Optimum wavelength combinations for retinal vessel oximetry

Sensitivity of visible range multi-wavelength algorithms for retinal tissue oximetry to acquisition parameters

This study examined the sensitivity of broadband spectroscopy algorithms for retinal tissue oximetry to spectral acquisition parameters. Monte Carlo simulations were conducted on a 4-layer retinal model to assess the impact of various parameters. The optimal spectral range for accurate measurements was determined to be 530 nm to 585 nm. Decreased spectral resolution below 4 nm significantly reduced accuracy. Using an acquisition area larger than the blood vessel resulted in an underestimation of oxygen saturation, especially for high values. A threshold was observed where increased light intensity had no significant impact on measurement variability. The study highlights the importance of informed parameter selection for accurately assessing retinal microcapillary oxygenation and studying local hemodynamics.

Robust non-contact peripheral oxygenation saturation measurement using smartphone-enabled imaging photoplethysmography

We propose a robust non-contact method to accurately estimate peripheral oxygen saturation (SpO₂) using a smartphone-based imaging photoplethysmography. The method utilizes the built-in color camera as a remote sensor and the built-in flashlight as illumination to estimate the SpO₂. Following the ratio of ratios between green and red channels, we introduce a multiple linear regression algorithm to improve the SpO₂ estimation. The algorithm considers the ratio of ratios and the reflectance images recorded at the RGB channels during a calibration process to obtain a set of weighting coefficients to weigh each contributor to the final determination of SpO₂. We demonstrate the proposed smartphone-based method of estimating the SpO₂ on five healthy volunteers whose arms are conditioned by a manual pressure cuff to manipulate the SpO₂ between 90∼100% as detected simultaneously by a medical-grade pulse oximeter. Experimental results indicate that the overall estimated error between the smartphone and the reference pulse oximeter is 0.029 ± 1.141%, leading to a 43% improvement over the conventional ratio of ratios method (0.008 ± 2.008%). In addition, the data sampling time in the current method is 2 seconds, similar to the sampling cycle used in the commercial medical-grade pulse oximeters.

The Various Oximetric Techniques Used for the Evaluation of Blood Oxygenation

Adequate oxygen delivery to a tissue depends on sufficient oxygen content in arterial blood and blood flow to the tissue. Oximetry is a technique for the assessment of blood oxygenation by measurements of light transmission through the blood, which is based on the different absorption spectra of oxygenated and deoxygenated hemoglobin. Oxygen saturation in arterial blood provides information on the adequacy of respiration and is routinely measured in clinical settings, utilizing pulse oximetry. Oxygen saturation, in venous blood (SvO₂) and in the entire blood in a tissue (StO₂), is related to the blood supply to the tissue, and several oximetric techniques have been developed for their assessment. SvO₂ can be measured non-invasively in the fingers, making use of modified pulse oximetry, and in the retina, using the modified Beer–Lambert Law. StO₂ is measured in peripheral muscle and cerebral tissue by means of various modes of near infrared spectroscopy (NIRS), utilizing the relative transparency of infrared light in muscle and cerebral tissue. The primary problem of oximetry is the discrimination between absorption by hemoglobin and scattering by tissue elements in the attenuation measurement, and the various techniques developed for isolating the absorption effect are presented in the current review, with their limitations.

Convolutional Neural Networks for Spectroscopic Analysis in Retinal Oximetry

Retinal oximetry is a non-invasive technique to investigate the hemodynamics, vasculature and health of the eye. Current techniques for retinal oximetry have been plagued by quantitatively inconsistent measurements and this has greatly limited their adoption in clinical environments. To become clinically relevant oximetry measurements must become reliable and reproducible across studies and locations. To this end, we have developed a convolutional neural network algorithm for multi-wavelength oximetry, showing a greatly improved calculation performance in comparison to previously reported techniques. The algorithm is calibration free, performs sensing of the four main hemoglobin conformations with no prior knowledge of their characteristic absorption spectra and, due to the convolution-based calculation, is invariable to spectral shifting. We show, herein, the dramatic performance improvements in using this algorithm to deduce effective oxygenation (SO₂), as well as the added functionality to accurately measure fractional oxygenation (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${{\bf{SO}}}_{{\bf{2}}}^{{\boldsymbol{f}}{\boldsymbol{r}}}$$\end{document}SO2fr). Furthermore, this report compares, for the first time, the relative performance of several previously reported multi-wavelength oximetry algorithms in the face of controlled spectral variations. The improved ability of the algorithm to accurately and independently measure hemoglobin concentrations offers a high potential tool for disease diagnosis and monitoring when applied to retinal spectroscopy.

