Fluid mask 3.0.21/31/2024 ![]() Serranho proposed a mathematical formulation to produce synthetic OCT boundaries on 10 manually segmented healthy human OCT data. In recent years, synthetic retinal fundus images, as well as OCT retinal images have been proposed. In previous works, synthetic images are developed as a benchmark to compare and evaluate the performance of medical image processing algorithms such as CT scans, MRI, and ultrasound images. Denoising methods can also be evaluated in different noise levels, on abnormal structures, and in 3D data. On the other hand, the proposed synthetic data can be used in the evaluation of segmentation methods to measure the robustness to noise and the performance in presence of abnormalities. Furthermore, the synthetic OCTs include the boundary locations for each scan, which is appropriate in applications like segmentation. To deal with mentioned limitations, the proposed method produces the new OCT data with a similar appearance to limited local datasets. Furthermore, most of the online datasets only contain the label for each image and delineation of the boundaries is rarely given. Namely, if the network is trained on less similar online datasets, the difference in device-specific parameters, size, resolution, noise, etc., will affect the performance on local datasets. ![]() Even though a great number of OCT data is available online, they may not necessarily resemble specific datasets. Today, deep learning algorithms are widely used in this application and their performance is in presence of abnormalities. Over the past two decades, researches on OCT image processing has been devoted to the main areas: segmentation of the retinal layers, classification, enhancement, and denoising. Manual analysis of this data is tedious, time-consuming, and prone to error. 1c) and each B-scan is composed of 1D signals in the z-direction (A-scans in Fig. 1a) is composed of cross-sectional images called B-scan (x–z images in Fig. Stacks of cross-sectional images yield three-dimensional (3D) volumes with detailed data for diagnosis of retinal diseases such as age-related macular degeneration, diabetic retinopathy, glaucoma, diabetic macular edema, etc. Optical coherence tomography (OCT) is one of the advanced and recent imaging techniques that provide valuable information for ophthalmologists with high-resolution images from transverse cuts of the retina. The primary task of the retina is to convert luminous energy into analyzable signals for the brain. The retina is an important component in the eye, made up of several inter-retinal layers. The proposed algorithm provides a unique benchmark for comparison of OCT enhancement methods and a tailored augmentation method to overcome the limited number of OCTs in deep learning algorithms. Regarding the texture features of the synthesized datasets, Q-Q plots were used, and even in cases that the points have slightly digressed from the straight line, the p-values of the Kolmogorov–Smirnov test rejected the null hypothesis and showed the same distribution in texture features of the real and the synthetic data. Visual inspection of the synthesized vessels was also promising. Statistical comparison of thickness maps showed that synthetic dataset can be used as a statistically acceptable representative of the original dataset ( p > 0.05). In this method, a limited number of OCT data is used during the training step and the Active Shape Model is used to produce synthetic OCTs plus delineation of retinal boundaries and location of abnormalities. We propose a new synthetic method, applicable to 3D data and feasible in presence of abnormalities like diabetic macular edema (DME). Due to complicated data structure in retinal OCTs, a limited number of delineated OCT datasets are already available in presence of abnormalities furthermore, the intrinsic three-dimensional (3D) structure of OCT is ignored in many public 2D datasets. Synthetic OCT was recently considered to provide a benchmark for quantitative comparison of automatic algorithms and to be utilized in the training stage of novel solutions based on deep learning. Nowadays, retinal optical coherence tomography (OCT) plays an important role in ophthalmology and automatic analysis of the OCT is of real importance: image denoising facilitates a better diagnosis and image segmentation and classification are undeniably critical in treatment evaluation.
0 Comments
Leave a Reply.AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |