Sensor catheter and signal processing for measuring flow velocity in a blood vessel

Abstract

1. An apparatus for measuring a flow velocity within a blood vessel (100), comprising: a catheter (104) configured to be inserted into the blood vessel (100); a plurality of flow velocity sensors (106) coupled to the catheter (104); a sensor network (108) coupled to the plurality of flow velocity sensors (106); and a processor (110) coupled to the sensor network (108), wherein each of the plurality of flow velocity sensors (106) is configured to sense a velocity of blood flow, and an output of the sensor network (108) is configured to be input to a mathematical model (152) stored in the processor (110), and the mathematical model (152) is configured to calculate the flow velocity within the blood vessel (100) in which the catheter (104) is positioned.

Description

[Technical field]

[0001] The present invention relates to an apparatus and associated method for measuring flow velocity in a blood vessel. The apparatus includes a catheter, a plurality of flow velocity sensors coupled to the catheter, a sensor network coupled to the plurality of flow velocity sensors, and a processor coupled to a system network. [Background technology]

[0002] There is a class of surgical procedures called interventional or minimally invasive procedures that involves the introduction of catheters into the human vascular system to measure quantities such as pressure and blood flow velocity. In a typical procedure, a catheter is introduced through an opening in a blood vessel, such as an opening in the groin or brachial artery. Through this opening, the catheter is then advanced through the blood vessel to a region of interest, often into a coronary artery branch. The interventional procedure is then performed in the region of interest.

[0003] Two parameters commonly measured by catheters are pressure and blood flow velocity, which are then processed to calculate two indices called fractional flow reserve (FFR) and coronary flow velocity reserve (CFVR).

[0004] Both pressure-derived myocardial fractional flow reserve (FFR) and coronary flow velocity reserve (CFVR), measured by noninvasive stress testing, have been evaluated as predictors of inducible ischemia and indicative of adverse events after stent implantation. Combining pressure and flow velocity as an index of hyperemic stenosis resistance significantly improves the diagnostic accuracy assessed by noninvasive ischemia testing, especially when results between conventional parameters are inconsistent.

[0005] The relationship between distal coronary velocity and transstenotic pressure gradient is determined almost entirely by the coronary stenosis and is therefore, by definition, well suited to assess its hemodynamic severity.

[0006] In all vascular procedures, for example coronary or peripheral procedures, such as stenting of stenoses or balloon dilatation, flow velocity and pressure measurements can always be used as diagnostic tools or to monitor the success of the procedure.

[0007] A commonly used technique for measuring blood flow velocity and indices derived therefrom is ultrasound. A catheter is equipped with piezoelectric crystals that, when excited, can emit ultrasonic waves. The ultrasound waves are then reflected by the natural scattering of blood, and measurements of the Doppler frequency shift or time of flight are used to derive the velocity of the fluid.

[0008] For example, to describe fluid flow within a blood vessel, both pressure and flow velocity may be required, and accurate knowledge of both pressure and flow velocity leads to a complete characterization of fluid flow and the definition of important diagnostic quantities such as peripheral vascular impedance. For example, a combined measurement of blood flow velocity and pressure may be used as part of a guidewire to assess the level of stenosis.

[0009] Measurements made with an ultrasound / Doppler sensor may be affected by the angle of incidence of the ultrasound as well as the direction of blood flow velocity according to the following equation: JPEG2024525076000002.jpg9150, where f d = Frequency shift f s = Frequency of the sound source v - velocity of the fluid c-speed of sound cosθ - the angle between the direction of velocity and the direction of emitted sound It is.

[0010] Therefore, the measurements are highly dependent on the angle θ, which cannot be controlled in an intravascular procedure; i.e., the angle θ depending on the position of the catheter can range from 0° to 90° giving completely different measurements.

[0011] A common problem is that the instantaneous blood flow velocity signal cannot be measured accurately enough to be dependent on the average blood flow velocity, so intracoronary adenosine is repeatedly injected.

[0012] WO 2019 / 149954 describes the placement of sensors to provide information on the alignment of a catheter (within a blood vessel). The contents of WO 2019 / 149954 are incorporated herein by reference. This application describes a method for aligning a sensor within a blood vessel in the yaw direction (see the z-axis in FIG. 1).

[0013] In this application, it was assumed that the pitch direction could be neglected due to the radial symmetry of the vessel. Identifying only the yaw direction may not be sufficient to characterize the flow in space, since the information provided is only about the plane identified by the sensor and information about the third dimension is missing. Information about roll therefore represents an important degree of freedom that needs to be considered.

