FIELD

[0001] The present disclosure relates to on line and real-time optic nerve blood oxygenation mapping.

BACKGROUND

[0002] The human eye retina contains neuronal structures that require a consistent supply for nutrients and oxygen to satisfy the metabolism. The alterations of the retina's blood flow are considered as the principal cause for the majority of ocular diseases. The vascular insufficiency of the optic nerve's head plays a significant role in the development of the optical neuropathy glaucomatous [1]. Accordingly, there is a large interest to develop techniques to evaluate blood flux and blood oxygenation in retinal vessels. The retinal structures are visible through the eye optical system and therefore the eye offers a great opportunity to investigate these structures by non-invasive methods and in real time.

[0003] Furthermore, the methods which propose to evaluate noninvasively retinal blood oxygenation can exploit the specific spectral signature differences between hemoglobin and oxyhemoglobin. Several technologies and methods have proposed to measure the retina's blood oxygenation by photometric methods using some specific wavelengths in visible or near infrared spectral zones [3] [5] [6]. These methods are based on theoretical studies [7] suggesting that by taking photometric measurements at three specific wavelengths, it is possible to estimate the blood oxygenation in a sample of blood without calibration. Two companies, Oxymap Inc., Reykjavik and Imedos UG, Jena, propose a commercial apparatus which estimates the blood oxygenation in retinal vessels by means of only two wavelengths of photometric measurements (i.e. 586 nm and 605 nm). These techniques use a preliminary calibration and the measurements for blood oxygenation are limited for large vessels like arteries and veins. The lens absorption related to the age of the patient is considered one of the main factors that affect the retinal oximetry measurements performed by these techniques [8].

[0004] The retinal structures represent a heterogeneous medium with specific optical properties for each retinal location and for each individual. Thus, the present interest for retinal blood oxygenation measurements is to develop mathematical models to estimate blood oxygenation by means of photometric methods under free calibration conditions, for various retinal locations and eye optical properties.

[0005] Modern methods propose to estimate blood oxygenation in retinal vessels with the full spectrum of the reflectometry function, measured by multichannel spectrometers that perform rapid photometric measurements at a large range and at a high number of wavelengths [12] [13].

[0006] Furthermore, the mathematical models developed to explain the full retinal spectrum reflectometry measurements are valuable for understanding the mechanism of the light absorption and reflection on the retinal structures like optic nerve capillaries, veins and arteries.

[0007] The results from these studies have proposed a mathematical Equation (1 ), capable to explain the full reflectometry absorption function measured in the zone of visible wavelengths. Equation (1 ) has been expressed as a linear combination of two terms, SOHb{X) and SHb{X), representing the normalized spectral absorption functions of the hemoglobin and oxyhemoglobin with a term (λ ^'η ) representing the ocular media absorption with scattering. One constant factor (k) has also been included in the model, to balance the data on the vertical scaling. A multi-Gaussian function

r-L

[0008] has been included in the model to compensate for the non-compatibility of the model and the experimental data in the red spectral zone. The multi-Gaussian function compensates for the light scattered on the red blood cells specific to the large retinal vessels that expose a high volume of blood in the optical path.

Ai A h = J?J * .S . , [A H- in , * 5 \ A. ) - m . * A. ^{' "} -+ ^■ m . ■ ^' - - X μ . <.- ^■ \

- (1 )

[0009] The procedure proposes to compute the "m " parameters by applying the model to the spectral absorption data Α(λ) in the spectral zone from 520 nm to 680 nm. The model results also represent a solution for the spectral reflectometry data in the spectral zone from 430 nm to 520 nm. By this method, the blood oxygenation of the retinal vessels can be estimated without a preliminary calibration on the oxygenation scale. The model computation is adapted for measurements obtained in various retinal vessels; the accuracy for the blood oxygenation derivation is not affected by the eye optical properties.

[0010] As well, Equation (1 ) model is useful to understand the mechanism of the light absorption and reflection on the retinal structures like the optic nerve capillaries, veins and arteries.

[0011] The results of this model suggest that in the spectral zone from 430 nm to 590 nm, the reflectometry function in capillaries, veins and arteries contains the light scattered on the frontal configuration of the blood vessel in combination with the light absorbed by the ocular media and the red blood cells near the endothelial zone of blood vessels. In fact, for this spectral zone, the light reflected on the high volume of the red blood cells specific to the large retinal vessels is not significant and the multi-Gaussian function

[0012] from the Equation (1 ) becomes inappropriate. Since the reflectometry absorption function of the optic nerve capillaries, veins and arteries in the spectral zone from 430 nm to 590 nm, are free from light scattered on the red blood cell, the oxyhemoglobin blood content can be derived by a simplified linear model procedure reflected in Equation (2). The latter equation is composed by four terms; two terms for the hemoglobin and oxyhemoglobin absorption, SOHb{X) and respectively SHb{X), and two terms for the non-hemoglobin media absorption, an exponential term (λ ^'η ) for scattering and lens absorption and respectively a vertical scaling (k) factor.

