视网膜光学相干层析图像的预处理方法

光相干层析技术(OCT)是近来发展起来的成像技术,利用光相干层析技术可以得到清晰的视网膜以及黄斑区的层析图。主要论述了在开发视网膜光相干层析图像识别和临床诊断系统中采用的图像预处理方法。利用图像分割、图像增强等处理技术,实现了图像的自动边缘检测与轮廓提取。通过对视网膜光相干层析图像进行预处理,为进一步的图像识别与分析打下了良好的基础。对黄斑囊样水肿、黄斑裂孔以及正常的视网膜光相干层析图像进行了实验,均取得了较好的效果。     

The potential optical coherence tomography in tooth bleaching quantitative assessment

In this paper, we report the outcomes from a pilot study on using OCT functional imaging method to evaluate and quantify color alteration in the human teeth in vitro. The image formations of the dental tissues without and with treatment 35% hydrogen peroxide were obtained by an OCT system at a 1310 nm central wavelength. One parameter for the quantification of optical properties from OCT measurements is introduced in our study: attenuate coefficient (μ). And the attenuate coefficient have significant decrease (p < 0.001) in dentine as well as a significant increase (p < 0.001) in enamel was observed during tooth bleaching process. From the experimental results, it is found that attenuate coefficient could be useful to assess color alteration of the human tooth samples. OCT has a potential to become an effective tool for the assessment tooth bleaching. And our experiment offer a now method to evaluate color change in visible region by quantitative analysis of the infrared region information from OCT.     

Compact,low-cost Ti:Al2O3 laser for in vivo ultrahigh-resolution optical coherence tomography

A compact, low-cost, prismless Ti:Al2O3 laser with 176-nm bandwidth (FWHM) and 20-mW output power was developed. Ultrahigh-resolution ophthalmic optical coherence tomography (OCT) ex vivo imaging in an animal model with approximately 1.2-microm axial resolution and in vivo imaging in patients with macular pathologies with approximately 3-microm axial resolution were demonstrated. Owing to the pump laser, this light source significantly reduces the cost of broadband OCT systems. Furthermore, the source has great potential for clinical application of spectroscopic and ultrahigh-resolution OCT because of its small footprint (500 mm x 180 mm including the pump laser), user friendliness, stability, and reproducibility.     

Imaging of subcutaneous microcirculation vascular network by double correlation Optical Coherence Tomography

Spatial distribution of superficial blood vessels in human skin in vivo has been observed by using the double correlation Optical Coherence Tomography (OCT). To remove background noise, reduce the artifacts associated with patient motions and to increase the overall quality of the experimental OCT images an adaptive Wiener filtering technique has been employed. Fourier domain correlation has been subsequently applied to enhance spatial resolution of images of vascular network in human skin in vivo. Image processing has been performed on Graphics Processing Units (GPUs) utilizing Compute Unified Device Architecture (CUDA) framework in the frequency‐domain. This approach allows carrying out image processing in parallel significantly speeding up the computations. The presented results show that the double correlation method permits obtaining 2D/3D OCT images of subcutaneous microcirculation vascular network and its spatial distribution within the human skin with higher spatial resolution compare to the other OCT correlation‐based techniques developed earlier.     

Optical Coherence Tomography Premiere in Korea

Optical Coherence Tomography (OCT) is considered as fundamentally new image modality having advantages such as high-speed, real-time cross sectional imaging with non-invasive in vivo investigation of biological tissue and its potential for clinical application is highly evaluated. In spite of these merits, the research activity in OCT has not taken much attention until recently. However the situation in Korea is gradually changing due to understanding of its versatile applications. Active research and commercialization are underway. Bulk optics-based OCT system is changed into alignfree fiber-optics system. With our expertise in the fabrication of fiber grating devices and strong background in control of fiber based system, we hope to contribute in developing simplified all-fiber OCT system with better performance. As a leading research group, working on OCT, we have firstly implemented fiber-based OCT system in Korea and demonstrated the imaging performance with glass plate and onion.     