[0014] For example, the catheter can be in an optimal position where the face of the sensor is aligned with the flow (Figure 2, roll = 0°), or in an angled position (Figure 3, roll = 90°) where the sensor is placed in a region of minimum velocity (see velocity distribution around the cylinder in Figure 4). These different positions can result in significantly different flow velocity readings.

[0015] A possible solution to this problem could be manual correction by the operator, who orients the catheter with the flow by manually rotating the catheter to obtain measurements that allow the best estimation of the flow.

[0016] This is not always possible as the stiffness of the catheter does not allow 1:1 control of the distal tip when it is manipulated proximally, which can result in stick-slip type rotation and suboptimal catheter placement. Furthermore, it is preferable to minimize the number of manipulations within the vessel to reduce the risk of loosening plaque within the vessel.

[0017] Furthermore, in Figure 5, the flow velocity distribution around the cylindrical shape is observed, - In the boundary layer at the surface facing the flow, the velocity is very low because in this region the flow is stopped by the cylinder. - In the boundary layers above and below the surfaces, the flow is accelerated by the cylinder, so the velocity is high. - The flow is very low in the boundary layer at the rear surface of the cylinder because the flow is in the shadow of the cylinder itself. It becomes clear that...

[0018] This means that it is difficult for a single sensor placed on the surface of a catheter that is not perfectly aligned with the flow to measure the actual vascular flow velocity due to the fact that there are non-negligible geometries in the flow that may accelerate or decelerate the flow velocity in either case, altering the measurement.

[0019] The device may have to be optimized with respect to the positioning of the sensor on the catheter, but in any case it would be desirable to create a device that allows for better measurement accuracy. There is therefore a need to provide a concept for equipping catheters with a sensor device that allows improved measurement of blood flow velocity, thereby giving accurate information regarding the correct positioning of the catheter. Such needs may be met by the subject matter of the claims. Summary of the Invention

[0020] The invention is set out in the independent claims. Preferred embodiments of the invention are outlined in the dependent claims. According to a first aspect, an apparatus for measuring flow velocity in a blood vessel is described. The apparatus comprises a catheter configured to be inserted into the blood vessel, a plurality of flow velocity sensors coupled to the catheter, a sensor network coupled to the plurality of flow velocity sensors, and a processor coupled to the sensor network. Each of the plurality of flow velocity sensors is configured to sense a velocity of blood flow. An output of the sensor network is configured to be input to a mathematical model stored in the processor. The mathematical model is configured to calculate the flow velocity in the blood vessel in which the catheter is located.

[0021] The catheter may be any commercially available catheter configured to be inserted into a blood vessel. The catheter may comprise any suitable material, such as, for example, PVC and / or rubber and / or silicone. The catheter is preferably shaped to allow multiple flow rate sensors to be coupled or placed on the catheter. The catheter is preferably cylindrical in cross section, but may alternatively be cubic, triangular, or custom shaped. In some instances, the catheter is instead an elongated body or element that may not be suitable for use as a catheter. Here, in this description, blood vessels are described. However, the catheter / elongated body or element may be inserted into any channel through which a fluid, such as, for example, air and / or water and / or oil, flows.

[0022] The flow rate sensor may be a sensor configured to measure the velocity of blood flow in a vector format. The vector format preferably includes three components corresponding to the x-axis, y-axis, and z-axis of the sensor. A plurality of flow rate sensors may be coupled to the sensor network to form part of the sensor network. In other words, a plurality of flow rate sensors may be configured as part of the sensor network. The flow rate sensor may be tilted (θ>0°) with respect to the blood flow velocity together with the catheter. For example, if the catheter is tilted at an angle greater than 0° or less than 0°, the two measurements may affect each other, i.e., one sensor exchanges power with a slightly warmed fluid, resulting in the two measurements being different. These measurements may then be sent to the sensor network individually or may be combined before sending to the sensor network.

[0023] The sensor network comprises components coupled to a number of flow rate sensors to enable data transmission regarding flow rate. The number of flow rate sensors may be coupled to the sensor network by wired or wireless coupling. If the coupling is wired, the wires coupling the components preferably have a small diameter so as not to interfere with readings from the flow rate sensors. If data from the flow rate sensors are transmitted individually to the sensor network, the sensor network may combine the received data measurements.

[0024] The processor is coupled to the sensor network and configured to receive data transmitted by the sensor network. The processor may be capable of calculating flow velocity via a mathematical model. The processor may be coupled to a memory that may be configured to store the results of the processor and / or store the mathematical model used by the processor. If the sensor network does not combine data measurements received from multiple flow velocity sensors, the processor may combine these measurements before using the measurements in the model. In some examples, the received measurements are not used in the model.