A {A ) = n ₁ ⁱ : _a t i J + « _i ^* s _ia ( i ) + « Λ " * k ₍₂₎

[0013] Using a multichannel setup, it is possible to measure the retinal absorption reflectometry for 200 different wavelengths in the spectral zone from 430 nm to 590 nm at every 200 ms, from one small optic nerve area location representing a 200 μίτι diameter, and then to compute one value of blood oxygenation at every 200 ms.

[0014] The multi-channel technology represents a tool to estimate on-line and real-time retinal blood oxygenation by measurements of retinal absorption reflectometry at numerous wavelengths. [0015] With such technology, it is useful to understand the blood oxygenation state of the optic nerve papilla of glaucoma patients compared to normal blood oxygenation of the optic nerve papilla in healthy patients.

[0016] A study [22] has proposed to explore the optic nerve papilla of normal and glaucomatous patients by blood oxygenation measurements. The results have demonstrated that, for the normal patients, there is a relative homogeneous blood oxygenation in capillary structures of the optic nerve papilla. Furthermore, the glaucomatous papilla could include zones where the blood oxygenation rate is either lower or higher compared to the limit of normal blood oxygenation. These results indicate that a method which illustrates the mapping for blood oxygenation of the optic nerve papilla would be a useful technique to better discriminate the atrophied zones in the optic nerve papilla of glaucomatous patients.

[0017] A possibility to obtain the blood oxygenation mapping of the optic nerve with the multispectral analysis model could be realized by using a hyperspectral imaging technique. Such technique consists in taking images from the optic nerve papilla for a multitude of wavelengths. Also, in conformity with the multispectral model, such technology should collect 200 spectral data from a minimum of 100 x 100 different locations of the papilla at a 500 ms interval. This task represents a high volume of data that the existing computers are unable to collect, solve and store at such a speed.

[0018] A solution to this problem would be to identify a limited number of specific wavelengths of the visible spectral zone so that the blood oxygenation values, calculated from a reduced number of reflectometry measurements at those specific wavelengths, remain equivalent with the blood oxygenation values, calculated from a full spectral reflectometry function. [0019] In fact, the fundamental question for retinal blood oxygenation measurements by photometric methods is to identify the specific wavelengths in relation with the spectral absorption functions of hemoglobin and oxyhemoglobin and to propose free calibration methods and models to derive the blood oxygenation by photometric measurements in relation with these specific wavelengths.

SUMMARY

[0020] The present disclosure relates to a method for differentiating a blood absorption contribution and an absorption contribution by other optical structures of a patient's eye to a full spectral retinal reflectometry absorption function, comprising: performing reflectometry absorption measurement on the vessels of the patient's retina at a first number of wavelengths in the visible spectral area, corresponding to isosbestic points where hemoglobin and oxyhemoglobin present similar absorption coefficients; estimating the absorption contribution by the other optical structures of the patient's eye at each wavelength as a difference between the reflectometry absorption measurement and a weighting parameter that amplifies the hemoglobin absorption coefficient; and adjusting the weighting parameter so that points representing the differences are collinear when representing these points on a logarithmic scale of the wavelengths.

[0021] The present disclosure also relates to a device for differentiating a blood absorption contribution and an absorption contribution by other optical structures of a patient's eye to a full spectral retinal reflectometry absorption function, comprising: means for performing reflectometry absorption measurement on the vessels of the patient's retina at a first number of wavelengths in the visible spectral area, corresponding to isosbestic points where hemoglobin and oxyhemoglobin present similar absorption coefficients; means for estimating the absorption contribution by the other optical structures of the patient's eye at each wavelength as a difference between the reflectometry absorption measurement and a weighting parameter that amplifies the hemoglobin absorption coefficient; and means for adjusting the weighting parameter so that points representing the differences are collinear when representing these points on a logarithmic scale of the wavelengths.

[0022] The foregoing and other objects, advantages and features of the present disclosure will become more apparent upon reading of the following non- restrictive description of illustrative embodiments thereof, given by way of example only with reference to the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In the appended drawings:

[0024] Figure 1 is a graph showing an example of fitting results between measured absorption values and model results by applying the Equation (2) model to (a) all 400 absorption values of a full wavelengths reflectometry function and (b) four absorption values from a wavelengths zone between 520 nm and 580 nm and an absorption value of 447 nm for results validation;

[0025] Figure 2 is a graph showing blood oxygenation values computed by applying the Equation (2) model to (a) all the 400 absorption values of a full wavelengths reflectometry function and (b) five specific absorption values as shown in Figure 1 preserved from the overall reflectometry function;