多尺度条件卷积的OCT视网膜图像降噪研究

散斑噪声存在于光学相干层析成像(OCT)中,影响OCT图像质量.在使用OCT设备诊断各种常见眼科疾病时,高质量的OCT图像是极为重要的.利用深度神经网络对OCT图像进行降噪处理,使图像在保留空间结构细节的基础上能展示更多的信息.提出了一种基于残差学习网络的新型OCT图像降噪网络-CMCNN,其具有多尺度、多权重和多层次...     

New Approaches in Broadband Fiber-Optical Interferometry for Optical Coherent Tomography

We have developed new methods of IR ultrabroadband interferometry for increasing the limiting capabilities of optical coherent tomography (OCT). In particular, this permitted us to create an optical coherent microscope combining the principles of confocal microscopy and OCT. To increase the longitudinal resolution, we synthesized an ultrabroadband optical spectrum based on two superluminescent diodes with separated radiation spectra. The resulting longitudinal resolution was 3.4 m and the transverse resolution was 3.9 m. Images of the inner structure of biological tissues are obtained at the cellular level. Experimental results of a study of biological tissue by the cross-polarization OCT are described.     

Quantification of molecular diffusion in arterial tissues with optical coherence tomography and fluorescence microscopy

Alternations in vascular permeability for different molecules, drugs, and contrast agents might be a significant early marker of development of various diseases such as atherosclerosis. However, up to date experimental studies of molecular diffusion across vascular wall have been limited. Recently, we demonstrated that the Optical Coherence Tomography (OCT) technique could be applied for noninvasive and nondestructive quantification of molecular diffusion in different biological tissues. However, the viability of the OCT-based assessment of molecular diffusion should be validated with established methods. This study focused on comparing molecular diffusion rates in vascular tissues measured with OCT and standard fluorescent microscopy. Noninvasive quantification of tetramethylrhodamine (fluorescent dye) permeability in porcine vascular tissues was performed using a fiber-based OCT system. Concurrently, standard histological examination of dye diffusion was performed and quantified with fluorescent microscopy. The permeability of tetramethylrhodamine was found to be (2.08 ± 0.31) × 10^?5 cm/s with the fluorescent technique (n = 8), and (2.45 ± 0.46) × 10^?5 cm/s with the OCT (n = 3). Good correlation between permeability rates measured by OCT and histology was demonstrated, suggesting that the OCT-based method could be used for accurate, nondestructive assessment of molecular diffusion in multilayered tissues.     

Determination of optical properties of oxidative bleaching human dental tissue samples using optical coherence tomography

Oxidative bleaching changes of human teeth induced changes in the optical properties of dental tissue. We introduced 1310 nm wavelengths of optical coherence tomography (OCT) attenuation coefficient method which is a relatively novel and rarely reported methodology to measure the correlation coefficient during the teeth oxidative bleaching procedure. And the quantitative parameters of enamel optical thickness and disruption of the entrance signal (DES) were extracted from the OCT images. The attenuation coefficient of the bleached tissue is 6.2 mm⁻¹ which is significant (p < 0.001) higher than that unbleached sample is 1.4 mm⁻¹. But attenuation coefficient varied significantly (p < 0.001) between 5.9 and 1.5 mm⁻¹ in dentine which is downtrend. Furthermore, the persistence of bleaching oxidation in 35% hydrogen peroxide-induced optical thickness of enamel is similar with unbleached tissue which may indicate the refractive index of enamel is unchanged. Moreover, disruption of the entrance signal (DES) analysis showed that remarkable difference was appeared at enamel surface. The results indicate that optical properties of oxidative bleaching human dental tissue can be determined by attenuation coefficient using OCT system.     

Rotational dynamics of gold nanoparticle chains in water solution

In this study, the optical properties of two nano-sized polymer colloids in optical coherence tomography (OCT) were compared in vitro with respect to their potential use as contrast agents. We used two types of particles: compact hydrophobic spherical polystyrene (PS) particles and soft water-swollen nanogel (NG) particles both with grafted hydrophilic shell, both prepared at two different sizes (PS at 300 and 150?nm, NG at 300 and 200?nm). The OCT backscattering signals of the particles in a vessel-mimicking highly scattering agar/TiO₂ phantom were compared on either number of particles or weight percent. Larger particles and higher concentrations produced higher OCT contrast. At each concentration tested, a markedly higher contrast was achieved by PS particles than NG particles. PS particles generated a markedly higher OCT contrast than the phantom at concentrations of at least 1?×?10¹⁰ or 0.1?% for PS 300?nm and at least 3?×?10¹¹ particles/mL or 0.4?% for PS 150?nm. The contrast generated by NG 300?nm was above the phantom contrast at concentrations of at least 3?×?10¹¹ particles/mL or 1?%, whereas NG 200?nm only at 4?%. At any given weight percent, the differences in OCT contrast between differently sized particles were much less evident than in the comparison based on particle number. PS 300?nm generated also a good contrast ex vivo on chicken muscle tissue. These results strongly suggest that PS spheres have strong potential as intravascular OCT contrast agent, while NG particles need further contrast enhancer for being used as OCT contrast agent.     