[0025] In some examples, the sensor network further comprises a pressure sensor, which is configured to sense the pressure in the blood vessel. The combination of the flow rate sensor and the pressure sensor may enable an indication of hyperemic stenosis resistance. The pressure sensor may be a piezoelectric pressure sensor based on the principle of Fabry-Perot interferometer and / or an optical pressure sensor. The signal from the pressure sensor may be used as an input for the mathematical model described herein. When the signal of the pressure sensor is used in combination with the flow rate sensor, this may provide a measurement of peripheral / vascular resistance, which may be particularly advantageous in clinical applications. This may significantly improve the diagnostic accuracy assessed by non-invasive ischemia testing, especially when results between conventional parameters are not consistent.

[0026] In some examples, the mathematical model includes or is a function, a polynomial function, a regression model, a lumped parameter model, a decision tree, a random forest, a neural network, or a numerical model, and the mathematical model is configured to output a velocity vector of the flow velocity. The mathematical function may be selected based on the parameters to be measured by the flow velocity sensor and / or the environment in which the catheter resides.

[0027] In some examples, the velocity vector is independent of the catheter orientation, which may allow for blood flow velocity measurements regardless of the orientation of the flow sensor, which may allow for more accurate measurements of blood flow velocity.

[0028] In some examples, the quality of the sensed parameter is configured to be evaluated by at least one of a regression coefficient, a correlation coefficient, or a fitting coefficient. This may allow the processor to compare results from the flow velocity sensor with expected results from a mathematical model. The quality of the results may then be transmitted to a display screen viewable by a user, allowing the user to modify the catheter position based on the displayed quality.

[0029] In some examples, the mathematical model includes information regarding the geometry of the catheter and / or the effect of the catheter on the flow, the information being configured to enable the mathematical model to compensate for the geometry of the catheter and / or the effect of the catheter on the flow, which may allow for more accurate measurements of blood flow because obstructions to blood flow caused by the catheter within the blood vessel may be mitigated and captured in the mathematical model.

[0030] In some examples, the output of the mathematical model is signaled to the user. This signaling may be done via a system of LEDs and / or a display screen. The displayed output may be the blood flow rate and / or pressure within the blood vessel. This may allow for a safer operation by allowing the user to abort the procedure if the blood flow rate and / or pressure fall outside of predefined parameters.

[0031] In some instances, the mathematical model is configured to be tailored to be specific to different vessel geometries and flow conditions, which may allow for more accurate measurements of blood flow. In particular, the mathematical model may be altered based on the diameter of the vessel (e.g., aorta or coronary artery) and / or based on retrograde flow conditions (e.g., arterial vessels vs. venous vessels).

[0032] In some examples, the mathematical model is adjusted to identify laminar and / or transition and / or turbulent flow conditions. The mathematical model may be adjusted by changing the parameters and / or hyperparameters of the model being used. The mathematical model may be adjusted by an experimental method in which a series of benchmark tests are performed and then the results of these tests are used to adjust the mathematical model. In some examples, the results are input into regression algorithms and the results of these algorithms are used to adjust the model. Additionally or alternatively, the mathematical model may be adjusted by an experimental method in which different simulations may be performed and then parameters are extracted from these simulations. These parameters may then be used to adjust the model. Additionally or alternatively, the mathematical model may be adjusted by machine learning methods. The use of machine learning is known to those skilled in the art. This may allow for more accurate measurement of blood flow velocity and / or may allow such measurement in case of problems occurring during the intervention. This may then allow the user to abort the intervention, thereby leading to a safer intervention process.

[0033] In some examples, the flow sensors are hot wire anemometer sensors, which may be particularly useful in situations where turbulent flow is present and may also enable analog outputs that may provide the opportunity for conditionally sampled time and frequency domain analysis and / or measurement of multi-component flows.

[0034] In some examples, each of the multiple flow rate sensors is configured to thermally affect at least one other flow rate sensor in the multiple flow rate sensors. The alignment of the catheter with respect to the blood flow may be determined by using thermal crosstalk between the multiple sensors. For example, when the catheter is precisely aligned with the flow rate stream, i.e., at an angle of 0° with respect to the blood flow, the multiple sensors may give the same measurement. This may allow for more accurate blood flow velocity measurements.

[0035] In some examples, the mathematical model is a numerical model, the mathematical model includes the Navier-Stokes equations, and the output of the Navier-Stokes equations is compared to the sensed blood flow velocity and a merit index is calculated based on the comparison. This allows the mathematical model to model specifically for viscous fluids, thereby allowing for more accurate measurement of blood flow velocity. The merit index lets the user know whether the calculated blood flow velocity is reliable, thereby maintaining the safety of the intervention.