[0026] Figure 3 is a graph showing collinear results obtained from five absorption values corresponding to the wavelengths for isosbestic points; [0027] Figure 4 is a graph showing blood oxygenation values calculated by a method applied to only eight absorption values at specific wavelengths (four isosbestic and four for maxima differences) selected from a full spectrum reflectometry function; in this figure are shown corresponding blood oxygenation values calculated by using the model of Equation (2) [12] applied to the 400 absorption values of a full spectrum reflectometry function;

[0028] Figure 5 is a schematic flow chart showing the operations of a method for evaluating blood oxygenation in vessels of a patient's retina;

[0029] Figure 6 is a schematic block diagram showing the modules of a device for evaluating blood oxygenation in vessels of a patient's retina;

[0030] Figure 7 is a schematic diagram of a typical design for a hyperspectral fundus camera with a rotating filter wheel setup;

[0031] Figure 8 is a schematic diagram of a design to combine six laser beams in one light beam using beam splitters;

[0032] Figure 9A-9F shows six images of the left eye optic nerve, scaled from the spectral absorption values for six wavelengths (four isosbestic and two non- isosbestic wavelengths) where, in each image, the absorption values were computed by the ratio between the image of a white uniform diffuser of a phantom eye fundus and the reflected image from the patient's optic nerve, to illustrate operation of the method and device for evaluating blood oxygenation in vessels of a patient's retina;

[0033] Figure 10 shows an optic nerve image edited from the blood oxygenation values, computed by the proposed method applied to the absorption values illustrated in Figure 9A-9F, to illustrate the results obtained using the method and device for evaluating blood oxygenation in vessels of a patient's retina; and

[0034] Figure 11 is a simplified block diagram of an example configuration of hardware components forming the blood oxygenation evaluating device.

DESCRIPTION

[0035] There are several wavelengths in the visible spectral area for which the model represented by Equation (2) becomes optimal to calculate blood oxygenation when the reflectometry absorption measurements are performed for these wavelengths. These wavelengths meet the following:

[0036] 1. The wavelengths corresponding to equal optical absorption coefficients (isosbestic) for hemoglobin and oxyhemoglobin;

[0037] 2. The wavelengths corresponding to a local maximum for optical absorption differences between hemoglobin and oxyhemoglobin.

[0038] The present description relates to a method to compute blood oxygenation that uses a reduced number of wavelengths to be measured in the visible spectral area to confirm that oxygenation values, calculated using this method, are equivalent to the oxygenation values calculated by Equation (2) but using numerous spectral values of a full spectrum reflectometry absorption function.

[0039] The present description also proposes to estimate the blood oxygenation in numerous zones of the optical nerve papilla in order to constitute a mapping technology of the blood oxygenation, on line and in real-time, for the optical nerve area. This technology would be useful in clinical ophthalmology to visualize the optic nerve's atrophic zone which is specific for various eye diseases such as glaucoma.

[0040] The first step is to identify specific wavelengths in the visible spectral area so that the reflectometry measurements from these wavelengths allow the reconstitution of the full spectral reflectometry absorption function as measured by the multichannel technique.

[0041] It has been demonstrated that the above Equation (2), able to explain the reflectometry absorption function from the optic nerve capillaries [12], is expressed by a linear combination of four terms:

A (-0 - "Ί ^{:t ( +} * ) + w ^* * * ⁺ « ^:t * (2)

[0042] It has also been demonstrated that applying Equation (2) to a reflectometry absorption function containing absorption values for only five different wavelengths, it is possible to compute a series of mi, m ₂ , m ₃ , m ₄ and n parameters. However, the blood oxygenation values presented in Figure 2 show that, by applying the Equation (2) model to five absorption values of specific wavelengths, the blood oxygenation values could be lower comparatively to the blood oxygenation values computed from full spectrum reflectometry function.

[0043] It is within the scope of the present description to use a limited number of wavelengths and to obtain blood oxygenation values equivalent with the blood oxygenation values computed from full spectrum reflectometry function. The question remains to compute blood oxygenation from a limited number of reflectometry absorption measurements at various specific wavelengths in the spectral visible area, so that the values of blood oxygenation estimated from these limited number of absorption values remain comparable to the values of blood oxygenation estimated by using the whole spectral reflectometry function. Preliminary considerations

[0044] The next point to be considered is related to the sensitivity of the Equation (2) model with respect to the reflectometry absorption measurements at specific wavelengths of the visible spectrum (visible spectral area) in order to derive the hemoglobin and oxyhemoglobin content [2].

[0045] In fact, the Equation (2) model cannot differentiate the amount of the hemoglobin (ml) and oxyhemoglobin (m2) if reflectometry absorption is measured only for the wavelengths where hemoglobin and oxyhemoglobin present similar absorption coefficients (isosbestic points) [3]. Actually, there are many models and techniques that propose photometric measurements at the wavelengths corresponding to the isosbestic points, useful to quantify the light absorption corresponding to the blood volume [9] [10] [1 1].

