Computer-implemented method, computer program, and surgical system for determining the blood volume flow through a portion of a blood vessel in a surgical area

Abstract

The present invention relates to a computer-implemented method (10) for determining the blood volume flow (I i , i = 1, 2, 3,...) through a part (90 BI ) of a blood vessel (88) in a surgical area (36) by means of a fluorophore. A plurality of images (801, 802, 803, 804,...) are provided, which are based on fluorescence in the form of light having a wavelength within the fluorescence spectrum of the fluorophore and which show the part (90 i ) of the blood vessel (88) at different recording time points (t1, t2, t3, t4,...). By processing at least one of the provided images (801, 802, 803, 804,...), the diameter (D) and length (L) of the part (90 i ) of the blood vessel (88) and the time interval for the fluorophore to propagate through the part (90 i ) of the blood vessel (88) are determined, the time interval describing the characteristic transit time (T) of the fluorophore in the part (90 i ) of the blood vessel (88), wherein a vascular model (I) for the part (90 i ) of the blood vessel (88) is described, the vascular model describing the part (90 i ) of the blood vessel (88) as a flow channel (94) having a length (L), having a wall (95) with a wall thickness (d) and having a free cross-section Q. A fluid flow model (II) for the vascular model (I) is assumed, the fluid flow model describing the local flow velocity (122) at different positions on the free cross-section Q of the flow channel (94) in the vascular model (I) and a fluorescence model (III) is assumed, the fluorescence model describing the spatial probability density of the intensity of the light emitted at different positions on the free cross-section Q of the flow channel (94) in the vascular model (I), the light being emitted by the fluid doped with the fluorophore flowing through the free cross-section Q of the flow channel (94) in the vascular model (I) when irradiated with fluorescence excitation light. The blood volume flow (I BI ) is determined as the fluid flow guided through the flow channel (94) in the vascular model (I), the fluid flow being calculated according to the length (L) and diameter (D) of the part (90 i ) of the blood vessel (88) and the characteristic transit time (T) of the fluorophore in the part (90 i ) of the blood vessel (88) by means of the fluid flow model (II) and the fluorescence model (III).

Description

Technical Field

[0001] The invention relates to a computer-implemented method for determining the blood volume flow through a portion of a blood vessel in a surgical area by means of a fluorophore. In this case, a plurality of images are provided, which are based on fluorescence in the form of light having a wavelength within the fluorescence spectrum of the fluorophore and which show the portion of the blood vessel at different recording time points. By processing the provided images, the diameter and length of the portion of the blood vessel and the time interval of propagation of the fluorophore through the blood vessel are determined, which time interval describes a characteristic transit time of the fluorophore in the blood vessel. The invention also relates to a computer program and a surgical system for determining the blood volume flow through a portion of a blood vessel in a surgical area by means of a fluorophore. Background Art

[0002] For example, the determination of blood volume flow is of interest in neurosurgery, since it can be used to check the success of bypass revascularization, aneurysm clipping or aneurysm treatment. The movement of plasma proteins contained in human blood can be made visible by doping the blood with fluorescent dyes, such as the dye indocyanine green (ICG), which bind to plasma proteins in the blood and can be excited to fluoresce by irradiation with light of a suitable wavelength. If such fluorescent dyes are introduced into the patient's blood flow, the blood volume flow in the patient's blood vessels can be inferred by evaluating the corresponding video images using a camera system designed to capture the fluorescence of the fluorescent dye.

[0003] A computer-implemented method of the type mentioned at the outset for determining the blood volume flow through a portion of a blood vessel in the surgical area is known from “Quantitative fluorescence angiography for neurosurgical interventions by Claudia Weichelt et al., BiomedTech 2013, Vol. 58, No. 3, pp. 269-279”. The determination of the blood volume flow in a blood vessel is described therein, into which the dye ICG is introduced and for which a video sequence is recorded with the aid of a camera. Here, the surgeon selects the start and end points of a portion of the blood vessel of interest in order to determine the blood volume flow. At these points, the intensity of the fluorescence is determined over time and smoothed, and then the time offset between the intensity curves and the length of the portion of the blood vessel are determined. The blood flow velocity is then determined from the time offset and the length, in order to calculate the blood volume flow from this with the aid of a cross section of the blood vessel.

[0004] In "Quantitative fluorescence angiography forneurosurgical interventions by Claudia Weichelt et al., Biomed Tech 2013, Vol. 58, No. 3, pp. 269-279", on page 274 in the section "Phantom measurements", it is mentioned that the values ​​of blood volume flow calculated in this publication differ from the values ​​measured during the experiment by a factor that depends on the diameter of the blood vessels. However, the rule used to calculate this factor is not stated.

[0005] In "Tsukiyama, A.; Murai, Y.; Matano, F.; Shirokane, K.; Morita, A.:Optical effects on the surrounding structure during quantitative analysisusing indocyanine green videoangiography: A phantom vessel study, J.Biophotonics, 2018, 11.-ISSN 1864-0648", it is described that the local intensity of the fluorescence detected from the observation window with the aid of a surgical microscope in a surgical field with blood vessels depends not only on the distance from the observation window to the vessel but also on the thickness of the vessel and the spatial environment of the vessel. It is pointed out there that this correlation must be taken into account when performing a quantitative analysis of the fluorescence in order to infer the blood flow in the vessel.

[0006] "Xu, J., Song, S., Li, Y., and Wang, R.: Complex-based OCTangiography algorithm recovers microvascular information superior toamplitude or phase-based algorithm in phase-stable systems [In phase-stable systems, the complex-based OCT angiography algorithm recovers microvascular information better than amplitude- or phase-based algorithms], Physics in medicine and biology, Vol. 63, December 19, 2017, 1 - ISSN 1361-6560" shows that the retina is examined by means of OCT in order to visualize blood vessels there in particular by evaluating the phase and amplitude of the OCT signal.

[0007] It is known from "Quantitative Blood Flow Assessment by MultiparameterAnalysis of Indocyanine Green Video Angiography [Quantitative blood flow assessment by multi-parameter analysis of indocyanine green video angiography] by Saito et al., World Neurosurgery, 2018, Vol. 116, pp. 187 - 193" the analysis of multiple measurable variables in video data in the case of adding ICG and irradiating the surgical area with fluorescence excitation light. Here, the temporal intensity distribution is examined at only one point of the blood vessel, and by comparison with experiments, the gradient is found to be the best indicator of blood volume flow. However, no specific calculation rules for determining the blood volume flow during surgery are described. Summary of the Invention

[0008] The object of the present invention is to accurately determine the blood volume flow through a part of a patient's blood vessel, especially during surgery.

[0009] This object is solved by the method for determining the blood volume flow through a part of a blood vessel as described in claim 1, the computer program as described in claim 14, and the device as described in claim 15. Advantageous embodiments and improvements of the present invention are described in the dependent claims.

[0010] Currently, the term blood volume flow represents the volume V of blood flowing through a part of a blood vessel with diameter D and length L per unit time t:

[0011]

[0012] wherein,

[0013] .

[0014] The computer-implemented method according to the invention for determining the blood volume flow through a part of a blood vessel in a surgical area by means of a fluorophore as set forth in claim 1 comprises the following method steps:

[0015] Providing a plurality of images which are based on fluorescence in the form of light having wavelengths within the fluorescence spectrum of the fluorophore and which show the part of the blood vessel at different successive recording time points. By processing the provided images, the diameter and length of the part of the blood vessel and the time interval of propagation of the fluorophore through the part of the blood vessel are determined, the time interval characterizing the transit time of the fluorophore in the part of the blood vessel. Adapting a blood vessel model to at least one of the provided images by means of image processing, the blood vessel model depicting the part of the blood vessel as a flow channel having a length, having a wall with a wall thickness and having a free cross-section Q. Providing a fluid flow model for the adapted blood vessel model , the fluid flow model depicting the local flow velocities at different positions on the free cross-section Q of the flow channel in the adapted blood vessel model. Furthermore, providing a fluorescence model , the fluorescence model depicting the spatial probability density of the intensity of the light emerging at different positions on the free cross-section Q of the flow channel in the adapted blood vessel model, the light being emitted by a fluid doped with the fluorophore and flowing through the free cross-section Q of the flow channel in the adapted blood vessel model when irradiated with fluorescence excitation light. Furthermore, determining the blood volume flow as the fluid flow guided through the flow channel in the adapted blood vessel model, the fluid flow being calculated by means of the provided fluid flow model and the provided fluorescence model in accordance with the length and diameter of the part of the blood vessel and the transit time of the fluorophore in the part of the blood vessel.

[0016] For the part of the blood vessel, processing the blood vessel model by processing at least one of the provided images , the blood vessel model depicting the part of the blood vessel as a flow channel having a length, having a wall with a wall thickness and having a free cross-section Q. Furthermore, processing the fluid flow model for the blood vessel model , the fluid flow model depicting the local flow velocities at different positions on the free cross-section Q of the flow channel in the blood vessel model . Furthermore, assuming a fluorescence model , the model depicting in the blood vessel model The spatial probability density of the intensity of the light emitted from different positions on the free cross-section Q of the flow channel, the light being emitted when the fluid flowing through the free cross-section Q of the flow channel in the blood vessel model doped with fluorophores is irradiated with fluorescence. Finally, the blood volume flow rate is determined as the fluid flow rate of the fluid flowing through the flow channel in the blood vessel model, and the fluid flow rate is calculated according to the length and diameter of this part of the blood vessel and the characteristic transit time of the fluorophore in this part of the blood vessel by means of a fluid flow model and a fluorescence model The fluid flowing through the flow channel in the blood vessel model, and the fluid flow rate is calculated according to the length and diameter of this part of the blood vessel and the characteristic transit time of the fluorophore in this part of the blood vessel by means of a fluid flow model and a fluorescence model The fluid flowing through the flow channel in the blood vessel model, and the fluid flow rate is calculated according to the length and diameter of this part of the blood vessel and the characteristic transit time of the fluorophore in this part of the blood vessel by means of a fluid flow model

[0017] The fluid flow model represents the mapping from the free cross-section Q, especially a partial region of the free cross-section, to the flow velocity, where the flow velocity can be specified one-dimensionally in the form of the magnitude of the flow velocity or in the form of an n-dimensional flow velocity vector having a magnitude and a direction:

[0018] .