A Dye-Free Analog to Retinal Angiography Using Hyperspectral Unmixing to Retrieve Oxyhemoglobin Abundance

Purpose

Retinal angiography evaluates retinal and choroidal perfusion and vascular integrity and is used to manage many ophthalmic diseases, such as age-related macular degeneration. The most common method, fluorescein angiography (FA), is invasive and can lead to untoward effects. As an emerging replacement, noninvasive OCT angiography (OCTA) is used regularly as a dye-free substitute with superior resolution and additional depth-sectioning abilities; however, general trends in FA as signified by varying intensity in images are not always reproducible in the fine structural detail in an OCTA image stack because of the source of their respective signals, OCT speckle decorrelation versus fluorescein emission.

Methods

We present a noninvasive/dye-free analog to angiography imaging using retinal hyperspectral imaging with a nonscanning spectral imager, the image mapping spectrometer (IMS), to reproduce perfusion-related data based on the abundance of oxyhemoglobin (HbO₂) in the retina. With a new unmixing procedure of the IMS-acquired spectral data cubes (350 × 350 × 43), we produced noninvasive HbO₂ maps unmixed from reflectance spectra.

Results

Here, we present 15 HbO₂ maps from seven healthy and eight diseased retinas and compare these maps with corresponding FA and OCTA results with a discussion of each technique.

Conclusions

Our maps showed visual agreement with hypo- and hyperfluorescence trends in venous phase FA images, suggesting that our method provides a new use for hyperspectral imaging as a noninvasive angiography-analog technique and as a complementary technique to OCTA.

Translational Relevance

The application of hyperspectral imaging and spectral analysis can potentially improve/broaden retinal disease screening and enable a noninvasive technique, which complements OCTA.

A multispectral microscope for in vivo oximetry of rat dorsal spinal cord vasculature

Quantification of blood oxygen saturation (SO₂) in vivo is essential for understanding the pathogenesis of diseases in which hypoxia is thought to play a role, including inflammatory disorders such as multiple sclerosis (MS) and rheumatoid arthritis (RA). We describe a low-cost multispectral microscope and oximetry technique for calibration-free absolute oximetry of surgically exposed blood vessels in vivo. We imaged the vasculature of the dorsal spinal cord in healthy rats, and varied inspired oxygen (FiO₂) in order to evaluate the sensitivity of the imaging system to changes in SO₂. The venous SO₂ was calculated as 67.8  ±  10.4% (average  ±  standard deviation), increasing to 83.1  ±  11.6% under hyperoxic conditions (100% FiO₂) and returning to 67.4  ±  10.9% for a second normoxic period; the venous SO₂ was 50.9  ±  15.5% and 29.2  ±  24.6% during subsequent hypoxic states (18% and 15% FiO₂ respectively). We discuss the design and performance of our multispectral imaging system, and the future scope for extending this oximetry technique to quantification of hypoxia in inflamed tissue.

Impact of vessel diameter and bandwidth of illumination in sidestream dark-field oximetry

We investigate the possibility of oxygen saturation estimation from images obtained by the sidestream dark-field (SDF) technique. The SDF technique is a method for microvascular imaging. In SDF imaging, light enters a tissue directly from illumination sources configured around a camera and then the camera captures the light scattered by the tissue. To advance the capability of the SDF imaging system, we develop a SDF oximetry method with LED illumination sources. In this paper, we evaluate some SDF oximetry methods from virtual SDF images obtained by the Monte Carlo photon propagation simulation. As a result, we verify that SDF imaging allows the estimation of oxygen saturation of the individual vessels from virtual images using the average extinction coefficients considering the bandwidth of the illumination and the effect of the integration of the camera.

