Optimization-based vessel segmentation pipeline for robust quantification of capillary networks in skin with optical coherence tomography angiography

Abstract

Optical coherence tomography angiography (OCTA) provides in-vivo images of microvascular perfusion in high resolution. For its application to basic and clinical research, an automatic and robust quantification of the capillary architecture is mandatory. Only this makes it possible to reliably analyze large amounts of image data, to establish biomarkers, and to monitor disease developments. However, due to its optical properties, OCTA images of skin often suffer from a poor signal-to-noise ratio and contain imaging artifacts. Previous work on automatic vessel segmentation in OCTA mostly focuses on retinal and cerebral vasculature. Its applicability to skin and, furthermore, its robustness against imaging artifacts had not been systematically evaluated. We propose a segmentation method that improves the quality of vascular quantification in OCTA images even if corrupted by imaging artifacts. Both the combination of image processing methods and the choice of their parameters are systematically optimized to match the manual labeling of an expert for OCTA images of skin. The efficacy of this optimization-based vessel segmentation is further demonstrated on sample images as well as by a reduced error of derived quantitative vascular network characteristics.

Assessing the impact of aging and blood pressure on dermal microvasculature by reactive hyperemia optical coherence tomography angiography

“…2 ) captured the baseline perfusion of the microvascular network. The microvasculature was automatically segmented and analyzed by quantifying the vessel density (VD), length of the vascular network (LVN), and number of branch points (NBP) 32 . At the peak of reactive hyperemia perfusion (Figs.…”

“…All data sets were deidentified and automatically segmented vessels (Fig. 2 ) were analyzed by various quantitative metrics described in detail in earlier works 17 , 32 . In summary, the vessel density (VD) is based on the area of segmented vessels (red pixels) as a percentage of the entire image (Fig.…”

OCTAVA: An open-source toolbox for quantitative analysis of optical coherence tomography angiography images

“…Other works have employed a Frangi filter to improve visibility of vessels in OCTA images of the retina [ 49 ], choroid [ 67 ], and skin [ 41 ]. Despite its widespread use, the Frangi filter has also been criticized in some studies for introducing errors in the vessel architecture, either by missing vessels rendered in the image with low SNR or generating spurious vessels depending on the structure of the background noise [ 68 , 69 ]. Even using the multiscale Frangi filter approach, it has been shown that the results are highly dependent on the range of vessel sizes within the image [ 70 ].…”

“…Motion artifacts are a main limiting factor in any segmentation-based analysis because the algorithm does not differentiate artifacts from vessels. Casper et al have recently proposed a segmentation technique which incorporates a refinement step to help correct for defocus and blurring caused by motion, which may prove helpful [ 69 ]. In any case, if motion artifact exceeds a certain level (such as the example in Fig 8A and 8B ), there is no choice but to exclude the images from analysis.…”

Generation of Augmented Capillary Network Optical Coherence Tomography Image Data of Human Skin for Deep Learning and Capillary Segmentation

“…Research has found that the presence of skin cancers often coincides with an abnormal proliferation of the surrounding blood vessel system [ 4 , 17 , 18 ], which delivers oxygen and nutrients that support the growth of the malignancies. FF-OCT provides the capability of resolving vessels and capillaries in high-resolution images [ 14 , 19 , 20 ], which could help physicians track the maturity of tumors and the effect of drug delivery, allowing them to examine the evolution of the nearby angiogenesis. Past studies [ 4 , 21 , 22 ] have also demonstrated that inhibiting angiogenesis is a viable approach for cancer therapy.…”