Depth-resolved vascular profile features for artery-vein classification in OCT and OCT angiography of human retina

Assessing spectral effectiveness in color fundus photography for deep learning classification of retinopathy of prematurity

Significance

Retinopathy of prematurity (ROP) poses a significant global threat to childhood vision, necessitating effective screening strategies. This study addresses the impact of color channels in fundus imaging on ROP diagnosis, emphasizing the efficacy and safety of utilizing longer wavelengths, such as red or green for enhanced depth information and improved diagnostic capabilities.

Aim

This study aims to assess the spectral effectiveness in color fundus photography for the deep learning classification of ROP.

Approach

A convolutional neural network end-to-end classifier was utilized for deep learning classification of normal, stage 1, stage 2, and stage 3 ROP fundus images. The classification performances with individual-color-channel inputs, i.e., red, green, and blue, and multi-color-channel fusion architectures, including early-fusion, intermediate-fusion, and late-fusion, were quantitatively compared.

Results

For individual-color-channel inputs, similar performance was observed for green channel (88.00% accuracy, 76.00% sensitivity, and 92.00% specificity) and red channel (87.25% accuracy, 74.50% sensitivity, and 91.50% specificity), which is substantially outperforming the blue channel (78.25% accuracy, 56.50% sensitivity, and 85.50% specificity). For multi-color-channel fusion options, the early-fusion and intermediate-fusion architecture showed almost the same performance when compared to the green/red channel input, and they outperformed the late-fusion architecture.

Conclusions

This study reveals that the classification of ROP stages can be effectively achieved using either the green or red image alone. This finding enables the exclusion of blue images, acknowledged for their increased susceptibility to light toxicity.

Functional OCT reveals anisotropic changes of retinal flicker-evoked vasodilation

The purpose of this study is to verify the effect of anisotropic property of retinal biomechanics on vasodilation measurement. A custom-built optical coherence tomography (OCT) was used for time-lapse imaging of flicker stimulation-evoked vessel lumen changes in mouse retinas. A comparative analysis revealed significantly larger (18.21%) lumen dilation in the axial direction compared to the lateral (10.77%) direction. The axial lumen dilation predominantly resulted from the top vessel wall movement toward the vitreous direction, whereas the bottom vessel wall remained stable. This observation indicates that the traditional vasodilation measurement in the lateral direction may result in an underestimated value.

Adaptive vessel tracing and segmentation in OCT enables the robust detection of wall-to-lumen ratio abnormalities in 5xFAD mice

The wall-to-lumen ratio (WLR) of retinal blood vessels promises a sensitive marker for the physiological assessment of eye conditions. However, in vivo measurement of vessel wall thickness and lumen diameter is still technically challenging, hindering the wide application of WLR in research and clinical settings. In this study, we demonstrate the feasibility of using optical coherence tomography (OCT) as one practical method for in vivo quantification of WLR in the retina. Based on three-dimensional vessel tracing, lateral en face and axial B-scan profiles of individual vessels were constructed. By employing adaptive depth segmentation that adjusts to the individual positions of each blood vessel for en face OCT projection, the vessel wall thickness and lumen diameter could be reliably quantified. A comparative study of control and 5xFAD mice confirmed WLR as a sensitive marker of the eye condition.

Alteration of perivascular reflectivity on optical coherence tomography of branched retinal vein obstruction

This study aimed to evaluate perivascular reflectivity in patients with branched retinal vascular obstruction (BRVO) using en-face optical coherence tomography (OCT). The study retrospectively analyzed 45 patients with recurrent BRVO, 30 with indolent BRVO, and 45 age- and sex-matched controls. Using a 3.0 × 3.0-mm deep capillary plexus slab on macular scans, OCT angiography (OCTA) and structural en-face OCT scans were divided into four quadrants. Obstructive quadrants of OCTA scans were binarized using a threshold value of mean + 2 standard deviation. The selected area of high signal strength (HSS) was applied to the structural en-face OCT scans, and the corrected mean perivascular reflectivity was calculated as the mean reflectivity on the HSS area/overall en-face OCT mean reflectivity. The same procedure was performed in the quadrants of the matched controls. Regression analysis was conducted on several factors possibly associated with corrected perivascular reflectivity. The perivascular reflectivity in the obstructive BRVO quadrant was significantly higher than in the indolent BRVO and control quadrants (P = 0.009, P = 0.003). Both univariate and multivariate regression analyses showed a significant correlation between the average number of intravitreal injections (anti-vascular endothelial growth factor or dexamethasone implant) per year and refractive errors and image binarization threshold and perivascular reflectivity (P = 0.011, 0.013, < 0.001/univariate; 0.007, 0.041, 0.005/multivariate, respectively). En-face OCT scans of the deep capillary plexus slab revealed higher perivascular reflectivity in recurrent BRVO eyes than in indolent BRVO and control eyes. The results also indicate a remarkable correlation between perivascular reflectivity and the average number of intravitreal injections, suggesting a link to recurrence rates.