[0019] The fluorescence model represents the mapping from the free cross-section Q, especially a partial region of the free cross-section, to real numbers:

[0020] .

[0021] Satisfying the model assumptions by means of the blood vessel model 、the fluid flow model and the fluorescence model makes it possible to determine the blood volume flow rate through this part of the blood vessel from the images provided during the operation by means of specific calculation rules. Thus, the method can be practically applied during the operation. Due to the different parameters of the model assumptions, the method for determining the blood volume flow rate can be flexibly adapted to different scenarios. Therefore, the method is also applicable to measuring the volume flow rate of a medium different from blood through vessels with different layer structures, where the medium has a characteristic fluid flow rate and the fluorescence has characteristic properties that can be represented in the model assumptions. By adapting the model assumptions to special circumstances during the operation, an improved accuracy of the blood volume flow rate determined by the method can also be achieved.

[0022] The method according to the present invention is based on the following assumptions: The considered part of the blood vessel is clear and fully visible in the provided images and is located in the focal plane of the image acquisition device. During the determination of the blood volume flow rate, the parameters of the image acquisition device and the blood vessel model 、the fluid flow model and the fluorescence model are not changed.

[0023] If the blood vessel model is a hollow cylinder having a determined length, a determined diameter, and a determined wall thickness for this part of the blood vessel, it is advantageous. This simplifies the blood vessel model , and thus also simplifies the calculation of blood volume flow, whereby calculation time can be saved.

[0024] Furthermore, if the fluid flow model describes laminar fluid flow through the flow channels of the blood vessel model , it is advantageous. This measure simplifies the fluid flow model , and thus also simplifies the calculation of blood volume flow, whereby calculation time can likewise be saved.

[0025] To determine the blood volume flow through this part of the blood vessel, furthermore, it is advantageous that the fluid flow model and the fluorescence model describe local segments of the blood vessel model such that the prototype of the fluid flow model and the prototype of the fluorescence model correspond to partial regions of the free cross-section Q of the blood vessel model . This in turn simplifies the method and makes the method applicable in practice. In particular, it is helpful that instead of defining the fluid flow model and the fluorescence model only on the diameter of the free cross-section Q on the free cross-section Q, the diameter is along a line that extends orthogonally to the centerline of the blood vessel and intersects the centerline. Each point then corresponds to the penetration depth of photons into the blood vessel.

[0026] The fluorescence model is preferably based on a computer simulation of irradiating the blood vessel model with fluorescence excitation light, where photons are assumed to be particles scattered at scattering centers. In this case, the scattering centers in the flow channels and the scattering centers in the wall of the blood vessel model have respectively characteristic scattering center distributions. In this case, the blood vessel model It is represented as a layer model with three layers: vessel wall-flow channel-vessel wall. In this layer model, the movement of a large number of photons is simulated by means of previously specified parameters of absorption, scattering and scattering anisotropy of photons in the medium within the layer and at the layer boundary, as described in the publication "L. Wang, S. Jacques, Monte Carlo Modeling of Light Transport inMulti-layered Tissues in Standard C [Monte Carlo Modeling of Light Transport in Multi-layered Tissues in Standard C], Computer Methods and Programs in Biomedicine, Vol. 47, No. 2, pp. 131-146, 1995", which is hereby fully cited and the disclosure of this publication is included in the description of the present invention. In order to determine the fluorescence model by means of computer simulation , in this case it is advantageous that the fluorescence model Prototype and vascular model The chord corresponds to the free cross section Q of the blood vessel model when irradiated with fluorescent excitation light. In this case, the chord of the cross section Q represents a line segment between any two points on the edge of the cross section Q, such that the line segment extends within the cross section Q. Fluorescence Model The penetration depth x is then mapped to the The ratio of the number of photons exported to the blood vessel model during the simulation The maximum penetration depth in corresponds to the value x. In this case, the fluorescence model can be understood as a probability density, where

[0027] .

[0028] An advantageous embodiment of the method provides that the fluid flow model Relative Fluid Flow Model Correspondingly, the relative fluid flow model is described in the vascular model The flow velocity at different positions on the free cross section Q of the flow channel relative to the reference flow velocity Relative flow distribution curve in the form of relative flow velocity. Relative fluid flow model The fluid flow model is as follows It turns out that:

[0029] .

[0030] Using the relative fluid flow model has the following advantages: the flow velocities at different positions relative to each other and relative to a reference flow velocity are known. Thus, the reference flow velocity can be inferred from the observation of the velocity at a certain position in the blood vessel . The reference flow velocity is selected by the user in this case. In this case, for example, the reference flow velocity can be selected as

[0031] - the flow velocity of the fluid flow model at a defined position in the blood vessel model , where for x

[0032] , , ,

[0033] - the maximum flow velocity of the fluid flow model in the blood vessel model , where

[0034] -

[0035] - the average flow velocity of the fluid flow model in the blood vessel model , where

[0036] .

[0037] By means of the fluid flow model is a relative fluid flow model which describes the local relative flow velocities at different positions on the free cross-section Q of the flow channel in the blood vessel model relative to the reference flow velocity . The blood volume flow I through a part of the blood vessel in the surgical area is determined with the aid of a fluorophore BI can be achieved with increased accuracy.

[0038] That is, this measure achieves: determination of the error in calculating the blood volume flow as mentioned in the above publication by Claudia Weichelt et al., which is determined from experiments and is in the form of a factor of the blood volume flow determined from the relative fluid flow model and the fluorescence model .

[0039] ​​In order to improve the accuracy of the method for determining blood volume flow, the present invention proposes to analytically determine the factors measured in the above-mentioned publication by Claudia Weichelt et al. In an advantageous refinement of the present invention, it is thus provided that the observed flow velocity is corrected by means of a correction factor :

[0040]

[0041] wherein the correction factor is determined according to the reference flow velocity of the relative fluid flow model .

[0042] In this case, the recognition on which the present invention is based is that the characteristic corresponding to the observed flow velocity of the fluorophore in a blood vessel section having a diameter D and a length L corresponding to the determined time interval

[0043]

[0044] depends particularly on the penetration depth of the fluorescence into the blood vessel through the time τ, since the proportion of the emitted fluorescence changes with the penetration depth of the fluorescence into the blood vessel. Therefore, the fluorescence model affects the observed flow velocity in this part of the blood vessel.

[0045] By combining the fluid flow model with the fluorescence model , the knowledge about the fluorescence emitted from different penetration depths can be included in the calculation of the flow velocity in the blood vessel model.

[0046] That is, the inventors have recognized that the intensity of the emitted light does not correspond to a uniform distribution over all penetration depths, but rather the proportion of the emitted light from certain penetration depths is particularly high and the proportion of the emitted light from other penetration depths is very low. This relationship is described by the fluorescence model .

[0047] Therefore, the recognition of the present invention is in particular that the correction factor can be determined from the fluid flow model and the fluorescence model , both of which are based on the blood vessel model .

[0048] For this purpose, the observed flow velocity observed in the blood vessel model is used as the expected value of the observed local flow velocity with respect to the fluid flow model according to the fluorescence model The spatial probability density of the intensity of the exiting light at different positions on the free cross-section Q of the flow channel 94 in the blood vessel model ( ) is calculated as follows:

[0049] .

[0050] If the fluid flow model corresponds to the selected reference flow velocity and the relative fluid flow model , then the following applies:

[0051]

[0052] And thus

[0053] .

[0054] Based on the flow velocity observed in the blood vessel model , the reference flow velocity

[0055]

[0056] can thus be inferred with the aid of a correction factor . This correction factor depends on the selected reference flow velocity in the fluid flow model .

[0057] Assume that the flow velocity actually observed in this part of the blood vessel corresponds to the flow velocity observed in the blood vessel model , i.e.,

[0058] ,

[0059] The reference flow velocity that can be expected in this part of the blood vessel can be determined based on the flow velocity observed in this part of the blood vessel. Therefore, when selecting a suitable reference flow velocity , the flow velocity observed in this part of the blood vessel is corrected with the aid of the correction factor :

[0060] .

[0061] For example, in order to obtain the value in the blood vessel model based on the flow velocity The average flow velocity in this part of the blood vessel at all positions of the free cross-section Q of the flow channel can be calculated as follows for the fluid flow model The average flow velocity in As the reference flow velocity :

[0062] .

[0063] According to the reference flow velocity , the associated relative fluid flow model Can be calculated as follows:

[0064] .

[0065] According to the relative fluid flow model And the fluorescence model , the correction factor can then be determined as follows :

[0066] .

[0067] Therefore, the correction factor Achieves: The flow velocity observed in this part of the blood vessel Is corrected as follows to the average flow velocity in the blood vessel Of the value expected in the fluid flow model :

[0068] .

[0069] Therefore, in this method, the fluid flow rate guided through the flow channel in the blood vessel model Is calculated by: From the relative fluid flow model with respect to the reference flow velocity And the fluorescence model Determine the correction factor And determine the fluorophore propagation velocity from the lengths L of the parts 901, 902, 903,... of the blood vessel 88 and from the characteristic transit times τ of the fluorophore in the parts 901, 902, 903,... of the blood vessel 88, and this fluorophore propagation velocity is corrected by means of this correction factor To a value corresponding to the reference flow velocity . In this case, the correction factor As the reciprocal of the expected value of the relative flow velocity in the relative fluid flow model According to the intensity of the emitted light at different positions on the free cross-section Q of the flow channel in the blood vessel model ​​The described spatial probability density is determined according to the following rules:

[0070] .