[0046] Furthermore, the photometric measurements at the isosbestic points are useful to separate the light absorbed by the blood structure from the light absorbed by the structures that do not contain blood in the overall absorption function.

[0047] The Equation (2) model contains two terms for the hemoglobin and oxyhemoglobin absorption (miSOHb(A) and m ₂ SHb( )) and two terms for the non- hemoglobin media absorption (m ₃ ^'n and m ₄ k). To separate the hemoglobin (m _? ) and oxyhemoglobin (m ₂ ) contributions in the overall reflectometry absorption function, one can introduce in the computation model the absorption values for the wavelengths where the ratio between the absorption coefficients of the hemoglobin and oxyhemoglobin represents a local maxima difference. [0048] On the visible spectrum scale, there are seven wavelengths identified as isosbestic points and seven wavelengths where the ratio between the absorption coefficients of the hemoglobin and oxyhemoglobin represents local maxima.

[0049] Previous studies [12] [22] have proposed to calculate the "m " parameters by applying the Equation (2) model to the spectral absorption data Α(λ) from the 520 nm to 680 nm spectral area. Also, these results represent a solution for the absorption data in the spectral zone from 430 nm to 520 nm.

[0050] In the wavelength zone from 520 nm to 680 nm, there are four wavelengths where the ratio between the absorption coefficients of the hemoglobin and oxyhemoglobin represents local maxima, but out of these wavelengths, only three wavelengths present significant absorption differences between hemoglobin and oxyhemoglobin.

[0051] Figure 1 shows an example of fitting results between the measured reflectometry absorption measurements and the model results by applying the Equation (2) model to:

[0052] (a) All the 400 absorption values of a full wavelengths reflectometry absorption function; and

[0053] (b) Four reflectometry absorption measurements from the wavelength zone between 520 nm and 580 nm and an absorption value of 447 nm for the results validation.

[0054] The absorption values for five specific wavelengths, employed to fit the absorption data shown in Figure 1 , were preserved from the overall reflectometry absorption function of the optic nerve papilla obtained with a multi-channel system (400 discrete wavelengths between 420 nm to 680 nm).

[0055] The results illustrated in Figure 1 show that the multi-channel reflectometry absorption function from the optic nerve structures could be reconstituted with the Equation (2) model, applied to only the five reflectometry absorption measurements at specific wavelengths.

[0056] Figure 2 shows the blood oxygenation values computed by applying the Equation (2) model to:

[0057] (a) All the 400 reflectometry absorption measurements of a multi- and full-wavelengths reflectometry absorption function; and

[0058] (b) Five specific reflectometry absorption measurement values as shown in Figure 1 preserved from the overall reflectometry function.

[0059] The results from Figure 2 show that the blood oxygenation values calculated by applying the Equation (2) model to five reflectometry absorption measurements at specific wavelengths are lower compared to the blood oxygenation values calculated by applying the Equation (2) model to a full spectrum multi-channel reflectometry absorption function.

[0060] A question remains to propose a method for blood oxygenation computation and to identify a number of specific wavelengths from the visible spectral area to be used for reflectometry absorption measurements, so that the values for retina blood oxygenation estimated with such a method applied to these absorptions values remain similar to the blood oxygenation values estimated by using a full wavelengths spectral reflectometry absorption function. Method for retinal blood oxygenation computation using a limited number of reflectometry absorption measurements at specific wavelengths

[0061] The Equation (2) model which explains the full spectrum reflectometry absorption function of the optic nerve papilla contains two terms for the hemoglobin and oxyhemoglobin absorption (miSOHb(X) and m ₂ SHb(X)), one term for the non- hemoglobin ocular media absorption and scattering {m ₃ X ⁿ ) and one term for the vertical scaling {m ₄ k).

l (- = ^!* * ⁺ " ^" < *) + « i ^■ -i ^" * ⁺ w * * (2)

[0062] The term for the ocular media absorption and scattering {m ₃ X ⁿ ) is specific for each subject and for each retinal location. A wrong estimation for this term could affect drastically the blood oxygenation measurement values derived from the Equation (2) model.

[0063] Besides, if the term for ocular media absorption and scattering {m ₃ X ⁿ ) would be known or resolved following specific measurements and derivations, the Equation (2) solutions could become an easy task. In this case, solution of the Equation (2) model may be obtained from only three reflectometry absorption measurements at three specific wavelengths.

[0064] Additionally, the Equation (2) model can be simplified when the term for vertical scaling {m ₄ k) is reduced to zero. The term m ₄ k becomes insignificant when the reference light used for computing the retina reflectometry absorption function, is recorded through a phantom eye [12].

[0065] In conclusion, the reflectometry absorption function that characterizes the retinal reflectometry contains two overall components:

[0066] (a) One component is due to the absorption by the blood located in the retina structures. This component is modulated by the hemoglobin and oxyhemoglobin content (miSOHb(X) + m ₂ SHb(A)); and

[0067] (b) A second component is due to the optics of the eye, the structures of the retina, as well as any kind of noise that may be typical for the method of measurement {m ₃ Z ⁿ + m ₄ k).