Deep learning for artery-vein classification in optical coherence tomography angiography

Major retinopathies can differentially impact the arteries and veins. Traditional fundus photography provides limited resolution for visualizing retinal vascular details. Optical coherence tomography (OCT) can provide improved resolution for retinal imaging. However, it cannot discern capillary-level structures due to the limited image contrast. As a functional extension of OCT modality, optical coherence tomography angiography (OCTA) is a non-invasive, label-free method for enhanced contrast visualization of retinal vasculatures at the capillary level. Recently differential artery-vein (AV) analysis in OCTA has been demonstrated to improve the sensitivity for staging of retinopathies. Therefore, AV classification is an essential step for disease detection and diagnosis. However, current methods for AV classification in OCTA have employed multiple imagers, that is, fundus photography and OCT, and complex algorithms, thereby making it difficult for clinical deployment. On the contrary, deep learning (DL) algorithms may be able to reduce computational complexity and automate AV classification. In this article, we summarize traditional AV classification methods, recent DL methods for AV classification in OCTA, and discuss methods for interpretability in DL models.

An open-source deep learning network AVA-Net for arterial-venous area segmentation in optical coherence tomography angiography

BACKGROUND

Differential artery-vein (AV) analysis in optical coherence tomography angiography (OCTA) holds promise for the early detection of eye diseases. However, currently available methods for AV analysis are limited for binary processing of retinal vasculature in OCTA, without quantitative information of vascular perfusion intensity. This study is to develop and validate a method for quantitative AV analysis of vascular perfusion intensity.

METHOD

A deep learning network AVA-Net has been developed for automated AV area (AVA) segmentation in OCTA. Seven new OCTA features, including arterial area (AA), venous area (VA), AVA ratio (AVAR), total perfusion intensity density (T-PID), arterial PID (A-PID), venous PID (V-PID), and arterial-venous PID ratio (AV-PIDR), were extracted and tested for early detection of diabetic retinopathy (DR). Each of these seven features was evaluated for quantitative evaluation of OCTA images from healthy controls, diabetic patients without DR (NoDR), and mild DR.

RESULTS

It was observed that the area features, i.e., AA, VA and AVAR, can reveal significant differences between the control and mild DR. Vascular perfusion parameters, including T-PID and A-PID, can differentiate mild DR from control group. AV-PIDR can disclose significant differences among all three groups, i.e., control, NoDR, and mild DR. According to Bonferroni correction, the combination of A-PID and AV-PIDR can reveal significant differences in all three groups.

CONCLUSIONS

AVA-Net, which is available on GitHub for open access, enables quantitative AV analysis of AV area and vascular perfusion intensity. Comparative analysis revealed AV-PIDR as the most sensitive feature for OCTA detection of early DR. Ensemble AV feature analysis, e.g., the combination of A-PID and AV-PIDR, can further improve the performance for early DR assessment.

Integrating a pressure sensor with an OCT handheld probe to facilitate imaging of microvascular information in skin tissue beds

Optical coherence tomography (OCT) and OCT angiography (OCTA) have been increasingly applied in skin imaging applications in dermatology, where the imaging is often performed with the OCT probe in contact with the skin surface. However, this contact mode imaging can introduce uncontrollable mechanical stress applied to the skin, inevitably complicating the interpretation of OCT/OCTA imaging results. There remains a need for a strategy for assessing local pressure applied on the skin during imaging acquisition. This study reports a handheld scanning probe integrated with built-in pressure sensors, allowing the operator to control the mechanical stress applied to the skin in real-time. With real time feedback information, the operator can easily determine whether the pressure applied to the skin would affect the imaging quality so as to obtain repeatable and reliable OCTA images for a more accurate investigation of skin conditions. Using this probe, imaging of palm skin was used in this study to demonstrate how the OCTA imaging would have been affected by different mechanical pressures ranging from 0 to 69 kPa. The results showed that OCTA imaging is relatively stable when the pressure is less than 11 kPa, and within this range, the change of vascular area density calculated from the OCTA imaging is below 0.13%. In addition, the probe was used to augment the OCT monitoring of blood flow changes during a reactive hyperemia experiment, in which the operator could properly control the amount of pressure applied to the skin surface and achieve full release after compression stimulation.

MF-AV-Net: an open-source deep learning network with multimodal fusion options for artery-vein segmentation in OCT angiography

This study is to demonstrate the effect of multimodal fusion on the performance of deep learning artery-vein (AV) segmentation in optical coherence tomography (OCT) and OCT angiography (OCTA); and to explore OCT/OCTA characteristics used in the deep learning AV segmentation. We quantitatively evaluated multimodal architectures with early and late OCT-OCTA fusions, compared to the unimodal architectures with OCT-only and OCTA-only inputs. The OCTA-only architecture, early OCT-OCTA fusion architecture, and late OCT-OCTA fusion architecture yielded competitive performances. For the 6 mm×6 mm and 3 mm×3 mm datasets, the late fusion architecture achieved an overall accuracy of 96.02% and 94.00%, slightly better than the OCTA-only architecture which achieved an overall accuracy of 95.76% and 93.79%. 6 mm×6 mm OCTA images show AV information at pre-capillary level structure, while 3 mm×3 mm OCTA images reveal AV information at capillary level detail. In order to interpret the deep learning performance, saliency maps were produced to identify OCT/OCTA image characteristics for AV segmentation. Comparative OCT and OCTA saliency maps support the capillary-free zone as one of the possible features for AV segmentation in OCTA. The deep learning network MF-AV-Net used in this study is available on GitHub for open access.