[0071] Preferably, the determination of the blood volume flow is carried out with the aid of a look-up table which contains pre-calculated correction factors for different fluid flow models and / or different fluorescence models . For example, the associated correction factor can be pre-calculated according to the diameter D of the hollow cylinder of the vascular model and stored in the table as a tuple [diameter, k]. This saves calculation time and simplifies the method.

[0072] In order to determine the length, average diameter, center line of this part of the blood vessel as parameters of the vascular model , in order to determine the parameters of the fluid flow model and / or the parameters of the fluorescence model , advantageously, the selected image is determined from the provided images with the aid of criteria regarding the image brightness of the individual image points of the image, i.e., the intensity of the image points, where this image brightness is a measure of the intensity of the emitted light detected by means of the image acquisition device.

[0073] This criterion corresponds to the state in which the blood vessels in the image are filled with the fluorescent agent to the maximum extent. This measure improves the accuracy of the determination of the blood volume flow.

[0074] In order to determine the length and / or the diameter of this part of the blood vessel, preferably the center line of this part of the blood vessel is determined in at least one of the provided images. The center line forms the central axis of this part of the blood vessel, such that the distance to the blood vessel wall is the same at any position of the center line. The center line can be determined from the provided images with the aid of image processing. This ensures the automatic determination of the center line, such that during the method for determining the blood volume flow, as little interaction or no interaction at all from the surgeon is required, so that the method can be applied in practice.

[0075] To this end, in the selected image, this part of the blood vessel is first determined by means of an image segmentation method. In this case, the segmentation represents an image that indicates, for each pixel, the class to which the pixel belongs. In particular, the segmentation is a binary image, where the value 1 means that the pixel belongs to this part of the blood vessel, and the value 0 means that the pixel does not belong to this part of the blood vessel. To ensure short computation times, in particular, an adaptive thresholding method is suitable as the segmentation method, such as that described in the publication "Nobuyuki Otsu, A threshold selection method from gray-level histogram, IEEE Trans. Sys. Man. Cyber. 1979, Vol. 9, No. 1, pp. 62 - 66", hereby incorporating the full text of this publication by reference and including the disclosure of this publication in the description of the present invention.

[0076] Other segmentation methods known to those skilled in the art from this literature, in particular methods for segmenting blood vessels in medical images, can also be used instead of these methods. The surgeon can adapt the segmentation and / or select this part of the blood vessel in which the blood volume flow should be measured. Using the image segmentation method has the following advantages: the method can operate as automatically as possible without any effort on the part of the surgeon and is therefore suitable for use during surgery in practice.

[0077] To determine the centerline from the segmentation of this part of the blood vessel, preferably in a first step, the pixels on the central axis of this part of the blood vessel are determined by processing the provided image, in particular the segmentation. To this end, morphological operations, such as the so-called erosion or the so-called Voronoi diagram or other algorithms, can be applied to the segmented image, for example, the skeletonization algorithm described in the article "FixedTopology Skeletons, P. Golland, W. Grimson, International Conference on Computer Vision and Pattern Recognition (CVPR), 2000", hereby incorporating the full text of this article by reference and including the disclosure of this article in the description of the present invention.

[0078] Then, broken lines are determined from the respective pixels on the central axis by connecting adjacent pixels. To improve the accuracy of this method, the connection structure of these pixels is adapted based on the pixel neighborhood of the pixels along the central axis to minimize the length of the broken line. Thereby, the discretization error is reduced. In particular, in this case, for each pixel along the center line, the pixel neighborhood around this pixel is considered, such as an 8-pixel neighborhood. The L-shaped connection structure between three consecutive pixels is replaced by a diagonal connection structure in this case in order to obtain a center line whose length corresponds as much as possible to the length of this part of the blood vessel. By fitting a continuous function, in particular a Bezier spline, to the broken line with minimized length, the discretization error can be further reduced and thus the accuracy of this method can be improved.

[0079] Preferably, it is determined by the offset of the temporal development of the image brightness at at least two different segments of this part of the blood vessel through the images provided by the processing over time, where continuous functions are respectively fitted to the temporal development of the image brightness at different segments of this part of the blood vessel. The transit time describes the characteristic time interval for the fluorophore to propagate in this part of the blood vessel. To determine the transit time, a starting point and an end point are determined on the center line of this part of the blood vessel, between which the blood volume flow should be determined.

[0080] In this case, the starting point is within the range between 5% and 15% of the total length of this part of the blood vessel, preferably at 10% of the total length of this part of the blood vessel, and the end point is within the range between 80% and 95% of the total length of this part of the blood vessel, preferably at 90% of the total length of this part of the blood vessel. Thereby, inaccuracies in determining the center line (such as through erosion of segmentation) are avoided, which particularly occur at the start and end of a part of the blood vessel, and thus the accuracy of this method is improved. In this case, the starting point and the end point can be automatically determined with the help of the center line and the described range, or they can be specified by the surgeon in the selected image. For the starting point and the end point, segments around the respective points are determined, such as a 5×5 pixel neighborhood with the starting point or the end point as the center point. To determine the intensity distribution, i.e., the distribution of the image brightness at the respective points, the intensity in the segment is calculated by averaging the intensities of all pixels in the segment in at least a plurality of the provided images. Averaging the intensity of the entire segment improves the accuracy of this method. The intensity values of the starting point and the end point are plotted at different recording time points of the image. A continuous function is fitted to these measured values in order to obtain the continuous temporal development of the image brightness at the respective points. In particular, a gamma function is used for this purpose. The temporal offset of the image brightness curves at the starting point and the end point of this part of the blood vessel represents the transit time.

[0081] The diameter of the flow channel cannot be directly determined from the segmentation of this part of the blood vessel, because the segmentation does not distinguish between the flow channel and the blood vessel wall. Assuming that the cross-section of the blood vessel model is circular, the total diameter of the cross-section of the blood vessel model can be determined with the help of the segmentation. For this purpose, starting from the points on the center line, circles around the corresponding points are determined respectively, and the radius of the circle is continuously increased until the circle edge coincides with the edge of the segmentation. The average value of the diameters of these circles corresponds to the diameter of the circular cross-section of the blood vessel model . To also determine the diameter of the flow channel, it is necessary to distinguish between the flow channel and the blood vessel wall in the segmented part of the provided image. In order to be able to make this distinction with as high an accuracy as possible, it is advantageous if the width and wall thickness of the flow channel of the blood vessel model are determined with the help of a criterion regarding the intensity distribution orthogonal to the center line of this part of the blood vessel in one or more of the provided images. In particular, in this case it is advantageous if the criterion for the curve of the intensity distribution takes into account the curvature of the curve of the intensity distribution orthogonal to the center line of this part of the blood vessel. That is, the recognition of the present invention is that the points on the boundary between the flow channel and the blood vessel wall respectively correspond to the minimum values of the curvature of the intensity distribution that is orthogonal to the center line of this part of the blood vessel. With the help of this criterion, the segmentation of the flow channel can then be determined from the segmentation of this part of the blood vessel. Then, the diameter of the flow channel can be determined as described above for the total diameter by increasing the diameter of the circle on the center line of this part of the blood vessel until the circle edge coincides with the flow channel edge.

[0082] Then, according to the transit time τ, the length L and diameter D of the flow channel of this part of the blood vessel, and the correction factor , the blood volume flow rate can be calculated as

[0083] .

[0084] Finally, it is also advantageous that, for the calculated blood volume flow rate in this part of the blood vessel, a confidence interval is determined with the help of error simulation based on the diameter and / or length and / or transit time and / or correction factor and / or blood vessel model and / or fluid flow model and / or fluorescence model and / or the shape of the center line of this part of the blood vessel. Thus, the accuracy of the blood volume flow rate determined by this method is quantified for the surgeon. This measure also helps the method to be usable in practice.

[0085] An advantageous refinement of the method provides that, in order to determine the blood volume flow through the blood vessels in the surgical area with the aid of fluorophores, the blood vessels are divided into a plurality of parts, and the blood volume flow in these parts is determined with the aid of the method described above for determining the blood volume flow in a part of a blood vessel (I Bi ). In this case, the blood volume flow is determined under the following conditions: at the branch of the blood vessel, the sum of the blood volume flow (I Bi ) towards the branch corresponds to the sum of the blood volume flow (I Bi ) from the branch. The assumption on which the method is based is that the volume of the blood remains constant over the distribution of the blood vessels. To ensure this, the determination of the blood volume flow in the individual parts of the blood vessels can be formulated as an optimization problem over all parts, where the volume conservation at the branches is included as a secondary condition. This has the advantage that the blood volume flow can be determined with higher accuracy by simultaneously determining it in a plurality of mutually dependent parts of the blood vessels.

[0086] The invention also extends to a computer program having program code for performing the above method steps when the computer program is loaded into and / or executed in a computer unit.

[0087] Furthermore, the invention extends to a surgical system for determining the blood volume flow (I Bi ) through a part of a blood vessel in a surgical area with the aid of fluorophores. The surgical system comprises: an illumination device for providing fluorescence excitation light for the surgical area; an image acquisition device for providing a plurality of images which are based on light having a wavelength within the fluorescence spectrum of the fluorophore and which show the part of the blood vessel at different recording time points; and a computer unit having a computer program which has program code for performing the above method steps for determining the blood volume flow. BRIEF DESCRIPTION OF THE DRAWINGS

[0088] The advantageous embodiments of the invention, which are schematically depicted in the drawings, are described below.