Method to differentiate the blood absorption contribution from the absorption contribution of the other optical structures of the patient's eye to the full spectral retinal reflectometry absorption function

[0068] Referring to Figure 5, operation 501 , reflectometry absorption measurement is performed on the vessels of the patient's retina at a number, for example between 3 and 7, of wavelengths in the visible spectral area. These wavelengths correspond to isosbestic points where hemoglobin and oxyhemoglobin present similar absorption coefficients.

[0069] Then, in operation 502 of Figure 5, the absorption contribution by the other optical structures of the patient's eye at each wavelength are estimated as a difference between the reflectometry absorption measurement and a weighting parameter that amplifies the hemoglobin absorption coefficient.

[0070] For that purpose, Equation (2) is expressed for only the wavelengths in which the hemoglobin and the oxyhemoglobin present identical absorption coefficients, i.e. Equation (2i) is obtained for the wavelengths of isosbestic points. Equation (2i) contains only one term (m _h ) for a total of hemoglobin and oxyhemoglobin contribution. The parameter m _h = mi+m ₂ represents a term that weights (weighting parameter) the total blood absorption contribution in the reflectometry absorption function.

A(f.. ^' t - iH), *Sii,, t s. fn α ^' " - ui -J:

(2i)

[0071] By reordering the terms, Equation (2i) becomes Equation (2ii).

A( .i) - mh tSs fAi) -? j-k =

(2ii)

[0072] The format of Equation (2ii) lets us understand that the light absorbed for the ocular media absorption and scattering {m ₃ i ⁿ ) can be estimated by the difference between the reflectometry absorption measurements Α(λ,) at the isosbestic wavelengths and a given value m _h (weighting parameter) that amplifies the hemoglobin absorption coefficients.

[0073] A logarithmic transformation of Equation (2i) reveals that the diagram of the points representing the values for the ocular media absorption and scattering {m ₃ i ⁿ ) are collinear when representing these points on a logarithmic scale of wavelengths which correspond to the isosbestic points.

(2iii)

[0074] A conclusion is to adjust, in operation 503 of Figure 5 and as illustrated in Figure 3, the weighting parameter m _h which weights the total blood absorption contribution to the reflectometry function so that the points representing the logarithmic values obtained from the difference between the reflectometry absorption measurements and the amount of the absorption coefficients for the hemoglobin at the isosbestic wavelengths, become collinear in relation with the logarithmic scale of the corresponding wavelengths.

[0075] The regularisation of parameter m _h lets us find the (m ₃ ) and (n) parameters of Equation (2iii). Generally, at least three measurements of the reflectometry absorption at the wavelengths corresponding to isosbestic values are used to evaluate the (m _h ), (m ₃ ) and (n) parameters of Equation (2iii).

Method to evaluate blood oxygenation in vessels of a patient's retina

[0076] The method (Figure 5) to evaluate blood oxygenation in vessels of a patient's retina comprises the method (Operations 501 -503 of Figure 5) to differentiate the blood absorption contribution from the absorption contribution of the other optical structures of the patient's eye to the full spectral retinal reflectometry absorption function.

[0077] In operation 504 of Figure 5, reflectometry absorption measurement is performed on the vessels of the patient's retina at a number of wavelengths (at least one) where the absorption coefficients between the hemoglobin and the oxyhemoglobin present a maximum difference, i.e. reflectometry absorption measurement on the vessels of the patient's retina at a number of wavelengths (at least one) where a ratio between the absorption coefficients of the hemoglobin and oxyhemoglobin presents a local maximum.

[0078] In operation 505 of Figure 5, blood oxygenation values of the vessels of the patient's retina are derived from at least a portion of the reflectometry absorption measurements and the absorption contributions.

[0079] For that purpose, the oxyhemoglobin contribution (blood oxygenation) to the overall reflectometry absorption function is found by resolving Equation (2) using the reflectometry absorption measurement on the vessels of the patient's retina at a number of wavelengths (at least one) where a ratio between the absorption coefficients of the hemoglobin and oxyhemoglobin presents a local maximum. Equation (2) can be solved since {m ₃ X") is know, {m ₄ k) is as mentioned herein above reduced to zero, and the ratio between (miSOHb( )) and (m ₂ SHb(A)) is know.

[0080] In accordance with the above, relatively to the mathematical signification of various elements constituting Equation (2), it is possible to quantify the blood oxygenation by using a retinal reflectometry absorption function estimated at four specific wavelengths. First, for example, three measurements of the reflectometry absorption at wavelengths corresponding to the isosbestic values are used to evaluate the (m _h ), (m ₃ ) and (n) parameters of Equation (2iii). Second, it is possible to derive the oxyhemoglobin blood content (blood oxygenation) from at least one reflectometry absorption measurement at a wavelength where a difference between the absorption coefficients for the hemoglobin and oxyhemoglobin presents a local maximum.