[0036] In some examples, the mathematical model is a lumped parameter model, the lumped parameter model including or consisting of discrete entities configured to approximate the behavior of the outputs of the multiple flow rate sensors, the lumped parameter model comprising: It is specified by JPEG2024525076000003.jpg9150, where Q is the thermal energy in Joules, h is the heat transfer coefficient between the catheter and the blood flow, A is the surface area for heat transfer, T is the temperature at the surface of the catheter, and T env is the temperature of the environment and ΔT(t) is the time-dependent thermal gradient between the environment and the catheter. This may allow for improved blood velocity measurements and / or increased user safety by allowing the results of the calculations to be easily shown to the user.

[0037] In some examples, the device further includes an alarm that notifies the user if the blood flow rate is outside of a predetermined range, which may improve the safety of the intervention by easily notifying the user if there is a problem with the intervention procedure.

[0038] According to a second aspect, a method for measuring a flow velocity in a blood vessel by an apparatus is described. The apparatus includes a catheter configured to be inserted into the blood vessel, a plurality of flow velocity sensors coupled to the catheter, a sensor network coupled to the plurality of flow velocity sensors, and a processor coupled to the sensor network. The method includes sensing a velocity of blood flow in the blood vessel by each of the plurality of flow velocity sensors, inputting an output of the sensor network into a mathematical model stored in the processor, and calculating, by the mathematical model, the flow velocity in the blood vessel in which the catheter is located.

[0039] The catheter may be any commercially available catheter configured to be inserted into a blood vessel. The catheter may comprise any suitable material, such as, for example, PVC and / or rubber and / or silicone. The catheter is preferably configured to allow multiple flow rate sensors to be coupled to the catheter. The catheter is preferably cylindrical in cross section, but may alternatively be cubic, triangular, or custom shaped.

[0040] The flow rate sensor may be a sensor configured to measure the velocity of blood flow in a vector format. The vector format preferably includes three components corresponding to the x-axis, y-axis, and z-axis of the sensor. A plurality of flow rate sensors may be coupled to a sensor network to form part of the sensor network. In other words, a plurality of flow rate sensors may be configured as part of the sensor network.

[0041] The sensor network comprises components coupled to a number of flow rate sensors to enable data transmission regarding flow rate. The flow rate sensors may be coupled to the sensor network by wired or wireless coupling. If the coupling is a wired coupling, the wires coupling the components preferably have a small diameter so as not to interfere with readings from the flow rate sensors.

[0042] The processor is coupled to the sensor network and configured to receive data transmitted by the sensor network. The processor may be capable of calculating the flow velocity via a mathematical model. The processor may be coupled to a memory that may be configured to store the results of the processor and / or store the mathematical model used by the processor.

[0043] In some instances, not all of these steps are required. In some instances, the steps may be in a different order. In some instances, some of the steps occur simultaneously.

[0044] As will be apparent to those skilled in the art, the description provided herein may be implemented under the use of hardware circuits, software means, or a combination thereof. The software means may relate to a programmed microprocessor or a general computer, an ASIC (Application Specific Integrated Circuit) and / or a DSP (Digital Signal Processor). For example, the processing unit may be at least partially implemented as a computer, a logic circuit, an FPGA (Field Programmable Gate Array), a processor (e.g., a microprocessor, a microcontroller (μC), or an array processor) / core / CPU (Central Processing Unit), an FPU (Floating Point Unit), an NPU (Numerical Processing Unit), an ALU (Arithmetic Logic Unit), a coprocessor (an additional microprocessor to support a main processor (CPU)), a GPGPU (General Purpose Computing on a Graphics Processing Unit), a multi-core processor (for parallel computing, such as performing arithmetic operations on multiple main processors and / or graphic processors simultaneously), or a DSP.

[0045] Moreover, as will be apparent to those skilled in the art, even if the details described herein are described in terms of a method, these details may be implemented or realized in a suitable device, computer processor, or memory connected to the processor, and the memory may be provided with one or more programs that, when executed by the processor, perform the method. Thus, methods such as swapping and paging may be deployed.