[0089] In the drawings:

[0090] Figure 1 A surgical microscope showing a system for determining the blood volume flow through a part of a blood vessel in a surgical area;

[0091] Figure 2 Showing a plurality of images of a surgical area with blood vessels;

[0092] Figure 3 Showing a blood vessel model;

[0093] Figure 4A Showing the segmentation of a part of a blood vessel;

[0094] Figure 4B Show a horizontal cross-section of a blood vessel model and parameters of the blood vessel model;

[0095] Figures 5A to Fig. 5F Show the determination of the centerline of a part of a blood vessel;

[0096] Fig. 6A Show a segment of the polyline (Polygonzug) shown in Figure 5E;

[0097] Figure 6B Show Fig. 6A The post-processing of the segment of the polyline shown in;

[0098] Figure 7A shows a blood vessel;

[0099] Figure 7B shows a part of a blood vessel;

[0100] Figure 7C shows the intensity distribution orthogonal to the centerline of a part of a blood vessel;

[0101] Fig.7D Show the flow channel edge points of a part of a blood vessel;

[0102] Figure 8 Show a fluid flow model with an absolute flow distribution curve;

[0103] Fig.9A Show a relative fluid flow model;

[0104] Fig. 9B Show a relative fluid flow model with a relative flow distribution curve in the form of a parabola for a blood vessel with a diameter of 3 mm;

[0105] Fig.10 Show the simulation of the propagation of photons in a blood vessel model for determining a fluorescence model;

[0106] Fig.11 Show the fluorescence model determined by simulation;

[0107] Fig.12 Show the flowchart of the simulation algorithm for the propagation of photons in a blood vessel model;

[0108] Figure 13A shows a horizontal cross-section of a part of a blood vessel, which part contains a first segment of a starting point and a second segment containing an end point;

[0109] Fig. 13B Show the vertical cross-section of the part of the blood vessel in Figure 13A along the penetration depth x and the associated flow distribution curve;

[0110] Figure 13C shows a relative fluid flow model;

[0111] Figure 13D shows a fluorescence model;

[0112] Figure 14A shows a horizontal cross-section of a part of a blood vessel that includes a first segment at the starting point and a second segment at the ending point;

[0113] Fig. 14B Shows the time evolution of the intensity in the first segment and the second segment of a part of a blood vessel for determining the transit time;

[0114] Fig.15 Shows an embodiment of a method step for determining the blood volume flow in a segment of a blood vessel in the surgical area; and

[0115] Fig.16 Shows an embodiment of a method step for determining the blood volume flow in a segment of a blood vessel in the surgical area. DETAILED DESCRIPTION

[0116] In Figure 1 The surgical microscope 12 shown in includes a system 14 for determining the blood volume flow I through a single part of the blood vessel 88 in the surgical area 36 Bi and is designed for neurosurgery. The surgical microscope 12 has a microscope main objective lens 20. The microscope main objective lens 20 is housed in the microscope body 22. The microscope body 22 includes an adjustable magnification system 24. The left and right observation beam paths 26, 28 pass through the microscope main objective lens 20. The binocular tube 30 is connected to the microscope body 22. In the left and right observation beam paths 26, 28, the binocular tube 30 includes an eyepiece lens 32 and a tube lens 34. Through the binocular tube 30, currently, the observer can stereoscopically observe the surgical area 36 at the patient's brain 37 using the left and right observer eyes 38, 40.

[0117] In the system 14 for determining the blood volume flow I Bi there is an illumination device 42. The illumination device 42 provides illumination light 46 for the surgical area 36 using the illumination beam path 44. The illumination device 42 has a xenon light source 48. The illumination device 42 includes other optical elements in the form of a lens 50, an optical waveguide 52, and an illumination objective lens 54. The light of the xenon light source 48 is coupled into the optical waveguide 52 through a lens system having the lens 50. The illumination light 46 reaches the surgical area 36 from the optical waveguide 52 through the illumination objective lens 54.

[0118] To adjust the spectral composition of the illumination light 46, the illumination device 42 includes a switchable filter assembly. The filter assembly includes an illumination filter 56. According to the arrow 59, the illumination filter 56 can be moved into the illumination beam path 44 and removed from the illumination beam path 44.

[0119] The illumination filter 56 is a band-pass filter. This band-pass filter is transmissive to light within the spectral range between 780 nm and 810 nm from the xenon light source 48. Light within the spectral ranges below 810 nm and above 780 nm is filtered out or strongly suppressed by the illumination filter 56.

[0120] In the microscope body 22, the observation filter 60 for the left observation beam path 26 and the observation filter 62 for the right observation beam path 28 are located on the side of the magnification system 24 facing away from the microscope main objective 20. According to the double arrows, the observation filters 60, 62 can be moved into or out of the observation beam paths 26, 28. On the one hand, the illumination filter 56 and on the other hand the observation filters 60, 62 have mutually matching filter characteristics. For fluorescence observation of the surgical area 36, the illumination filter 56 is switched into the illumination beam path 44, and the observation filters 60, 62 are arranged in the observation beam paths 26, 28.

[0121] System 14 for determining the blood volume flow I in the operating microscope 12 Bi has an image acquisition device 64 which is used to acquire images 801, 802, 803, 804,... of the surgical area 36. Observation light from the surgical area 36 can be conveyed from the right observation beam path 28 through the observation filter 62 via an output coupling beam splitter 66 with an optical axis 68 to the image acquisition device 64. An image sensor 70 is present in the image acquisition device 64. The image sensor 70 is sensitive to the emission wavelength of the fluorophore ICG, which is within the spectral range from 810 nm to 830 nm, and the fluorophore is introduced into the patient's bloodstream to determine the blood volume flow I in the blood vessels Bi .

[0122] The image sensor 70 of the image acquisition device 64 is connected to a computer unit 72. The computer unit 72 has an input unit 74 and includes a program memory 76. The computer unit 72 is connected to a screen 78. Images 801, 802, 803, 804,... of the surgical area 36 acquired at different recording time points t1, t2, t3, t4,... are displayed on the screen. The computer unit 72 controls a display 82. The display of the display 82 is superimposed via a lens 84 through a beam splitter 86 on the observation light in the right observation beam path 28. For the observer, the display of the display 82 is thus simultaneously visible with the surgical area 36 in the right eyepiece view (Okulareinblick) of the binocular tube 30.

[0123] Figure 2A plurality of images 801, 802, 803, …… showing the surgical area 36 are acquired by means of an image acquisition device 64 at different recording time points t1, t2, t3, ……. The images 801, 802, 803, …… each contain a blood vessel 88 which has three spatially separated portions 90, i = 1, 2, 3. In the images 801, 802, 803, ……, the blood volume flow I Bi is indicated by means of an arrow 92. Here, the direction of the arrow 92 corresponds to the direction of the blood volume flow I Bi , and the length of the arrow 92 corresponds to the amount of the blood volume flow I i in the corresponding portion 90 Bi of the blood vessel 88.

[0124] The blood volume flow I B in the entire blood vessel 88 is composed of the blood volume flow I i in the respective portions 90 BI (i = 1, 2, 3) of the blood vessel 88. At the branch 89 of the blood vessel 88 visible in Figure 2 , the blood volume flow I B1 of the blood flowing into the branch 89 is I B1 = 5 ml / s. The blood volume flows I B2 and I B3 in the portions 902, 903 after the branch 89 are I B2 = 2 ml / s and I B3 = 3 ml / s. Thus, I B1 = I B2 + I B3 applies, that is, the sum of the blood volume flows I B2 , I B3 after the branch 89 corresponds to the blood volume flow I B1 before the branch 89. This relationship means that the blood volume remains unchanged at the branch 89 in the vascular system.

[0125] The condition that the blood volume remains unchanged at the branch 89 in the vascular system can be used as a linear constraint for an optimization problem that simultaneously determines the blood volume flow I i on all portions 90 i (i = 1, 2, 3, ……) of the blood vessel 88 with the corresponding data given for the portions 90 BI .

[0126] A computer program is loaded into the program memory 76 of the computer unit 72, and this computer program is used to determine the blood volume flow I i through a portion 90Bi 。

[0127] The computer program contains a vascular model that describes the geometry of a part 90 of a blood vessel 88 i 。 。

[0128] Figure 3 Displays the vascular model which models a part 90 of the blood vessel 88 shown in Figure 2 as a hollow cylinder having a length L and a cylinder axis 91 through which the patient's blood flows, the hollow cylinder having a wall 95 with a wall thickness of d and forming a flow channel 94 delimited by the wall 95 of the hollow cylinder, wherein the flow channel 94 has a circular cross-section Q, an inner diameter D and an outer diameter G and can be flowed through by fluid in the direction of arrow 93 i (i = 1, 2, 3)(through which the patient's blood flows) as a hollow cylinder having a length L and a cylinder axis 91, the hollow cylinder having a wall 95 with a wall thickness of d and forming a flow channel 94 delimited by the wall 95 of the hollow cylinder, wherein the flow channel 94 has a circular cross-section Q, an inner diameter D and an outer diameter G and can be flowed through by fluid in the direction of arrow 93

[0129] In the computer program loaded into the program memory 76 of the computer unit 72, the vascular model is determined by processing at least one of the provided images 801, 802, 803, 804, …… of the part 90 of the blood vessel 88 i 。 To this end, the selected image is determined from a plurality of images in the images 801, 802, 803, 804, …… by means of a criterion regarding the image brightness of the image points of the image, i.e., the intensity of the image points. Since fluorescence results in a particularly high intensity of the image points in the image, the state in which the blood vessel in the image is maximally filled with the fluorescent agent is determined by the following criterion wherein,

[0130]

[0131] wherein, is the set of image points x in the image and is the image brightness of the image at that image point, which is currently referred to as the intensity of the image point x

[0132] The image with a high maximum intensity and at the same time having a large number of pixels presenting this maximum intensity value is determined by maximizing the value A

[0133] For the part 90 of the blood vessel 88 i as shown in Figure 4A the segmentation 96 of the part 90 of the blood vessel 88 and the starting point P1 and the ending point P2 are then determined. By means of the publication " i 。 A threshold selection method from gray-level histograms [A threshold selection method based on grayscale histogram], the Otsu thresholding described in "Nobuyuki Otsu, IEEE Trans. Sys. Man. Cyber., Vol. 9, No. 1, pp. 882 - 886, 1979" is used to determine the segmentation 96, and the entire publication is hereby incorporated by reference in its entirety and its disclosure is included in the specification of the present invention. The initial segmentation 96 is improved by means of a gradient-based segmentation method explained with reference to FIGS. 7A to 7F.

