Application of Improved Homogeneity Similarity-Based Denoising in Optical Coherence Tomography Retinal Images

Chen, Qiang; Sisternes, Luís de; Leng, Theodore; Rubin, Daniel L.

doi:10.1007/s10278-014-9742-8

Cited by 14 publications

(7 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The differences in refractive optics, axial length, retinal tissue composition (e.g., nerve fiber/nuclear/plexiform layer contributions), and optical scattering between the human and porcine eye might contribute to disparities in OCT-histology measurements. In addition, variable lateral/horizontal image stretching or distortion might occur with SD-OCT algorithmic image processing (Chen et al, 2015; Folgar et al, 2014; Garcia Garrido et al, 2015; Podoleanu et al, 2004; Uji et al, 2017; Uji et al, 2015).…”

Section: Discussionmentioning

confidence: 99%

Correlation of spectral domain optical coherence tomography with histology and electron microscopy in the porcine retina

Xie

Zhao

Tsai

et al. 2018

Experimental Eye Research

View full text Add to dashboard Cite

Spectral domain optical coherence tomography (SD-OCT) is used as a non-invasive tool for retinal morphological assessment in vivo. Information on the correlation of SD-OCT with retinal histology in the porcine retina, a model resembling the human retina, is limited. Herein, we correlated the hypo- and hyper-reflective bands on SD-OCT with histology of the lamellar architecture and cellular constituents of the porcine retina. SD-OCT images were acquired with the Heidelberg Spectralis HRA + OCT. Histological analysis was performed using epoxy resin embedded tissue and transmission electron microscopy. Photomicrographs from the histologic sections were linearly scaled to correct for tissue shrinkage and correlated with SD-OCT images. SD-OCT images correlated well with histomorphometric data. A hyper-reflective band in the mid-to-outer inner nuclear layer correlated with the presence of abundant mitochondria in horizontal cell processes and adjacent bipolar cells. A concentration of cone nuclei corresponded to a relative hypo-reflective band in the outer portion of the outer nuclear layer. The presence of 3 hyper-reflective bands in the outer retina corresponded to: 1) the external limiting membrane; 2) the cone and rod ellipsoid zones; and 3) the interdigitation zone of photoreceptor outer segments/retinal pigment epithelium (RPE) apical cell processes and the RPE. These correlative and normative SD-OCT data may be employed to characterize and assess the in vivo histologic changes in retinal vascular and degenerative diseases and the responses to novel therapeutic interventions in this large animal model.

show abstract

Section: Discussionmentioning

confidence: 99%

Correlation of spectral domain optical coherence tomography with histology and electron microscopy in the porcine retina

Xie

Zhao

Tsai

et al. 2018

Experimental Eye Research

View full text Add to dashboard Cite

show abstract

“…To segment the HRFs from the SD OCT image, we developed a novel algorithm, which first segments both the nerve fiber layer (NFL) and IS/OS boundaries and then constructs a narrow-band including HRFs. The details of this algorithm are as follows: A bilateral filter is applied to reduce image noise before follow-up processing 30 . Considering that NFL and OS have similar intensity with HRF, we intend to segment the down boundary of the NFL and OS layer to limit the regions where foci locate.…”

Section: Methodsmentioning

confidence: 99%

Multimodality analysis of Hyper-reflective Foci and Hard Exudates in Patients with Diabetic Retinopathy

Niu

Chen

et al. 2017

Sci Rep

Self Cite

View full text Add to dashboard Cite

To investigate the correlations between hyper-reflective foci and hard exudates in patients with non-proliferative diabetic retinopathy (NPDR) and proliferative diabetic retinopathy (PDR) by spectral-domain optical coherence tomography (SD OCT) images. Hyper-reflective foci in retinal SD OCT images were automatically detected by the developed algorithm. Then, the cropped CFP images generated by the semi-automatic registration method were automatically segmented for the hard exudates and corrected by the experienced clinical ophthalmologist. Finally, a set of 5 quantitative imaging features were automatically extracted from SD OCT images, which were used for investigating the correlations of hyper-reflective foci and hard exudates and predicting the severity of diabetic retinopathy. Experimental results demonstrated the positive correlations in area and amount between hard exudates and hyper-reflective foci at different stages of diabetic retinopathy, with statistical significance (all p < 0.05). In addition, the area and amount can be taken as potential discriminant indicators of the severity of diabetic retinopathy.

show abstract

“…One of the challenges in OCT imaging is the presence of speckle noise. To reduce image noise, we applied a bilateral filter for SD‐OCT images. Then, a layer segmentation method based on graph theory and dynamic programming was adopted to segment the NFL/GCL and the IS/OS layers to limit the HRF search regions.…”

Section: Methodsmentioning

confidence: 99%

Hyper‐reflective foci segmentation in SD‐OCT retinal images with diabetic retinopathy using deep convolutional neural networks

Xie

Niu

et al. 2019

Medical Physics

Self Cite

View full text Add to dashboard Cite

Purpose The purpose of this study was to automatically and accurately segment hyper‐reflective foci (HRF) in spectral domain optical coherence tomography (SD‐OCT) images with diabetic retinopathy (DR) using deep convolutional neural networks. Methods An automatic HRF segmentation model for SD‐OCT images based on deep networks was constructed. The model segmented small lesions through pixel‐wise predictions based on small image patches. We used an approach for discriminative features extraction for small patches by introducing small kernels and strides in convolutional and pooling layers, which was applied on the state‐of‐the‐art deep classification networks (GoogLeNet and ResNet). The features extracted by the adapted deep networks were fed into a softmax layer to produce the probabilities of HRF. We trained different models on a dataset with 16 HRF eyes by using different sizes of patches, and then, we fused these models to generate optimal results. Results Experimental results on 18 eyes demonstrated that our method is effective for the HRF segmentation. The dice similarity coefficient (DSC) for the foci area in B‐scan, projection images, and foci amount in B‐scan images reaches 67.81%, 74.09%, and 72.45%, respectively. Conclusions The proposed segmentation model can accurately segment HRF in SD‐OCT images with DR and outperforms traditional methods. Our model may provide reliable segmentations for small lesions in SD‐OCT images and may be helpful in the clinical diagnosis of diseases.

show abstract

Application of Improved Homogeneity Similarity-Based Denoising in Optical Coherence Tomography Retinal Images

Cited by 14 publications

References 36 publications

Correlation of spectral domain optical coherence tomography with histology and electron microscopy in the porcine retina

Correlation of spectral domain optical coherence tomography with histology and electron microscopy in the porcine retina

Multimodality analysis of Hyper-reflective Foci and Hard Exudates in Patients with Diabetic Retinopathy

Hyper‐reflective foci segmentation in SD‐OCT retinal images with diabetic retinopathy using deep convolutional neural networks

Contact Info

Product

Resources

About