[0081] In Figure 3 are presented collinear results obtained from seven reflectometry absorption measurements corresponding to the wavelengths for isosbestic points for optimal values of the parameter m _h in Equation (2/7/). The seven absorption values have been selected from a full spectrum reflectometry absorption function measured with a multi-channel setup from the optic nerve papilla.

[0082] Figure 4 shows blood oxygenation values calculated using the technique proposed in the present description applied to only six reflectometry absorption measurements at specific wavelengths (three isosbestic and three for maxima differences) selected from a full spectrum reflectometry function. Furthermore, Figure 4 shown the corresponding blood oxygenation values calculated by using the model of Equation (2) [12] applied to the 400 absorption values of a full spectrum reflectometry function.

[0083] The results presented in Figure 4 indicate that the blood oxygenation values estimated by a limited number (six) of reflectometry absorption measurements show a greater variability compared to the blood oxygenation values estimated from 400 values of a full spectrum reflectometry absorption function. These results could be in agreement with earlier studies [14], which have demonstrated that the precision of the blood oxygenation evaluation increases with approximately the square root of the number of measured wavelengths. However, this rule is not fully illustrated by these results because the 400 measured values from a full spectrum reflectometry absorption function do not have the same sensitivity to differentiate the hemoglobin from the oxyhemoglobin. In fact, on the spectral zone from 400 nm to 600 nm, there are seven wavelengths that present a maximum sensitivity to differentiate the hemoglobin from the oxyhemoglobin by a spectral absorption method.

[0084] For all other wavelengths of the visible spectral zone, the sensitivity diminishes so that, in the intervals determined by the seven wavelengths with a maximum sensitivity, there are found other seven wavelengths for which the coefficients of absorption for hemoglobin and oxyhemoglobin are equivalent (isosbestic) and therefore the sensitivity for differentiating the hemoglobin from the oxyhemoglobin is minimal. However, the absorption measurements for the isosbestic wavelengths are optimal to estimate the scattering and the ocular media contribution to a given reflectometry absorption measurement.

[0085] Referring to Figure 6: [0086] The operation 501 of performing reflectometry absorption measurement on the vessels of the patient's retina at a number of wavelengths in the visible spectral area, corresponding to isosbestic points where hemoglobin and oxyhemoglobin present similar absorption coefficients, and the operation 504 of performing reflectometry absorption measurement on the vessels of the patient's retina at a number (at least one) of wavelengths where a ratio between the absorption coefficients of the hemoglobin and oxyhemoglobin presents a local maximum are performed by a detector 601 ;

[0087] The operation 502 of estimating the absorption contribution by the other optical structures of the patient's eye at each wavelength as a difference between the reflectometry absorption measurement and a weighting parameter that amplifies the hemoglobin absorption coefficient is performed by an estimator 602;

[0088] The operation 503 of adjusting the weighting parameter so that points representing the differences are collinear when representing these points on a logarithmic scale of the wavelengths is performed by an adjuster 603; and

[0089] The operation 504 of deriving blood oxygenation values of the vessels of the patient's retina from at least of portion of the reflectometry absorption measurements and the absorption contributions is performed by a calculator 605.

Instrument for on line and real time blood oxygenation mapping

[0090] For obtaining the blood oxygenation mapping of the optic nerve by using the multispectral analysis of the model as proposed herein, a multispectral imagery technique that explores images of the optic nerve papilla for several wavelengths is used. [0091] In accordance with the results of the method for blood oxygenation evaluation as presented in the present description, the multispectral imagery technique characterizes every point of the image representing the optic nerve papilla by a number as low as four values of absorption evaluated for each of four specific wavelengths (for example three wavelengths for isosbestic points and one wavelengths for local maxima of differences between absorption coefficients for hemoglobin and oxyhemoglobin).

[0092] For example, to obtain an optic nerve papilla image with a reasonable resolution of 200 x 200 values for blood oxygenation computed from six specific wavelengths, 240 000 absorption values are collected and solved at every 500 ms. This task represents a volume of data that the current computers are capable to collect, to solve and to store on line every 500 ms.

[0093] To form the detector 601 , there are several types of multispectral imagery systems proposed in the literature that could be characterized in two main groups.

[0094] In the first group are included the hyperspectral image techniques. This technique collects image data for numerous narrow adjacent spectral bands. Therefore, a hyperspectral image contains spectral data to derive a continuous spectrum for each image pixel.

[0095] The hyperspectral devices use monochrome cameras synchronised with tunable filters, snapshot or Bragg-grating systems. The spectral information is obtained during the capture process [15] [16].

[0096] In a second group are included the multispectral imaging devices which allow to acquire images at several spectral bands. These devices are used principally to estimate typical spectral characteristics in an image.