[0134] For a portion 90 Figure 4A of the blood vessel 88 shown in i the segmentation, Figure 4B shows a horizontal cross-section of the blood vessel model and its parameters, which are determined from the segmentation 96 in a computer program for determining the blood volume flow I i through the portion 90 of the blood vessel 88 in the surgical area 36 by means of a fluorophore. Here, the centerline 98 is determined by means of erosion of the segmentation 96 of the portion 90 Bi of the blood vessel 88. In this case, the starting point P1 and the ending point P2 of the portion 90 i of the blood vessel 88 are defined, and the blood volume flow I i is determined between these two points. In this case, the starting point P1 and the ending point P2 are located on the centerline 98. In addition, the starting point P1 is located within the range between 5% and 15%, preferably 10%, of the total extension along the length of the portion 90 Bi of the blood vessel 88, and the ending point P2 is located within the range between 80% and 95%, preferably 90%, of the total extension along the length of the portion 90 i of the blood vessel. This avoids inaccuracies in determining the centerline 98, which occur especially at the beginning and end of the portion 90 i of the blood vessel 88. i

[0135] The starting point P1 and the ending point P2 can be automatically determined by means of image processing with the aid of the centerline 98 and the described ranges, or they can be specified by the surgeon in the selected image. The length L of the portion 90 i of the blood vessel 88 considered for blood volume flow determination is determined by determining the length of the centerline portion 99 between the starting point P1 and the ending point P2 of the centerline 98. To determine the total diameter G i of the portion 90 L of the blood vessel 88, the local total diameter G is determined at any point between the starting point P1 and the ending point P2 along the centerline 98 in the following way: A circle is defined around each point and the radius of the circle is continuously increased until the edge of the circle touches the edge of the blood vessel 88. In this case, as in Figure 4AIn [the figure], the edge of the blood vessel 88 can be determined by determining the edge of the segmentation 96 of the blood vessel 88. The total diameter G of the blood vessel 88 then corresponds to the average value of all local total diameters G L corresponds to.

[0136] To determine the diameter D of the flow channel 94 of the portion 90 of the blood vessel 88, the local diameter D is determined at any point along the center line 98 between the starting point P1 and the ending point P2 by the following method L : Define a circle around each point and continuously increase the radius of the circle until the edge of the circle touches the inner side of the wall 95 of the blood vessel 88. The inner side of the wall 95 of the blood vessel 88 can be determined by determining the edge of the segmentation 96 of the flow channel as in [Figures 7A to Fig.7D . The diameter D of the flow channel 94 then corresponds to the average value of all local diameters D L corresponds to.

[0137] Figures 5A to Fig. 5FDetermination of the center line 98 of a portion 90 of a blood vessel 88 and determination of the length L of the portion 90 of the blood vessel 88 between the starting point P1 and the end point P2 considered for the blood volume flow determination in a computer program is explained. FIG5A shows a portion 90 of a blood vessel 88 with a center line 98. In the image acquisition device 64, the portion 90 of the blood vessel 88 is imaged onto the region 71 of the image sensor 70 of the image acquisition device 64 shown in FIG5B, which region has a resolution that can be seen in FIG5C. FIG5D shows pixels belonging to the discretized center line 102 on the image sensor 70, and FIG5E shows a polyline 104 of the discretized center line 102. The computer program contains a post-processing routine that post-processes this polyline 104, as can be seen in FIG6, in order to avoid discretization errors as much as possible when determining the length of the center line 98 between the starting point P1 and the end point P2. These discretization errors are described as experimental results in the article "A. Naber, D. Berwanger, W. Nahm, In Silico Modelling of Blood Vessel Segmentations for Estimation of Discretization Error in Spatial Measurement and its Impact on Quantitative Fluorescence Angiography, 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 2019" and are an average of 6.3%. Fig. 5F As shown in , a continuous function 106 in spline form is fitted to the post-processed polyline 104 and the length L of the centerline portion 99 of the portion 90 of the blood vessel 88 between the start point P1 and the end point P2 is calculated by means of arc integration in order to further reduce the discretization error.

[0138] As in Fig. 6A and Figure 6B As shown in , the post-processing routine of the computer program corrects the center line 98 of the polyline 104 shown in FIG. 5E by taking into account the pixel neighborhood to reduce the discretization error. To do this, for each pixel along the center line 98, a pixel neighborhood 108, here an 8-pixel neighborhood 108, surrounding each pixel is considered. Thus, by Figure 6B The diagonal connection structure 112 in the embodiment replaces Fig. 6AThe L-shaped connection structure 110 in it. In this way, the length of the center line 98 is minimized, and the center line 100 to be post-processed is generated. As an alternative to the 8-neighborhood, other pixel neighborhoods can also be considered.

[0139] Figures 7A to Fig.7D explains determining the diameter D of the flow channel 94 and the wall thickness d of the blood vessel model in a computer program based on one of the images 801, 802, 803, 804,... acquired and thus provided by the image acquisition device 64. Since the wall 95 of the blood vessel scatters the fluorescence signal, the boundary between the flow channel 94 and the wall 95 of the blood vessel cannot be clearly identified in the images acquired by the image acquisition device 64. Figure 7A shows the blood vessel 88 of one of the images 801, 802, 803, 804,... and the selected part 97 therein. i In Figure 7B, the selected part 91 of the blood vessel 88 can be seen. i The local intensity distribution in the selected image is shown in Figure 7C as a curve 114 along a line segment x extending orthogonally to the center line 98 of the part 90 of the blood vessel 88. The blood vessel model i describes the part 90 of the blood vessel 88 as a flow channel 94 having a circular cross-section, which has a diameter D and is bounded by a wall 95 having a wall thickness d. The diameter D of the flow channel 94 and the wall thickness d of the blood vessel model are determined by means of a criterion regarding the intensity distribution i that is orthogonal to the center line 98 of the part 90 of the blood vessel 88 in one or more of the provided images 801, 802, 803, 804,.... The criterion regarding the intensity distribution for determining the diameter D and the wall thickness d of the blood vessel model i can be, for example, the curvature of the intensity distribution

[0140] orthogonal to the center line 98 of the flow channel 94, where the boundary between the flow channel 94 and the wall 95 is defined at so-called flow channel edge points 116, at which the curvature in the form of the second derivative of the intensity distribution reaches a minimum value looking outwards from the center line 98, and the intensity I(x) presents an intensity value 115. The criterion regarding the intensity distribution for determining the diameter D and the wall thickness d of the blood vessel model can be, for example, the curvature of the intensity distribution

[0141] The motivation for this standard is that the inventors have recorded images of a silicone tube with a known wall thickness, filled here with a blood-like medium and an ICG dye, using the surgical microscope 12, and have examined the intensity distribution orthogonal to the center line 98 of the silicone tube in the acquired images 801, 802, 803, 804, … This was repeated for different diameters of the silicone tube and different arrangements of the silicone tube under the surgical microscope 12. In this case, the inventors have found that the curvature of the intensity distribution, in particular orthogonal to the center line of the silicone tube in the acquired images 801, 802, 803, 804, … is suitable as a criterion for determining the diameter D of the flow channel 94. The curvature of the intensity distribution I(x) is determined in this case in the form of the second derivative of the intensity distribution I(x). In this method (Vorgehen), the boundary between the flow channel 94 and the wall 95 corresponds to the points where the curvature of the intensity distribution reaches a minimum when looking outwards from the center line 98.

[0142] To determine the boundary between the flow channel 94 and the wall 95, the computer program thus determines the flow channel edge points 116 for the points on the center line 98 of the part 90 i of the blood vessel 88 as the first two points 115 where the curvature of the intensity distribution I(x) in the form of the second derivative orthogonal to the center line 98 has a minimum. The local diameter D L at the corresponding points on the center line 98 is determined by the distance between these two flow channel edge points 116. By averaging the local diameters D L for all points on the center line 98, the diameter D of the flow channel 94 is determined. The distance between the flow channel edge points 116 and the edge of the segmented part 90 of the blood vessel 88 then corresponds to the local wall thickness d L . This local wall thickness is also averaged for all points along the center line 98 and thus the wall thickness d of the part 90 of the blood vessel 88 is determined. The segmentation 96 of the flow channel 94 is determined by connecting the flow channel edge points 116.

[0143] A computer program for determining the blood volume flow I Bi through a part i of the blood vessel 88 in the surgical area 36 with the aid of a fluorophore contains in addition to the blood vessel model a fluid flow model that describes the local flow velocity 122 at different positions on the free cross-section Q of the flow channel 94 in the blood vessel model

[0144] Figure 8 Shows along Figure 3 the blood vessel model in The diameter D of the cross-section Q of the flow channel 94 has a flow distribution curve 120 with different local flow velocities 122. Here, the fluid flow model of the computer program describes the laminar fluid flow through the blood vessel model of the flow distribution curve 120 of the flow channel 94. The fluid flow model is defined on the chord of the cross-section Q of the flow channel 94, i.e., along the diameter D, and has the following mapping rules:

[0145] .

