(E-181) Distinguishing blood vessels with spectroscopic vis-OCT

Introduction:: Vision is a highly complex task, with the retina having a large metabolic demand. Major retinal diseases, including diabetic retinopathy (DR), have been indicated to alter the metabolism of the eye. Diabetic Retinopathy (DR) is a leading cause of irreversible blindness in the United States, and one in three adults over 40 with diabetes is diagnosed with some form of DR [1]. Consequently, accurate measurement of blood oxygen saturation (sO2) is a potential biomarker for the management of many retinal pathologies. Using spectroscopic visible-light optical coherence tomography (vis-OCT), we identified patterns in the vis-OCT amplitude as a function of wavelengths offering optimal contrast between arteries, veins, and retinal tissue. This investigation will assist in the design of methods for measuring oxygen metabolism.


Materials and Methods:: We imaged three C57BL/6 mice with a custom retinal vis-OCT system. Our vis-OCT system has an axial resolution of ~1.3µm in tissue. Each volumetric scan was acquired with a temporal speckle averaging pattern where each lateral position was repeatedly imaged twice per volume and three volumes were acquired. Using a Gaussian STFT window, we captured the spectroscopic vis-OCT signal between the wavelengths of 535nm to 587nm in increments of 4nm. To examine differences in spectroscopic patterns between vessels and retinal tissue, we segmented major arteries and veins in our vis-OCT scans (fig. 1). As the measurement of oxygenation depends on the depth-dependent decay of the blood signal at different wavelengths, we compared the mean signal of vessels at different wavelengths and depths in the retina. We can fit the attenuation at each wavelength and compare it to the theoretical spectrum for oxyhemoglobin and deoxyhemoglobin [2]. To normalize depth, the surface of the retina was flattened to the internal limiting membrane of the retina (figure 2a). The vitreous/inner membrane boundary was segmented using thresholding [3]. To compare the signal at two wavelengths, we divided the mean spectroscopic signal at a given depth from different wavelengths. We calculated the mean signal from arteries and veins relative to background tissue and calculated a signal-to-noise ratio (SNR). This comparison was done for every wavelength band from 535nm to 587nm and depths from the inner limiting membrane to 75µm below the inner limiting membrane.


Results, Conclusions, and Discussions:: The ratio of wavelengths mean vis-OCT signal at 539nm and 567nm for depths from the inner limiting membrane to 75µm was found to be different for arteries and veins (p=0.0016, α=0.01) in (Figure 2b). Conversely, the ratio of signal from 535nm and 587nm was not a statistically significant difference (Figure 2c).

The wavelengths 539nm and 567nm showed high discrepancy in SNR in the veins and arteries due to the differences in the absorption coefficient for oxyhemoglobin and deoxyhemoglobin at these wavelengths. At 539nm, this value is 52,474 [cm-1/(moles/liter)] in oxyhemoglobin and 45,842 [cm-1/(moles/liter)] in deoxyhemoglobin [4]. At 567nm, this value is 38,333.6 [cm-1/(moles/liter)] in oxyhemoglobin and 47,888 [cm-1/(moles/liter)] in deoxyhemoglobin [4]. Thus, because of the large difference of the absorption coefficient in oxyhemoglobin, we detected a difference in the SNR between the venous and arterial blood vessels at the ratio between these wavelengths. Thus, the ratio of the spectroscopic signal between 539nm and 567nm can be used to differentiate arteries and veins in vis-OCT images.



The results show that using the ratio of the spectroscopic at wavelengths 539nm and 567nm distinguishes veins and arteries well. We conclude that the spectroscopic vis-OCT signal at different wavelengths may assist in the evaluation of retinal pathologies.




Acknowledgements (Optional): :

References (Optional): : [1] National Center for Chronic Disease Prevention and Health Promotion. “Diabetic Retinopathy.” Retrieved July 21, 2023, from https://www.cdc.gov/visionhealth/pdf/factsheet.pdf (2022)



[2] W. Liu et al., “Increased Retinal Oxygen Metabolism Precedes Microvascular Alterations in Type 1 Diabetic Mice,” Invest. Ophthalmol. Vis. Sci., vol. 58, no. 2, pp. 981–989, Feb. 2017, doi: 10.1167/iovs.16-20600.



[3] V. J. Srinivasan et al., “Characterization of Outer Retinal Morphology with High-Speed, Ultrahigh-Resolution Optical Coherence Tomography,” Invest. Ophthalmol. Vis. Sci., vol. 49, no. 4, pp. 1571–1579, Apr. 2008, doi: 10.1167/iovs.07-0838.



[4] “Tabulated Molar Extinction Coefficient for Hemoglobin in Water.” https://omlc.org/spectra/hemoglobin/summary.html (accessed Jul. 21, 2023).