Multilevel Deep Feature Generation Framework for Automated Detection of Retinal Abnormalities Using OCT Images

Optical coherence tomography (OCT) images coupled with many learning techniques have been developed to diagnose retinal disorders. This work aims to develop a novel framework for extracting deep features from 18 pre-trained convolutional neural networks (CNN) and to attain high performance using OCT images. In this work, we have developed a new framework for automated detection of retinal disorders using transfer learning. This model consists of three phases: deep fused and multilevel feature extraction, using 18 pre-trained networks and tent maximal pooling, feature selection with ReliefF, and classification using the optimized classifier. The novelty of this proposed framework is the feature generation using widely used CNNs and to select the most suitable features for classification. The extracted features using our proposed intelligent feature extractor are fed to iterative ReliefF (IRF) to automatically select the best feature vector. The quadratic support vector machine (QSVM) is utilized as a classifier in this work. We have developed our model using two public OCT image datasets, and they are named database 1 (DB1) and database 2 (DB2). The proposed framework can attain 97.40% and 100% classification accuracies using the two OCT datasets, DB1 and DB2, respectively. These results illustrate the success of our model.     

Adaptive-optics ultrahigh-resolution optical coherence tomography

Merging of ultrahigh-resolution optical coherence tomography (UHR OCT) and adaptive optics (AO), resulting in high axial (3 microm) and improved transverse resolution (5-10 microm) is demonstrated for the first time to our knowledge in in vivo retinal imaging. A compact (300 mm x 300 mm) closed-loop AO system, based on a real-time Hartmann-Shack wave-front sensor operating at 30 Hz and a 37-actuator membrane deformable mirror, is interfaced to an UHR OCT system, based on a commercial OCT instrument, employing a compact Ti:sapphire laser with 130-nm bandwidth. Closed-loop correction of both ocular and system aberrations results in a residual uncorrected wave-front rms of 0.1 microm for a 3.68-mm pupil diameter. When this level of correction is achieved, OCT images are obtained under a static mirror configuration. By use of AO, an improvement of the transverse resolution of two to three times, compared with UHR OCT systems used so far, is obtained. A significant signal-to-noise ratio improvement of up to 9 dB in corrected compared with uncorrected OCT tomograms is also achieved.     

An Approach to Optical Reflection Tomography along the Geometrical Thickness

We propose and demonstrate a novel optical reflection tomography along the geometrical thickness. This technique is based on simultaneous measurement of refractive index n and thickness t of a sample using the combination of a low coherence interferometer and confocal optics. The interferometer provides optical coherence tomography (OCT) of the dimension of the optical thickness (=n × t) along the optical axis, while the confocal optics gives us another type of reflection tomography, having the thickness dimension of nearly t/n along the optical axis. This sort of tomography can be called confocal reflection tomography (CRT) and has not yet been demonstrated, to our knowledge. Simple image processing of OCT and CRT results in the desired reflection tomographic image, showing two-dimensional refractive index distribution along the geometrical thickness.     

Range Technique in Scattering Medium Using a Needle-Fiber Optical Coherence Tomography System

A needle-fiber optical coherence tomography (OCT) system designed for guiding epidural anesthesia to a target is presented. In the preliminary experiment, a desired target in tofu as a strong scattering medium can be monitored over a sufficient range of several millimeters within a short measuring time. The axial resolution is 21 μm. In order to investigate the ranging performance, the relation between background noise levels and focus-scattering properties in a confocal optical system is numerically studied by use of Monte Carlo technique. Numerical results indicate that an appropriate focal length of the lens attached to the fiber should be used to separate the OCT signal from strong noise in strongly scattered medium.     