Wavelength selection method with standard deviation: application to pulse oximetry

Near-infrared spectroscopy provides useful biological information after the radiation has penetrated through the tissue, within the therapeutic window. One of the significant shortcomings of the current applications of spectroscopic techniques to a live subject is that the subject may be uncooperative and the sample undergoes significant temporal variations, due to his health status that, from radiometric point of view, introduce measurement noise. We describe a novel wavelength selection method for monitoring, based on a standard deviation map, that allows low-noise sensitivity. It may be used with spectral transillumination, transmission, or reflection signals, including those corrupted by noise and unavoidable temporal effects. We apply it to the selection of two wavelengths for the case of pulse oximetry. Using spectroscopic data, we generate a map of standard deviation that we propose as a figure-of-merit in the presence of the noise introduced by the living subject. Even in the presence of diverse sources of noise, we identify four wavelength domains with standard deviation, minimally sensitive to temporal noise, and two wavelengths domains with low sensitivity to temporal noise.

Retinal Oximetry Using Intravitreal Illumination

PURPOSE

To demonstrate spectroscopic retinal oximetry measurements on arteries and veins in swine using intravitreal illumination. Retinal arterial and venous saturations are measured for a range of inspired O2 levels after pars plana vitrectomy.

METHODS

Pars plana vitrectomy and intravitreal manipulations were performed on two female American Yorkshire domestic swine. Light from a scanning monochromator was coupled into a fiberoptic intraocular illuminator inserted into the vitreous. The retinal vessels were illuminated obliquely, minimizing vessel glints. Multispectral images of the retinal vasculature were obtained as the swine's arterial blood oxygen saturation was decreased from 100% to 67% in decrements of approximately 10%. Retinal vessel spectra were used to calculate oxygen saturation in selected arteries and veins. Arterial oxygen saturations were calibrated using blood gas analysis on blood drawn from a Swan-Ganz catheter placed in the femoral artery.

RESULTS

Oblique illumination of retinal vessels using an intravitreal fiberoptic illuminator provided a substantial reduction in the central vessel glint usually seen in fundus images, thus simplifying the analysis of spectral data. The vessel shadows were displaced from the vessel image simplifying the light paths in the eye. Using a full spectral analysis simplified by the light path reductions, we calculated retinal vessel saturations. The reduction of glint allowed for increased accuracy in measuring retinal vessel spectral optical density. Abnormally low retinal venous oxygen saturations were observed shortly after pars plana vitrectomy.

CONCLUSIONS

Retinal oximetry using intravitreal illumination has been demonstrated. As a research tool, intravitreal illumination addresses several difficulties encountered when performing retinal oximetry with transcorneal illumination.

Blood oxyhemoglobin saturation measurements by blue-green spectral shift

Previous work describing a resilient method for measuring oxyhemoglobin saturation using the blue-green spectral shift was performed using cell free hemoglobin solutions. Hemoglobin solution and whole blood sample spectra measured under similar conditions in a spectrophotometer are used here to begin evaluating the impact of cellular scattering on this method. The blue-green spectral shift with changing oxyhemoglobin saturation was preserved in these blood samples and the blue-green spectral shift was relatively unaffected by physiological changes in blood pH (6.6, 7.1, and 7.4), path length through blood (100 and 200 microm), and blood hematocrit (19 to 48%). The packaging of hemoglobin in red blood cells leads to a decreased apparent path length through hemoglobin, and an overall decrease in scattering loss with increasing wavelength from 450 to 850 nm. The negative slope of the scattering loss in the 476 to 516-nm range leads to a +3.0 nm shift in the oxyhemoglobin saturation calibration line when the blue-green spectral minimum in these blood samples was compared to cell free hemoglobin. Further research is needed to fully evaluate the blue green spectral shift method in cellular systems including in vivo testing.

Snapshot hyperspectral imaging in ophthalmology

Retinal imaging spectroscopy can provide functional maps using chromophore spectra. For example, oxygen saturation maps show ischemic areas from diabetes and venous occlusions. Obtaining retinal spatial-spectral data has been difficult due to saccades and long data acquisition times (>5 s). We present a snapshot imaging spectrometer with far-reaching applicability that acquires a complete spatial-spectral image cube in approximately 3 ms from 450 to 700 nm with 50 bands, eliminating motion artifacts and pixel misregistration. Current retinal spectral imaging approaches are incapable of true snapshot operation over a wide spectral range with a large number of spectral bands. Coupled to a fundus camera, the instrument returns true color retinal images for comparison to standard fundus images and for image validation while the patient is still dilated. Oxygen saturation maps were obtained with a three-wavelength algorithm: for healthy subjects arteries were approximately 95% and veins 30 to 35% less. The instrument is now undergoing clinical trials.