[0046] Even though some of the above aspects are described in relation to an apparatus, these aspects may also be applied to a method, and vice versa. These and other aspects of the invention will now be further described, by way of example only, with reference to the accompanying drawings, in which like reference numbers refer to like parts and in which: [Brief description of the drawings]

[0047] [Figure 1] 1 shows a three-axis geometry according to the prior art. [Diagram 2] FIG. 1 shows a diagram of a catheter and flow rate sensor according to the prior art. [Diagram 3] FIG. 1 shows a diagram of a catheter and flow rate sensor according to the prior art. [Figure 4] 1 shows the velocity distribution around a cylinder according to the prior art. [Diagram 5] 1 shows the velocity distribution around a cylinder according to the prior art. [Figure 6] FIG. 1 shows a diagram of a catheter and flow rate sensor according to an embodiment described herein. [Figure 7] FIG. 1 shows a diagram of a catheter, flow sensor, and sensor network according to embodiments described herein. [Figure 8] FIG. 2 illustrates a flow diagram of data flow according to embodiments described herein. [Figure 9] FIG. 1 shows a perspective view of a catheter according to an embodiment described herein. [Figure 10] 1 shows a cutaway view of a catheter according to an embodiment described herein. [Figure 11]FIG. 2 illustrates a flow diagram of data flow according to embodiments described herein. [Figure 12] FIG. 1 shows a block diagram of a method for measuring flow velocity in a blood vessel according to embodiments described herein. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0048] FIG. 1 shows a prior art three-axis geometry. In the prior art, the flow of blood in a vessel is primarily along the x-axis. The blood flow velocity is measured by a flow velocity sensor with respect to the yaw direction of the vessel, i.e., around the z-axis. The prior art does not consider the pitch direction, i.e., around the y-axis, due to the radial symmetry of the vessel. Thus, roll information represents an important degree of freedom that must be considered.

[0049] 2 and 3 show diagrams of a catheter and flow sensor according to the prior art. Blood vessel 10 includes blood flow in substantially the same direction as the direction of arrow 12. Positioned within blood vessel 10 is a catheter 14 having a single flow rate sensor 16 coupled thereto. Flow rate sensor 16 is configured to sense the rate of blood flow within blood vessel 10.

[0050] The catheter 14 may be optimally oriented as shown in FIG. 2 when the roll angle is 0°. In this optimal orientation, the major plane of the flow sensor 16 is aligned with the blood flow. Alternatively, the catheter 14 may be suboptimally oriented as shown in FIG. 3 when the roll angle is 90°. In this suboptimal orientation, the major plane of the flow sensor 16 is located in an area of ​​minimum velocity, i.e., hidden behind the catheter 14. Furthermore, because the flow sensor 16 is behind the catheter 14, the flow sensor may not accurately measure the velocity of the blood flow within the blood vessel 10.

[0051] 4 and 5 show the velocity distribution around a cylinder according to the prior art. FIG. 4 shows the velocity field when the catheter 14 is inside the blood vessel 10. Furthermore, in Figure 5, the flow velocity distribution around the cylindrical shape can be observed. - In the boundary layer at the surface 20 facing the flow, the velocity is very low because in this region the flow is stopped by the cylinder, - in the boundary layer at the upper and lower surfaces 22, 24 of the catheter 14, the velocity is higher because the flow is accelerated by the catheter 14; - in the boundary layer at the surface 26 behind the catheter 14, the flow velocity is very low because the flow is in the shadow of the catheter 14 itself; It becomes clear that...

[0052] This type of fluid dynamics is known to those skilled in the art. As mentioned above, this means that it is difficult for a single flow sensor placed on the surface of the catheter 14 that is not perfectly aligned with the flow to measure the actual vascular flow velocity. This is due to the fact that the catheter 14 has a non-negligible geometry in the flow and in any case the catheter 14 will accelerate or decelerate the flow velocity due to fluid dynamics.

[0053] FIG. 6 shows a diagram of a catheter and flow rate sensor according to an embodiment described herein. FIG. 6 shows a blood vessel 100 in which the blood flow therein is moving in a direction substantially parallel to the arrow 102. A catheter 104 is positioned within the blood vessel 100. Two flow rate sensors 106 are directly coupled to the catheter 104. The flow rate sensors 106 may be capable of sensing the velocity of the blood flow by pressure, in which case the pressure exerted by the blood flow on the flow rate sensors 106 may allow the flow rate to be calculated. The flow rate sensors 106 may comprise electronic circuitry that allows the flow rate sensors 106 to calculate the flow rate sensed by the sensors 106. In this embodiment, two flow rate sensors 106 are coupled to the catheter 104 on either side of the catheter 104, i.e., the sensors 106 are positioned 180° apart from each other. In some embodiments, the flow rate sensors 106 are not positioned 180° apart from each other, but may be any suitable angle apart from each other, such as, for example, 45°, 90°, or 120°. In some embodiments, there are three or more flow rate sensors 106. In embodiments with three or more flow sensors 106, the flow sensors 106 may be equally spaced apart from one another. For example, if there are three flow sensors 106, the sensors 106 may be spaced 120 degrees apart from one another. Alternatively, the sensors 106 may be spaced any suitable distance apart from one another.