[0097] There are a large number of papers that propose devices for multispectral imaging of the retina [8] [16] [20].

[0098] Several authors even suggest that the images obtained with these devices can be useful for assessing the blood oxygenation and to map various retinal structures for the blood oxygenation [15] [19] [20]. However, the methods for the blood oxygenation computation remain vague or necessitate a prior calibration.

[0099] A device 601 adapted to capture the optic nerve papilla images where each image pixel contains spectral data for, for example, six specific wavelengths of the visible spectral area, wherein the instrument conserves the basic concept of a retinal imaging setup represented by a coupling between a monochrome CCD device and a fundus camera.

[00100] The CCD camera (Figure 7) contains an active matrix of, for example, 1024 x 1024 pixels cooled for low noise detection. The multispectral imaging device uses a retinal lighting setup that emits six successive monochromatic lights on the same light beam.

[00101] There are several possibilities to design a light source that emits a temporal sequence of six or more monochromatic lights in a single light beam.

[00102] One possibility, easy to achieve and less expensive, is illustrated in Figure 7. Figure 7 shows a design for a multispectral fundus camera with a rotating filter wheel. [00103] The lighting setup shown in Figure 7 represents a coupling between a 1 -watt white light from a LED (Mr) collimated by lenses (SLc, L2 and LOB) and a high speed rotating filter wheel. There are six interferential filters with a 3 nm band width mounted on the rotating wheel. The adjacent filter transition is 30 ms. A period of 80 ms is allowed to take one monochromatic image if the exposure time for one monochromatic image is fixed at 50 ms. The six monochromatic images proposed to compute one image for blood oxygenation can be obtained in approximatively 500 ms. A triggering device (computer controller) coordinates the filter wheel position and the exposure time for the CCD camera.

[00104] Another interesting design for a light source which emits temporal sequences of six or more monochromatic lights in a single light beam is illustrated in Figure 8 and combine six or more lasers beams from Laser 1 - Laser 6 in one beam light by using beam splitters BSi - BS ₅ or optical fibers coupler. Figure 7 shows a typical design to combine six lasers beams in one beam light by using beam splitters. The advantage of a multi laser setup is that the wavelengths transition delay is reduced to zero. Another advantage to illuminate the retina by several wavelengths from a laser setup is that laser beams can be modulated to perform scanning laser imageries which significantly increases the signal noise ratio in the retinal image and therefore increase the accuracy for blood oxygenation estimation. This option is commercially available but is expensive.

[00105] Set-ups as illustrated in Figures 7 and 8 are well known to those of ordinary skill in the art and, for that reason, will not be further elaborated in the present disclosure.

[00106] As illustrated in Figure 1 1 , the device (identified as 1100 in Figure 11 ) for evaluating blood oxygenation in vessels of the patient's retina may comprise an input 1202, an output 1204, a processor 1206 and a memory 1208. [00107] The input 1020 may be configured to receive the reflectometry absorption measurements from the detector 601. The output 1204 may be configured to supply the blood oxygenation data or mapping determined from the reflectometry absorption measurements. The input 1202 and the output 1204 may be implemented in a common module, for example a serial input/output device.

[00108] The processor 1206 is operatively connected to the input 1202, to the output 1204, and to the memory 1208. The processor is realized as one or more processors for executing code instructions in support of the functions of the various modules of the method (operations 502, 504 and 505) and device (estimator 602, adjuster 603 and calculator 605) for evaluating blood oxygenation in vessels of the patient's retina.

[00109] The memory may comprise a non-transient memory for storing code instructions executable by the processor, specifically, a processor-readable memory comprising non-transitory instructions that, when executed, cause a processor (a) to control the detector 601 of Figures 7 and 8 and (b) to implement the modules (602, 603, 605) of the device of Figure 6 for evaluating blood oxygenation in vessels of the patient's retina and the operations (502, 503, 505) of the method for evaluating blood oxygenation in vessels of the patient's retina as described in Figure 5 of the present disclosure. The memory 1208 may also comprise a random access memory or buffer(s) to store intermediate processing data from the various functions performed by the processor.

[00110] Those of ordinary skill in the art will realize that the description of the method and device for evaluating blood oxygenation in vessels of the patient's retina are illustrative only and are not intended to be in any way limiting. Other embodiments will readily suggest themselves to such persons with ordinary skill in the art having the benefit of the present disclosure. Furthermore, the disclosed device and method may be customized to offer valuable solutions to existing needs and problems of acquiring blood oxygenation data.

[00111] In the interest of clarity, not all of the routine features of the implementations of the method and device for evaluating blood oxygenation in vessels of the patient's retina are shown and described. It will, of course, be appreciated that in the development of any such actual implementation of the device and method, numerous implementation-specific decisions may need to be made in order to achieve the developer's specific goals, such as compliance with application-, system-, network- and business-related constraints, and that these specific goals will vary from one implementation to another and from one developer to another. Moreover, it will be appreciated that a development effort might be complex and time-consuming, but would nevertheless be a routine undertaking of engineering for those of ordinary skill in the field of acquiring blood oxygenation data.