Oxyhemoglobin saturation measurements by green spectral shift

From an analysis of new hemoglobin solution transmission spectra at various oxygen saturations (SO2), path lengths, and pH, we find the determination of SO2 by using the classical oximetry technique to be poorly calibrated. We used this data set to develop a proposed method for SO2 determination based on the spectral shift of the hemoglobin transmission minimum between 475 and 510 nm. The method does not require accurate knowledge of hemoglobin extinction coefficients and is linear in relation to SO2 despite changes in path length, pH, or hemoglobin concentration.

Oximetry with a multiple wavelength SLO

Oximetry may be useful for understanding pathologies of the eye. We use a prototype scanning laser ophthalmoscope capable of simultaneous multiple wavelength imaging to record fundus images. The system is run under non-confocal conditions in using slit apertures with a width of up to 600 microm. The laser lines launched into the SLO were the 633 nm line of a HeNe-laser, and the 815 nm line from a tunable (700 to 900 nm) cw Ti:Sapphire-laser. These wavelengths were selected because of their availability and absorption characteristics. The difference in absorption at these two wavelengths is used to assess blood oxygen content. Images were averaged to improve the signal to noise ratio. This simultaneous method of measuring oxygen content may be preferred to other techniques such as sequential SLO imaging at different wavelengths, or spot analysis using a modified fundus camera. The advantages of this technique are that image registration is not required, and a large area of the retina can be assessed concurrently.

Wavelength dependence of the precision of noninvasive optical measurement of oxy-, deoxy-, and total-hemoglobin concentration

The precision of noninvasive optical measurement of the concentration changes in oxy-, deoxy-, and total-hemoglobin depends on wavelength. For estimating the precision, we calculated the noise level of the concentration changes as the uncertainty in measurements using several wavelength pairs of light. Seven laser diodes (664-848 nm) were used simultaneously for spectroscopic measurement of brain activity during finger motor stimulation. We also used the analysis of error propagation from the uncertainty in direct measurements of absorbance changes to estimate indirectly the uncertainty of concentration changes. The measurement of the concentration changes made using an 830/664-nm pair are two times (oxy-Hb) and six times (deoxy-Hb) more precise than those made using an 830/782-nm pair.

Trans-abdominal monitoring of fetal arterial blood oxygenation using pulse oximetry

Pulse oximetry (oxygen saturation monitoring) has markedly improved medical care in many fields, including anesthesiology, intensive care, and newborn intensive care. In obstetrics, fetal heart rate monitoring remains the standard for intrapartum assessment of fetal well being. Fetal oxygen saturation monitoring is a new technique currently under development. It is potentially superior to electronic fetal heart rate monitoring (cardiotocography) because it allows direct assessment of both the fetal oxygen status and fetal tissue perfusion. Here we present the analysis for determining the most optimal wavelength selection for pulse oximetry. The wavelengths we chose as the most optimal are the first in the range of 670-720 nm and the second in the range of 825-925 nm. Further, we discuss the possible systematic errors during our measurements and their contribution to the obtained saturation results. We present feasibility studies for fetal pulse oximetry, monitored noninvasively through the maternal abdomen. Our preliminary experiments show that the fetal pulse can be discriminated from the maternal pulse and thus, in principle, the fetal arterial oxygen saturation can be obtained. We present the methodology for obtaining these data, and discuss the dependence of our measurements on the fetal position with respect to the optode assembly.

Effect of multiple light paths on retinal vessel oximetry

Techniques for noninvasively measuring the oxygen saturation of blood in retinal arteries and veins are reported in the literature, but none have been sufficiently accurate and reliable for clinical use. Addressing the need for increased accuracy, we present a series of oximetric equations that explicitly consider the effects of backscattering by red blood cells and lateral diffusion of light in the ocular fundus. The equations are derived for the specific geometry of a scanning-beam retinal vessel oximeter; however, the results should also be applicable to photographic oximeters. We present in vitro and in vivo data that suggest the validity of these equations.