[0146] Therefore, the fluid flow model assigns to each position along the diameter D of the cross-section Q of the flow channel 94 the local flow velocity 122 at that position.

[0147] Fig.9A And Fig. 9B respectively show the relative fluid flow model , which describes the relative flow distribution curve 124 of the laminar fluid flow through the flow channel 94 of the blood vessel model .

[0148] The relative flow distribution curve describes the local relative flow velocity 126 at different positions on the diameter D of the cross-section Q of the flow channel 94 in the Figure 3 blood vessel model relative to the reference flow velocity . The reference flow velocity is determined in particular by selecting the determined flow velocity in the flow distribution curve 124 as shown in Fig.9A or by averaging all the local flow velocities 122 in the flow distribution curve 120 as can be seen in Fig. 9B and is determined by the local flow velocity 122 as described in Figure 8 . To illustrate the relative fluid flow model with the relative flow distribution curve 124 , the values of the flow distribution curve 120 in Figure 8 are divided by the value of the reference flow velocity . In Fig. 9B shown, the relative fluid flow model in the form of the fluid flow model describes the laminar flow with the relative flow distribution curve 124 in the form of the reference flow velocity relative to the average value of all local flow velocities 122 for a diameter D of 3 mm. The relative flow distribution curve 124 has a parabolic shape here

[0149]

[0150] Among them,

[0151] .

[0152] A computer program for determining the blood volume flow I of a part i of a blood vessel 88 in a surgical area 36 by means of a fluorophore Bi further includes a fluorescence model , which describes the spatial probability density of the intensity of the emitted light at different positions on the free cross-section Q of the flow channel 94 in the blood vessel model . The emitted light is emitted when the fluid flowing through the free cross-section Q of the flow channel 94 in the blood vessel model and doped with a fluorophore is irradiated with fluorescence excitation light. As shown in Fig.10 and described in the above-mentioned publication by L. Wang and S. Jacques, the fluorescence model is determined by Monte Carlo simulation of the propagation of photons 127 in the blood vessel model . In this case, the blood vessel model is assumed to be a three-layer model 128: blood vessel wall - flow channel - blood vessel wall. The layer model 128 is irradiated with light, and the path of the photons 127 in the layer model 128 is tracked. In this case, the photons 127 are assumed to be particles scattered at the scattering centers, where the scattering centers in the flow channel 94 and the wall 95 of the blood vessel model have characteristic scattering center distributions respectively. When hitting the scattering center, the photons 127 are scattered with a certain probability and absorbed with another probability. To determine the fluorescence model , for each emitted photon 127 that leaves the layer model 128 through the same layer as it enters the layer model 128 again, it is determined which photon has reached the maximum penetration depth of that photon 127 in the layer model 128. In this case, the different penetration depths of the photons 127 into the layer model 128 correspond to the diameter D of the cross-section Q of the flow channel 94. In this embodiment, it is advantageous to divide the diameter D into n equally large sub-segments, for example for n = 100.

[0153] Fig.10 The graph 131 in shows the proportion of photons 127 emitted from the layer model 128 for different penetration depths x along the diameter D, and the maximum penetration depth of these photons corresponds to the value x. Since the fluorescence model represents a probability density, the following applies:

[0154] .

[0155] This probability density is also a measure of the intensity of the emitted light at different positions on the free cross-section Q of the flow channel 94 in the blood vessel model because the probability that a photon 127 reaches a determined penetration depth is also a measure of the intensity of the light emitted from that depth.

[0156] Fig.11 shows the fluorescence model determined by means of the Monte Carlo simulation described above . The penetration depth x along the diameter D of the flow channel is plotted on the vertical axis. The proportion A of the photons 127 shown in Fig.10 is plotted on the horizontal axis, these photons having reached the penetration depth x maximally during the simulation and leaving again through the same wall 95 through which they entered the model via the blood vessel model .

[0157] Fig.12A flowchart showing the movement of photons in the layer model 128 as described in the above-mentioned publication by Wang and Jacques. In the initialization step 132, the program is initialized and parameters are loaded. Thereafter, in the photon introduction step 134, the simulation of photons 127 with a prescribed initial weight is started. In the query 136, it is checked whether the photon 127 has entered the layer model 128 or has been reflected by specular reflection before entering the layer model 128. If the photon 127 has entered the layer model 128 and is located in the layer 129, then in the step length calculation step 138, the step length of the photon is determined according to the characteristics of the medium in the layer 129 by means of the probability distribution on the free path length of the photon 127. In the boundary distance determination step 140, the distance to the layer boundary 130 of the next layer 129 in the movement direction of the photon 127 is determined. In the subsequent query 142, it is checked whether the photon 127 will reach or exceed the layer boundary 130 of the layer 129 in its movement direction with the determined step length in the next step. If this is not the case, then in the photon adaptation step 144, the position of the photon 127 is adapted by means of its step length. However, if the photon 127 reaches or exceeds the layer boundary 130, then in the query 146, it is checked whether the photon 127 is reflected at the layer boundary 130 or is transmitted to the next layer 129. If the photon 127 is transmitted to the next layer 129, then the parameters of the photon 127, such as the step length, are adapted to the next layer 129, and the above steps are repeated starting from the query 136 as to whether the photon 127 is located in the medium. If the photon 127 is reflected at the layer boundary 130 of the layer 129, then in the photon adaptation step 144, the position and direction of the photon are adapted by means of the parameters of the photon 127. In the weight absorption step 148, the weight of the photon 127 is reduced due to absorption at the interaction position. In this case, a part of the current weight of the photon 127 is deposited at a local position of the layer model 128, and the weight of the photon 127 is adapted. After the movement and weight reduction of the photon 127, the photon 127 is scattered in the scattering calculation step 150 by means of the characteristics of the medium and various statistically determined angles, and its parameters are adapted. In order to terminate photons 127 with a very low weight (the further movement of which has only a very small effect on the model), in the query 152, it is checked by means of a random value whether the photon 127 still survives in the simulation or whether the photon should be terminated in the photon end step 154. If the photon is the last photon 156 in the simulation, then in the final program end step 158, the program is ended. Otherwise, in the photon introduction step 134, the next photon 127 is introduced into the layer model 128.

[0158] In order to determine the fluorescence model by means of this simulation , the inventor has determined the movement of 1,000,000 photons 127 and in this case the following parameters are used for the layer model 128 with three layers 129:

[0159] Blood vessel wall Flow channel <![CDATA[Absorption coefficient µ a > <![CDATA[2.25 cm -1 > <![CDATA[7.38 cm -1 > <![CDATA[Scattering coefficient µ s > <![CDATA[200 cm -1 > <![CDATA[713 cm -1 > Refractive index n 1.44 1.38 Anisotropy g 0.99 0.99

[0160] Figure 13A shows a portion 90 of blood vessel 88 i of a horizontal cross-section having a starting point P1 and an ending point P2 on the centerline 98 of this portion 90 of the blood vessel 88 i and shows a first segment A1 containing the starting point P1 and a second segment A2 containing the ending point P2.

[0161] Fig. 13B For the first segment A1 in Figure 13A, the portion 90 of the blood vessel 88 is shown i along a vertical cross-section of the penetration depth . For each penetration depth x, the associated local relative flow velocity 126 is illustrated by the relative fluid flow model in Figure 13C and the proportion of photons 127 exiting from the corresponding penetration depth x is illustrated by the fluorescence model in Figure 13D . In this case, the relative fluid flow model is illustrated with respect to the average flow velocity of the fluid flow model . The relative fluid flow model and the fluorescence model can be used to determine the expected average flow velocity in this portion of the blood vessel in the following manner according to the flow velocity observed in this portion of the blood vessel in the form of a corrected flow velocity , as follows:

[0162]

[0163] where the correction factor

[0164] .

[0165] For the relative fluid flow model Fig. 9B shown with respect to the reference flow velocity in and the fluorescence model Fig.10 shown in the graph 131 of , thus for a diameter D of 3 mm, for example, the correction factor = 0.68 is obtained.

[0166] To save computational time, correction factors for different model parameters, e.g., for different diameters D of the flow channel 94, are pre-calculated and stored in a look-up table (LUT).

[0167] Figure 14A shows a horizontal cross-section of a portion 90 of the blood vessel 88 i thereof.

[0168] Fig. 14B showing the portion 90 of the blood vessel 88 i and the velocity observed in that portion of the blood vessel is calculated. For this purpose, the time evolution 160 of the intensity I in a first segment A1 and the time evolution 162 of the intensity in a second segment A2 are considered on a plurality of the provided images 801, 802, 803, 804,.... For this purpose, the intensity values I(t) are averaged over the pixels in the respective segments A1, A2. A continuous function 106 in the form of a gamma function is fitted to the discrete intensity values I(t) in order to obtain intermediate values and to be able to determine as accurately as possible the velocity observed in that portion of the blood vessel even at low image rates .

[0169] The current time shift determined by cross-correlation of the two curves then corresponds to the transit time τ. The blood flow direction can also be derived from this shift. It should be noted that the time shift of the two curves can in principle be determined by averaging the time shifts of different features of the two curves instead of by cross-correlation.

[0170] Based on the transit time τ, the length L and diameter D of the portion 90 of the blood vessel 88 i and the correction factor , the blood volume flow I Bi can then be calculated as

[0171] .

[0172] For the blood volume flow I Bi a confidence interval is specified as a function of the length L, diameter D, correction factor k, transit time τ and the shape of the centerline 98, where for this purpose the angle of a sub-part of the centerline with respect to the grid lines of the grid of the image sensor of the image acquisition device is also taken into account.

[0173] In this case the uncertainty is calculated according to DIN 1319 as the propagation of uncorrelated input uncertainties without assuming a normal distribution.