基于受激辐射信号的谱域光学相干层析分子成像方法

鉴于不同生理病理状态下组织复折射率实部的变化不大,传统光学相干层析(OCT)成像技术在分子特异性识别上存在先天不足.为此,本文提出了基于受激辐射信号的OCT成像方法,可在实现传统散射成像的同时,实现基于受激辐射信号的分子成像.在超高分辨率谱域OCT系统的基础上,通过增设光谱分光与调制抽运光支路,建立了基于单宽谱光源的抽运探测谱域OCT系统,详细推导了调制抽运下受激辐射信号的获取与成像公式.利用搭建的抽运探测谱域OCT系统,实现了瞬态受激辐射信号的相干探测.基于同时获取的受激辐射OCT信号和传统OCT信号,成功重构了氮化物粉末构建样品的基于受激辐射信号的分子对比OCT图像.     

Simultaneous optical coherence tomography--Indocyanine Green dye fluorescence imaging system for investigations of the eye's fundus

We have developed a dual-channel optical coherence tomography-Indocyanine Green dye (OCT-ICG) fluorescence system based on a previously reported ophthalmic OCT confocal imaging system. The confocal channel is tuned to the fluorescence wavelength range of the ICG, and light from the same optical source is used to generate the OCT image and to excite the ICG fluorescence. The system enables the clinician to visualize simultaneously en face OCT slices and corresponding ICG angiograms of the ocular fundus, displayed side by side. C-scan (constant depth) and B-scan (cross section) images are collected by a fast en face scan (T scan). The pixel-to-pixel correspondence between the OCT and angiography images allows the user to capture OCT B scans precisely at selected points on the ICG confocal images.     

Tracking optical coherence tomography

An experimental tracking optical coherence tomography (OCT) system has been clinically tested. The prototype instrument uses a secondary sensing beam and steering mirrors to compensate for eye motion with a closed-loop bandwidth of 1 kHz and tracking accuracy, to within less than the OCT beam diameter. The retinal tracker improved image registration accuracy to <1 transverse pixel (<60 microm). Composite OCT images averaged over multiple scans and visits show a sharp fine structure limited only by transverse pixel size. As the resolution of clinical OCT systems improves, the capability to reproducibly map complex structures in the living eye at high resolution will lead to improved understanding of disease processes and improved sensitivity and specificity of diagnostic procedures.     

Ultra-high resolution spectral domain optical coherence tomography using supercontinuum light source

An ultra-high resolution spectral domain optical coherence tomography (SD-OCT) was developed using a cost-effective supercontinuum laser. A spectral filter consists of a dispersive prism, a cylindrical lens and a right-angle prism was built to transmit the wavelengths in range 680–940 nm to the OCT system. The SD-OCT has achieved 1.9 μm axial resolution and the sensitivity was estimated to be 91.5 dB. A zero-crossing fringes matching method which maps the wavelengths to the pixel indices of the spectrometer was proposed for the OCT spectral calibration. A double sided foam tape as a static sample and the tip of a middle finger as a biological sample were measured by the OCT. The adhesive and the internal structure of the foam of the tape were successfully visualized in three dimensions. Sweat ducts was clearly observed in the OCT images at very high resolution. To the best of our knowledge, this is the first demonstration of ultra-high resolution visualization of sweat duct by OCT.     

Quadrature fringes wide-field optical coherence tomography and its applications to biological tissues

We propose and demonstrate quadrature fringes wide-field optical coherence tomography (QF WF OCT) to expand an optical Hilbert transformation to two-dimensions. This OCT simultaneously measures two quadrature interference images using a single InGaAs CCD camera to obtain en face OCT images. The axial and lateral resolutions are measured at 29 μm in air and 70 μm limited by a pixel size of camera using a superluminescent diode with a wavelength of 1.3 μm as the light source; the system sensitivity is determined to be −90 dB. The area of the en face OCT images is 4.0 mm × 4.0 mm (160 × 160  pixels). The OCT images are measured axially with steps of 10 μm. The en face OCT images of a in vivo human fingertip and a in situ rat brain are three-dimensionally measured up to the depth of about 3 mm with some degradations of a lateral resolution.