by evaluating phase differences between adjacent A-line scans.After this,many OCT angiography(OCTA)techniques have been proposed and used to image vascular networks in human eyes in vivo.Y.Jia,et al.proposed an OCTA algorithm named split-spectrum amplitude-decorrelation angiography(SSADA)in 2012.In the comparison with existing methods,the fundus image based on this OCTA algorithm has been highly improved in contrast,density and connectivity of the vasculature.High-quality OCT and OCTA can help doctors to improve the diagnosis accuracy and stability.The advanced OCT system is important for the quality improvement of the OCT and OCTA.But the post-processing of OCT and OCTA based on image processing technique is also important.As the basic operation of image processing,the registration and segmentation are really important for the post-processing of OCT and OCTA.For the registration,it can be used between two sequence obtained OCT or OCTA to correct the motion artifacts and improve the contrast,density and connectivity of the vasculature.It can be also used to register two image from the same patient with different scanning time.The registered images can reflect the variation of the lesion area during this time.The doctor can make a more accurate and efficient therapeutic schedule based on this variation.In addition,the registration between OCT or OCTA and several traditional fundus images are also important.Because the traditional fundus images such as color fundus photograph(CFPs),fluorescein fundus angiography(FFAs),and indocyanine green angiography(ICGAs)still have their own advantage.The doctor can overall considerate the registered fundus images to improve the accuracy and stability of the diagnosis.And for the segmentation,the most common application is the layer segmentation of retina and choroid.The en face OCT or OCTA based on the segmented layers are very important for the clinical treatment.On the other hand,it can be also used to segment the lesion area in OCT or OCTA.The quantification based on the segmented lesions can objectively reflect the condition of the patient.Based on the applications of registration and segmentation in OCT and OCTA,we have conducted four studies and the achievements were satisfactory.First,we developed a simple and effective strategy to remove motion artifacts using parallel-strip registration and merge two en face angiograms in the same transverse priority.It was very important for the clinical application of high speed(above 200 kHz)OCT system.Second,we propose a 3-D registration algorithm for OCT volume based on both angiographic and structural OCT information.The correction of motion artifacts on the volumetric data not only provides clinically useful cross-sectional images along slow axis but also improves the segmentation efficiency on retinal layers.Third,we proposed a registration method for the multi-modality of retina fundus photographs.Three traditional fundus photographs CFPs,FFAs,and ICGAs were respectively registered with the OCT fundus(OFIs).And this method can be also used to register the OFIs scanned from different time of same patient.Last,we propose an automated disc boundary detection and layer segmentation method for both structural OCT and OCTA of the optic disc.The en face OCT or OCTA based on the segmentation results were very important for the clinical treatment.The proposed four method are very important for the clinical applications of OCT and OCTA.
Keywords/Search Tags:OCT, affine registration, B-spline based free-form deformation, dynamic programming, retina
PDF Full Text Request
Related items