[0174] Fig.15 showing a method for determining a portion 90 of a blood vessel 88 in a surgical area 36 of a patient with the aid of a plurality of provided images 801, 802, 803, 804,.... iBlood volume flow I in Bi Flowchart of an embodiment of method 10. In image selection step 166, an image is selected from the provided images 801, 802, 803, 804,.... The selected image is segmented in image segmentation step 168. Diameter D is determined from the segmentation 96 in diameter calculation step 170. In addition, part 90 of blood vessel 88 is determined in centerline calculation step 172 i Centerline 98 of and the length L of the centerline.

[0175] In time evolution determination step 176, the time evolutions 160, 162 of the intensities in the first part and the second part are determined. In fitting step 178, a continuous function 106 is adapted to the time evolutions 160, 162. In transit time determination step 180, the transit time τ is calculated based on the offset between the continuous functions 106 adapted to the time evolutions 160, 162. Finally, in blood volume flow determination step 182, the blood volume flow I is determined from the determined data Bi .

[0176] Fig.16 Shows a method for determining a part 90 of blood vessel 88 in the surgical area 36 i Blood volume flow I in Bi Flowchart of another embodiment of method 10'. In this case, a video of the surgical area 36 under illumination light 46 is recorded using an image acquisition device 64 such that the fluorescent dye is visible. The video consists of a plurality of images 801, 802, 803, 804,.... These images are based on fluorescence in the form of illumination light 46 having a wavelength within the fluorescence spectrum of the fluorophore, and these images show part 90 of blood vessel 88 at different recording time points i . In image selection step 166, the selected image 164 is determined from among the plurality of images 801, 802, 803, 804,.... The selected image has the maximum number of color-saturated pixels of all the provided images 801, 802, 803, 804,.... Using an image segmentation method, in image segmentation step 168, part 90 of blood vessel 88 in the selected image 164 is determined i . In centerline calculation 172, part 90 of blood vessel 88 is calculated from the segmentation 96 i Centerline 98 of. Using the segmentation 96 and centerline 98 of part 90 of blood vessel 88 i , diameter D is calculated in diameter calculation 170. In start and end point calculation 184, start point P1 and end point P2 are determined on the centerline. In interpolation step 186, for example, the centerline 98 is interpolated using a Bessel spline segment by segment, and in length calculation 188, part 90 of blood vessel 88 is determined using the interpolated centerline 98i The length L. By means of the center line 98, the intensity distribution orthogonal to the center line 98 of the curve 114 and the portion 90 of the blood vessel 88 i of the segmentation 96, in the diameter calculation 170, the portion 90 of the blood vessel 88 is determined i of the diameter D. Then, for this diameter D, in the correction factor determination 192, a correction factor is determined by means of a look-up table . In addition, the characteristic passage time τ of the fluorophore is determined as the time interval during which the fluorophore passes through the portion 90 of the blood vessel 88 i propagates. For this purpose, in the time evolution determination 176, at two different segments A1, A2, where A1 contains the starting point P1 and A2 contains the end point P2, the time evolution 160, 162 of the image brightness is determined. For this purpose, the average value of the intensity over time in the corresponding segments A1, A2 is considered. In the fitting step 178, continuous functions 106, such as gamma functions, are respectively fitted to the measured values obtained from the time evolution 160, 162. In the passage time determination 180, the passage time τ is calculated from the time offset of the obtained curves. Finally, in the blood volume flow determination 182, from the portion 90 of the blood vessel 88 i of the length L and the diameter D and the passage time τ and the correction factor the blood volume flow I is obtained Bi . Additionally, in the confidence interval determination 190, a confidence interval is determined by means of the calculated parameters.

[0177] In summary, the following should be noted in particular: The present invention relates to a computer-implemented method 10 for determining the blood volume flow I through a portion 90 of a blood vessel 88 in a surgical area 36 by means of a fluorophore i (i = 1, 2, 3,...), wherein a plurality of images 801, 802, 803, 804,... are provided, which are based on fluorescence in the form of light having a wavelength within the fluorescence spectrum of the fluorophore, and which images show the portion 90 of the blood vessel 88 at different recording time points t1, t2, t3, t4,... BI , wherein, by processing the provided images 801, 802, 803, 804,..., the diameter D and the length L of the portion 90 of the blood vessel 88 and the time interval during which the fluorophore passes through the portion 90 of the blood vessel 88 are determined i , which time interval describes the characteristic passage time τ of the fluorophore in the portion 90 of the blood vessel 88 i , wherein, by processing the portion 90 of the blood vessel 88 i propagates, which time interval describes the characteristic passage time τ of the fluorophore in the portion 90 of the blood vessel 88 i in, wherein the portion 90 of the blood vessel 88 is processed i of the blood vessel model , which blood vessel model represents the portion 90 of the blood vessel 88 iA flow channel 94 is described as having a length L, a wall 95 with a wall thickness of d, and a free cross-section Q, wherein at least one of the provided images 801, 802, 803, 804,... is processed, and the method processes a fluid flow model for a blood vessel model , and the fluid flow model describes the local flow velocities 122 at different positions on the free cross-section Q of the flow channel 94 in the blood vessel model . The method processes a fluorescence model , and the fluorescence model describes the spatial probability density of the intensity of the emitted light at different positions on the free cross-section Q of the flow channel 94 in the blood vessel model . The light is emitted when the fluid flowing through the free cross-section Q of the flow channel 94 in the blood vessel model , which is doped with a fluorophore, is irradiated with fluorescence excitation light. And wherein, the blood volume flow rate I BI is determined as the fluid flow rate of the fluid guided through the flow channel 94 in the blood vessel model , and the fluid flow rate is calculated according to the length L and diameter D of the part 90 of the blood vessel 88 and the characteristics of the fluorophore in the part 90 of the blood vessel 88 through the characteristic time τ i using the fluid flow model i and the fluorescence model.

[0178] List of Reference Numerals

[0179] 10, 10’ Method

[0180] 12 Surgical microscope

[0181] 14 System for determining blood volume flow rate

[0182] 20 Microscope main objective lens

[0183] 22 Microscope body

[0184] 24 Magnification system

[0185] 26 Left observation light beam path

[0186] 28 Right observation light beam path

[0187] 30 Binocular tube

[0188] 32 Eyepiece lens

[0189] 34 Tube lens

[0190] 36 Surgical area

[0191] 37 Brain

[0192] 38 Left observer's eye

[0193] 40 Right observer's eye

[0194] 42 Lighting device

[0195] 44 Lighting beam path

[0196] 46 Lighting light

[0197] 48 Xenon light source

[0198] 50 Lens

[0199] 52 Light guide

[0200] 54 Lighting objective lens

[0201] 56 Lighting filter

[0202] 59 Arrow

[0203] 60 Observation filter for the left observation beam path

[0204] 62 Observation filter for the right observation beam path

[0205] 64 Image acquisition device

[0206] 66 Output coupling beam splitter

[0207] 68 Optical axis

[0208] 70 Image sensor

[0209] 71 Region

[0210] 72 Computer unit

[0211] 74 Input unit

[0212] 76 Program memory

[0213] 78 Screen

[0214] 801 Image 1

[0215] 802 Image 2

[0216] 803 Image 3

[0217] 804 Image 4

[0218] 82 Monitor

[0219] 84 Lens

[0220] 86 Beam splitter

[0221] 88 Blood vessel

[0222] 89 branches

[0223] 90, 90 i , for i = 1, 2, 3, …… part

[0224] 91 Cylinder axis

[0225] 92 Arrow

[0226] 93 Arrow

[0227] 94 Flow channel

[0228] 95 Wall

[0229] 96 Division

[0230] 97 i Image part

[0231] 98 Center line

[0232] 99 Center line part

[0233] 100 Post - processed center line

[0234] 102 Discretized center line

[0235] 104 Polyline

[0236] 106 Continuous function

[0237] 108 Pixel neighborhood

[0238] 110 L - shaped connection structure

[0239] 112 Diagonal connection structure

[0240] 114 Curve

[0241] 115 Intensity value

[0242] 116 Flow channel edge point

[0243] 120 Flow distribution curve

[0244] 122 Local flow velocity

[0245] 124 Relative flow distribution curve

[0246] 126 Local relative flow velocity

[0247] 127 Photon

[0248] 128 Layer model

[0249] 129 Layer

[0250] 130-layer boundary

[0251] 131 curve

[0252] 132 initialization

[0253] 134 photon introduction

[0254] 136 inquiry

[0255] 138 step size calculation

[0256] 140 boundary distance determination

[0257] 142 inquiry

[0258] 144 photon adaptation

[0259] 146 inquiry

[0260] 148 weight absorption

[0261] 150 scattering calculation

[0262] 152 inquiry

[0263] 154 photon end

[0264] 158 program end

[0265] 160 Time development of the intensity I(t) in the first segment

[0266] 162 Time development of the intensity I(t) in the second segment

[0267] 164 Selected image

[0268] 166 Image selection

[0269] 168 Image segmentation

[0270] 170 Diameter calculation

[0271] 172 Centerline calculation

[0272] 176 Time development determination

[0273] 178 Fitting

[0274] 180 Determination by time

[0275] 182 Blood volume flow determination

[0276] 184 Start and end point calculation

[0277] 186 Interpolation

[0278] 188 Length calculation

[0279] 190 Confidence interval determination

[0280] 192 Correction factor determination

[0281] I B Blood volume flow through blood vessel

[0282] I Bi Blood volume flow through the i-th part of blood vessel

[0283] ICG Indocyanine green

[0284] Blood vessel model

[0285] Fluid flow model

[0286] Relative fluid flow model

[0287] Fluorescence model

[0288] t1, t2, t3, t4 Recording time

[0289] L Length of part of blood vessel

[0290] Q Cross-section of flow channel

[0291] a Cylinder axis

[0292] D Diameter

[0293] D L Local diameter

[0294] G Total diameter

[0295] G L Local total diameter

[0296] d Wall thickness

[0297] d L Local wall thickness

[0298] P1 Starting point

[0299] P2 End point

[0300] Reference flow velocity

[0301] A1 First segment

[0302] A2 Second segment

[0303] τ Transit time

[0304] for reference flow velocity correction factor

[0305] flow velocity observed in the vascular model

[0306] flow velocity observed in a portion of the blood vessel.