[0054] FIG. 7 shows a diagram of a catheter, flow sensor, and sensor network according to embodiments described herein. FIG. 7 shows a sensor network 108 coupled to a flow sensor 106 on the catheter. The coupling between the flow sensor 106 and the sensor network 108 may be a wired or wireless coupling. In a wireless coupling, each flow sensor 106 may comprise a transmitter configured to transmit its sensed readings to the sensor network 108, which is configured to receive the transmitted readings. In this embodiment, the sensor network 108 may be located in or coupled to a computer and / or processor (see FIG. 8) and / or in a cloud storage. The sensor network 108 comprises two or more sensors 106 integrated in a sensor housing (see FIG. 9). It is particularly preferred that the sensors 106 are symmetrically positioned around the catheter 104. For example, if the catheter 104 has three sensors 106, the sensors 106 are preferably positioned 120° apart. This is particularly advantageous for measuring flow rates.

[0055] In a wired connection, the sensor network 108 may include wires coupled to each flow sensor 106, the wires configured to transmit sensed readings of the flow sensors 106 to, for example, a computer and / or processor. The wires may also be coupled to the catheter 104 such that they do not become loose within the blood vessel 100 and / or minimize disruption of the fluid dynamics around the catheter. Additionally or alternatively, the wires may be located within the catheter 104.

[0056] The sensor network 108 may also be able to collate sensed readings from multiple flow rate sensors 106 into a single three-dimensional velocity vector. In some embodiments, the sensor network does not collate the sensed readings but keeps them in separate data flows.

[0057] For ease of illustration, the catheter is shown in cutaway view in Figure 7. As will be appreciated by those skilled in the art, the sensor network 108 may be in any suitable orientation relative to the flow sensor 106 and catheter 104. In some examples, additional sensors may be present, such as pressure sensors and / or sensors that sense collisions with the walls of the blood vessel 100 and / or any other suitable sensors configured to assist in the intervention.

[0058] FIG. 8 shows a flow diagram of data flow according to embodiments described herein. FIG. 8 illustrates that sensed readings from the flow sensor 106 are transmitted to a processor 110 via a sensor network 108 of which the flow sensor 106 is a part. The processor 110 may comprise a processing unit. The processor may be coupled to a memory unit and / or a display unit, as described in more detail below. The coupling between the flow sensor 106 and the sensor network 108 may be wired or wireless, as previously described. The coupling between the sensor network 108 and the processor 110 may be wired or wireless. In a wired coupling, an electrical cable of the sensor network 108 may transmit the sensed readings directly to the processor 110, as previously described. In a wireless coupling, the sensor network 108 may comprise a transmitter configured to transmit the sensed readings of the flow sensor 106, and the processor 110 may comprise a receiver configured to receive the sensed readings from the sensor network 108.

[0059] FIG. 9 shows a perspective view of a catheter according to an embodiment described herein. FIG. 9 shows a catheter 104 configured to be inserted into a blood vessel 104. On the catheter are a number of flow rate sensors 106 and a pressure sensor 107. The pressure sensor may be similar to the pressure sensor 107 described above. The sensors 106, 107 are located within a housing 112 located at the distal end of the catheter 104. The sensor housing 112 may comprise any material suitable for use in a catheter, such as plastic. The housing 112 is preferably flexible to allow easier movement of the catheter 104. The sensor network 108 is preferably located within the housing 112. At the distal end of the catheter 104 is a tip 114 suitable for intervention. The sensor network 108 transmits measurement data from the number of sensors 106, 107 via a hypotube 116 to a processor 110 located at an external location. Alternatively, the hypotube 116 may be any type of tube suitable for transmitting data and being used as part of the catheter 104. The remaining parts of the catheter 104 are known to those skilled in the art.

[0060] FIG. 10 shows a cutaway view of a catheter according to an embodiment described herein. The catheter 104 preferably has sensors 106, 107 positioned at equal angular intervals around the catheter 104. For example, if the catheter 104 has three sensors 106, 107, the sensors 106, 107 are preferably positioned 120° apart. This is particularly advantageous for measuring flow velocity. In some embodiments, the catheter 104 comprises a channel 118. This channel 118 is preferably used to transmit data from the sensors 106, 107 to the sensor network 108 and then to the processor 110. Alternatively, this channel 118 may be used for additional instruments that may need to be used during the intervention, such as a guidewire.