[00112] In accordance with the present disclosure, the modules, processing operations, and/or data structures described herein may be implemented using various types of operating systems, computing platforms, network devices, computer programs, and/or general purpose machines. In addition, those of ordinary skill in the art will recognize that devices of a less general purpose nature, such as hardwired devices, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), or the like, may also be used. Where a method comprising a series of operations is implemented by a processor, computer or a machine and those operations may be stored as a series of non-transitory code instructions readable by the processor, computer or machine, they may be stored on a tangible and/or non- transient medium.

[00113] Modules of the method and device for evaluating blood oxygenation in vessels of the patient's retina as described herein may comprise software, firmware, hardware, or any combination(s) of software, firmware, or hardware suitable for the purposes described herein.

[00114] In the method for evaluating blood oxygenation in vessels of the patient's retina as described herein, the various operations may be performed in various orders and some of the operations may be optional.

[00115] Although the present disclosure has been described hereinabove by way of non-restrictive, illustrative embodiments thereof, these embodiments may be modified at will within the scope of the appended claims without departing from the spirit and nature of the present disclosure.

References

[00116] The following references, at least some of which are mentioned in the disclosure, are incorporated herein by reference.

[1] Hayreh SS., 2001 , Risk factors in AION. Ophthalmology, Oct; 108(10): 1717-8.

[2] Harris A., et al., 2003, A review of methods for human retinal oximetry Opht.

Surg. Lasers Im. 34.

[3] Pittman RN and Duling BR 1975 A new method for the measurement of percent oxyhemoglobin J. Appl. Physiol. 38.

[5] Beach JM., et al., 1999, Oximetry of retinal vessel by dual-wavelength imaging: calibration and influence of pigmentation J. Appl. Physiol 86. Hardarson HS., et al., 2006, Automatic retinal oximetry, Invest. Ophthalmol, Vis. Sci. 47.

Delori FC, 1988, Noninvasive technique for oximetry of blood in retinal vessels, Applied Optics, 15;27(6): 1 1 13-25.

Yip W., et al., Reliability and Determinants of Retinal Vessel Oximetry Measurements in Healthy Eyes, IOVS, Vol. 55, No. 1 1 , 7105, November 2014.

Ramanujam N., Phelps J., Systems and methods for determining hemoglobin concentration utilizing diffuse reflectance at isosbestic wavelengths, wo 201 1068998 A2.

Phelps J., et al., Rapid radiometric determination of hemoglobin concentration using UV-VIS diffuse reflectance at isosbestic wavelengths, OPTICS EXPRESS Vol. 18, No. 18, August 2010.

Candon Lau, et al, 2009, Re-Evaluation of model-based light-scattering spectroscopy for tissue spectroscopy. Journal of biomedical Optics, 14(2).

Diaconu V., 2009, Multichannel spectroreflectometry: a noninvasive method for assessment of on-line hemoglobin derivatives Appl. Opt. 48.

Vucea V., et al., 201 1 , Blood Oxygenation Measurements by Multi-Channel Reflectometry on the Venous and Arterial Structures of the Retina. Applied Optics 201 1 vol 50 No 26. [14] Schweitzer D., et al., 1995, Calibration-free measurement of the oxygen saturation in retinal vessels of men. Proc. SPIE 2393.

[15] ZUZAK, K., et al., DIGITAL LIGHT PROCESSING HYPERSPECTRAL IMAGING APPARATUS, US 2013/0296708 A1 .

[16] Tkaczyk T., et al., IMAGE MAPPING SPECTROMETERS, US 8,654,328 B2.

[17] Francisco H., et al., Imaging systems and applications: Introduction to the feature, APPLIED OPTICS / Vol. 53, No. 13 / 1 May 2014.

[18] Martinez M., et al., Combining transverse field detectors and color filter arrays to improve multispectral imaging systems APPLIED OPTICS, Vol. 53, No. 13, May 2014.

[19] Khoobehi K., et al., METHOD FOR EVALUATING RELATIVE OXYGEN SATURATION IN BODY TISSUES, US 7,949,387 B2.

[20] Jasper S. Y., et al., PAPER BASED SPECTROPHOTOMETRIC DETECTION OF BLOOD HEMOGLOBIN CONCENTRATION, US

8,730,460 B2.

[21 ] Han Q. Le and Yang Wang, Semiconductor Laser Multi-Spectral Sensing and Imaging, Sensors, 10, 544-583, 2010.

[22] Diaconu, V; Van Loc Tran, Optic Nerve Capillaries Blood Oxygenation in Normal and Primary Open-Angle Glaucoma Subjects Invest. Ophthalmol. Vis. Sci. 2013; 54(15):4456.