Claims

1. A computer-implemented method (10) for determining the blood volume flow (I BI ) through a portion (901, 902, 903,...) of a blood vessel (88) in a surgical area (36) by means of a fluorophore Among them, Multiple images (801, 802, 803, 804, ……) are provided, which are based on fluorescence in the form of light having wavelengths within the fluorescence spectrum of the fluorophore, and which images show the part (901, 902, 903, ……) of the blood vessel (88) at different recording times (t1, t2, t3, t4, ……). Wherein, by processing the provided images (801, 802, 803, 804, ……), the diameter (D) and length (L) of the part (90) of the blood vessel (88) and the time interval for the fluorophore to propagate through the part (901, 902, 903, ……) of the blood vessel (88) are determined, and this time interval describes the characteristic passage time (τ) of the fluorophore in the part (901, 902, 903, ……) of the blood vessel (88). wherein, by means of image processing, a vascular model is adapted to at least one of these provided images (801, 802, 803, 804,...), the vascular model describing the portion (901, 902, 903,...) of the blood vessel (88) as a flow channel (94) having a length (L), having a wall (95) with a wall thickness (d) and having a free cross-section (Q); wherein, a fluid flow model for the adapted blood vessel model is provided ; The fluid flow model describes local flow velocities (122) at different positions on a free cross-section (Q) of a flow channel (94) in the adapted blood vessel model ; Characterized in that Provide a fluorescence model The fluorescence model describes the spatial probability density of the intensity of the light exiting at different positions on the free cross-section (Q) of the flow channel (94) in the adapted blood vessel model The light is emitted when the fluid flowing through the free cross-section (Q) of the flow channel (94) in the adapted blood vessel model, which is doped with a fluorophore, is irradiated with fluorescence excitation light; and ​ Wherein, the blood volume flow rate (I BI ) is determined as the fluid flow rate guided through the flow channel (94) in the adapted blood vessel model , and the fluid flow rate is calculated according to the length (L) and diameter (D) of the portion (901, 902, 903,...) of the blood vessel (88) and the characteristic passage time (τ) of the fluorophore in the portion (901, 902, 903,...) of the blood vessel (88) by means of the provided fluid flow model and the provided fluorescence model .

2. The computer-implemented method according to claim 1, wherein Collect a plurality of images (801, 802, 803, 804...) using an image acquisition device (64), wherein during determination of the blood volume flow (I BI ), the parameters of the image acquisition device (64), the blood vessel model the fluid flow model and the fluorescence model are not changed.

3. The computer-implemented method according to claim 1 or 2, characterized in that, This blood vessel model is a hollow cylinder with a length (L), a diameter (D), and a wall thickness (d). And / or This fluid flow model describes laminar fluid flow through the flow channel (94) of this blood vessel model .

4. The computer-implemented method according to claim 1 or 2, characterized in that, This fluorescence model is based on the simulation of irradiating the blood vessel model with fluorescence excitation light wherein photons (127) are assumed to be particles scattered at scattering centers, and the scattering centers in the flow channel (94) and the scattering centers in the wall (95) of the blood vessel model have characteristic scattering center distributions respectively.

5. The computer-implemented method according to claim 4, wherein, This fluorescence model has a prototype corresponding to the chord of the free cross-section (Q) of this blood vessel model . This chord represents the penetration depth of photons (127) into this blood vessel model when irradiated with fluorescence excitation light, and this fluorescence model maps the penetration depth x to the proportion of photons (127) exiting from this blood vessel model . The maximum penetration depth of these photons in this blood vessel model corresponds to the value x during this simulation.

6. The computer-implemented method according to claim 1 or 2, characterized in that, The fluid flow model is a relative fluid flow model The relative fluid flow model describes the local relative flow velocity (126) at different positions on the free cross-section (Q) of the flow channel (94) in the blood vessel model relative to the reference flow velocity (v R ).

7. The computer-implemented method according to claim 6, wherein is guided through the blood vessel model The fluid flow rate of the fluid flowing through the flow channel (94) in is calculated as follows: based on the relative fluid flow model with respect to the reference flow velocity (v R ) and the fluorescence model to determine a correction factor (k_v R ) and to determine the fluorophore propagation velocity based on the length (L) of the portion (901, 902, 903,...) of the blood vessel (88) and based on the characteristic transit time (τ) of the fluorophore in the portion (901, 902, 903,...) of the blood vessel (88), the fluorophore propagation velocity being corrected by means of the correction factor (k_v R ) to a value corresponding to the reference flow velocity (v R ).

8. The computer-implemented method according to claim 7, wherein, The correction factor (k_v R ) is determined according to the following rule as the reciprocal of the expected value of the relative flow velocity in the relative fluid flow model based on the spatial probability density described by the intensity of the emitted light at different positions on the free cross-section (Q) of the flow channel in the blood vessel model for the blood vessel model through the fluorescence model:

9. The computer-implemented method according to claim 1 or 2, characterized in that, The length (L) of the part (901, 902, 903, ……) of the blood vessel (88) and / or the diameter (D) of the part (901, 902, 903, ……) of the blood vessel (88) are determined by means of the center line (98) of the part (901, 902, 903, ……) of the blood vessel (88) in at least one of the provided images (801, 802, 803, 804, ……).

10. The computer-implemented method according to claim 9, wherein The center line (98) of the part (901, 902, 903, ……) of the blood vessel (88) is determined by the following method: By processing the provided images (801, 802, 803, 804, ……), the pixels on the center line (98) of the part (901, 902, 903, ……) of the blood vessel (88) are determined. A broken line (104) is determined from the pixels of the center line (98). By adapting the connection structure (110, 112) of these pixels based on the pixel neighborhood (108) of the pixels along the center line (98), the length of the broken line (104) is minimized. And a continuous function (106) is fitted to the minimized broken line (104).

11. The computer-implemented method according to claim 1 or 2, characterized in that The fluid flow model and the fluorescence model describe local segments (A1, A2) of the blood vessel model such that the prototype of the fluid flow model and the prototype of the fluorescence model represent partial regions of the free cross-section (Q) of the blood vessel model ; And / or The length (L), diameter (D), centerline (98) of the portion (90) of the blood vessel (88), the blood vessel model The fluid flow model and / or the fluorescence model is determined by a criterion regarding the intensity (I(x)) which is a measure of the image brightness of the image points of the selected image (164). And / or The passage time (τ) is determined by processing the provided images (801, 802, 803, 804, ……) according to the offset of the temporal development (160, 162) of the image brightness at at least two different segments (A1, A2) of the part (901, 902, 903, ……) of the blood vessel (88), and continuous functions (106) are respectively fitted to the temporal development (160, 162) of the image brightness at these different segments (A1, A2) of the part (901, 902, 903, ……) of the blood vessel (88). And / or For the blood volume flow rate (I BI ) in the calculated portion (901, 902, 903, ……) of the blood vessel (88), a confidence interval is determined by error simulation based on the diameter (D) and / or length (L) and / or the transit time (τ) and / or correction factor (k_v R ) and / or the blood vessel model and / or the fluid flow model and / or the fluorescence model and / or the shape of the centerline (98) of the portion (901, 902, 903, ……) of the blood vessel (88); And / or The part (90) of the blood vessel (88) is determined by processing the provided images (801, 802, 803, 804, ……) using an image segmentation method.

12. The computer-implemented method according to claim 1 or 2, characterized in that, The blood vessel model The width of the flow channel (94) and the wall thickness (d) of the wall (95) are determined by a criterion of a curve (114) of the intensity distribution orthogonal to the center line (98) of the portion (90) of the blood vessel (88) in one or more of the provided images (801, 802, 803, 804,...).

13. The computer-implemented method according to claim 12, wherein Standard consideration of the curve (114) of the intensity distribution concerns the minimum of the curvature of the curve (114) of the intensity distribution that is orthogonal to the center line (98) of the section (901, 902, 903, ……) of the blood vessel (88).

14. A computer-implemented method (10) for determining the blood volume flow (I B ) through a blood vessel (88) in a surgical area (36) by means of a fluorophore, wherein The blood vessel (88) is divided into a plurality of portions (901, 902, 903, ……) and the blood volume flow rates (I Bi ) of these portions (901, 902, 903, ……) are determined respectively according to one of claims 1 to 13 under the following conditions: at a branch (89) of the blood vessel (88), the sum of the blood volume flow rates (I Bi ) towards the branch (89) corresponds to the sum of the blood volume flow rates (I Bi ) from the branch (89).

15. A computer program product comprising a computer program having program code for performing all the method steps as set forth in any one of claims 1 to 14 when the computer program is loaded into and / or executed in a computer unit (72).

16. A surgical system for determining the blood volume flow (I BI ) through a portion (901, 902, 903, ……) of a blood vessel (88) in a surgical area (36) by means of a fluorophore, the surgical system An illumination device (42) for providing fluorescence excitation light for the surgical area (36); An image acquisition device (64) for providing a plurality of images (801, 802, 803, 804, ……), the images being based on light having a wavelength within the fluorescence spectrum of the fluorophore and the images showing the section (90) of the blood vessel (88) at different time points (t1, t2, t3, t4, ……); And a computer unit (72) comprising the computer program product as claimed in claim 15.

Images