[0061] FIG. 11 shows a flow diagram of data flow according to embodiments described herein. In the processor 110, the processor 110 receives the measured sensory data 150 from the flow sensor 106 via the sensor network 108, as previously described. The processor 110 receives, either within the processor 110 itself or via a memory unit (not shown), a mathematical model 152. The mathematical model 152 may be capable of estimating an expected flow velocity sensed by the flow sensor 106 and / or the flow velocity sensed by the flow sensor 106 may be input into the model 152 itself.

[0062] The mathematical model 152 may have different properties. For example, the model 152 may be a polynomial function whose coefficients are adjusted to provide an accurate flow measurement. JPEG2024525076000004.jpg6150, where f( ) is a polynomial function, x1,…,x n are the different measurement inputs, JPEG2024525076000005.jpg713 is a velocity vector. The velocity vector output by the polynomial function is a vector containing three components, each component associated with a different axis of the vessel, the x-axis, the y-axis, and the z-axis. In some embodiments, the velocity vector has fewer than three components. The axes represented by this reduced vector can be changed based on the user's preferences.

[0063] The mathematical model 152 may be a polynomial function as described above and / or may include at least one of the following mathematical models 152. In all of the mathematical models 152 described below, one of ordinary skill in the art will understand the limitations of each model 152 and possible modifications that may be made to the model 152.

[0064] - Univariate regression A feature is selected that adequately represents the effect of the orientation of the catheter 104, which feature allows the correction to compensate for the fluid dynamics around the catheter 104. The feature may be, for example, a quantity representing the raw signals received from the sensors 106, 107. The feature may additionally or alternatively be a ratio between the raw signals received from the sensors 106, 107. Equation 1 is a function of the feature x, the objective y, and the weights c to represent the fluid dynamics around the catheter 104. i shows a general representation of a univariate regression equation with , which can be used to estimate flow velocity. The objective may be, for example, the flow velocity in the blood vessel 100, and the weights may be model parameters. The selection of features and objective (correction), as well as the order n of the model, are possible hyperparameters. There is little possibility for tuning this model 152.

[0065] [Formula 1] JPEG2024525076000006.jpg13150

[0066] -Multivariate Linear Regression Multivariate linear regression is similar to univariate regression. Equation 2 expresses the i-th feature value x i , objective y, and weight c i We show a general representation of multivariate linear regression with the objective (correction), the number of features m, and the configuration of these features being possible hyperparameters.

[0067] [Formula 2] JPEG2024525076000007.jpg13150

[0068] -Decision Tree A decision tree algorithm includes or consists of nodes, branches, and leaves. During fitting of the model 152, a comparison can be established for each node. Depending on the value given, the decision follows one of two branches to the next node. Finally, when the bottom leaf of the decision tree is reached, a decision is made. Decision trees are highly tunable and can provide various hyperparameters that can be tuned, as listed in Table 1 below.

[0069] [Table 1] Overview of available hyperparameters for decision trees TIFF2024525076000008.tif36157

[0070] Decision trees can predict a wide range of data due to their tunability and versatility. Furthermore, decision trees can be used for regression as well as classification. However, this gives the risk of overfitting the model to the data. Furthermore, for large trees with a large depth and a large number of nodes / leaves, the implementation and prediction can be time-consuming. As will be appreciated by those skilled in the art, the various parameters can be, for example, the blood flow velocity of a single component of the velocity vector, and / or the increase or decrease in velocity, and / or whether the component of the velocity vector is outside the range of a given parameter.

[0071] -Random Forest The Random Forest model is an ensemble method, so that the estimates of multiple models are taken into account for the final estimation. In the case of Random Forest, the ensemble may include or consist of a certain amount of decision trees "n_estimators". Each tree is specified individually for a sample of the complete data "max_samples". These two parameters are additional hyperparameters to the parameters from the individual decision trees and can be used to tune the model 152. Finally, each decision tree makes a prediction and the final result is the average of the individual estimates of the decision trees.

[0072] - Centralized parameters A lumped parameter physics model simplifies complex physical phenomena into a topology that includes or consists of discrete entities that approximate the behavior of a distributed system. This model 152 may be defined by: where Q is the thermal energy in Joules, h is the heat transfer coefficient between the catheter 104 and the blood flow, A is the surface area for heat transfer, T is the temperature of the surface of the catheter 104, and T env is the temperature of the environment, and ΔT(t) is the time-dependent thermal gradient between the environment and the catheter 104.

[0073] -Deep Neural Networks for Regression A neural network is a collection of connected nodes or units in a layered structure. Each unit corresponds to a nonlinear function that takes as input the output of the previous layer and gives as output a weighted sum after applying a nonlinear function, i.e. an activation function. The last layer is the output layer, which does away with the nonlinear functions seen in previous layers and outputs a weighted sum of its inputs.

[0074] To calculate the weights for each weighted sum, a neural network is trained by processing input and output examples to optimize a loss function. The neural network uses an optimization algorithm based on gradient descent and backpropagation to adjust the weights of all nodes in all layers. This can enable particularly accurate deep learning mathematical models.

[0075] -Recurrent Neural Networks Recurrent neural networks (RNNs) are a subclass of neural networks that take into account previous outputs by having hidden states, and therefore exhibit dynamic behavior over time. Gated recurrent units (GRUs) and long short-term memory units (LSTMs) are subtypes of RNNs that address the vanishing gradient problem, allowing them to capture long-term dependencies.

[0076] When training such a network, backpropagation is performed at each time point (backpropagation through time), so that every time point influences the weights of each single unit. This may allow for particularly accurate models due to improvements in deep learning techniques.

[0077] -Numerical model The results of a numerical solution of the Navier-Stokes equations, which describe the flow velocity distribution around the catheter 104, are used to compare measurements taken by the flow velocity sensors. A figure of merit can be generated to evaluate the fit between the measurements and the model 152, which defines the three-dimensional flow velocity vectors.

[0078] In addition to the above, the mathematical model 152 may include a correction component configured to correct for the effects of fluid dynamics surrounding the catheter 104 and / or unknown orientation of the flow rate sensor 106 and / or the catheter 104. The correction component may be at least one of the following:

[0079] - Output compensation Considering the effect of the yaw direction on the output-velocity curve of the sensor samples may allow possible ways to compensate for the yaw direction. The yaw direction may be compensated for by adjusting the output value of the sensor 106 depending on the orientation of the flow velocity sensor 106 and / or catheter 104 so that the output is within a certain range regardless of the orientation. There are two possible approaches to adjusting the output. First, the output can be shifted up or down by adding / subtracting the output value depending on the orientation. Second, the output can be scaled by a factor that depends on the orientation.

[0080] Generally, the required correction is defined by determining a representative power-speed curve and calculating the necessary shift / coefficient to move the measured value to the determined curve. As possible representative pv curves, the minimum, average and maximum values ​​of the selected data set can be considered. However, due to the flattening behavior of the pv curve, the minimum value is not considered to be a good choice, since it brings a high risk that the correction value may be outside the valid range.

[0081] Finally, it is necessary to define the partition of the data set for which the representative curve is calculated. Two main approaches are therefore identified: First, each sensor sample is considered individually. This means that a representative curve is calculated for each sensor sample. The advantage is that this ignores the differences between the sensor samples. However, this requires a separate calibration for each sensor 106 to determine the sensor's behavior. Secondly, all sensors 106 are considered with respect to the representative curve. The advantage is that this correction allows a good generalization of the sensor's 106 behavior. However, this may lead to a loss of accuracy if the differences between the samples are not taken into account.

[0082] -Evaluation of measurement accuracy The quality of the data can be assessed using regression coefficients, correlation indices, and indices for assessing the fitting of the data to the model, which can give information about the quality of the measurements collected. In this way, the system model-sensor network 108 can also estimate non-optimal measurements caused by non-optimal exposure to flow, such as when the sensor 106 is touching the wall of the blood vessel 100. The non-optimal measurements can be estimated by comparing different measurements in the network 108 and by defining indicators that may indicate poor data quality (e.g., correlations with regression indicators of models or functions).

[0083] The model 152 may be selected by a user based, for example, on the diameter of the blood vessel 100, where the catheter 104 is positioned, and / or the retrograde flow of the blood vessel 100. This modification may be made by the user via a display coupled to the processor 110. Alternatively, the modification of the model 152 may be made automatically by the processor 110 in the form of a machine learning algorithm within the processor 110.

[0084] After the model 152 is selected, the sensed readings 150 are then compared to the results of the model 152 in a comparison unit 154. The sensed readings 150 may be run through the model 152 before the comparison is made. The comparison unit 154 then transmits data regarding this comparison to a quality measurement unit 156. The quality measurement unit 156 measures the quality of the sensed reading 150 with respect to the estimated velocity vector from the model 152. The quality measurement unit 156 then makes a decision on the quality of the sensed reading 150. This decision may then be displayed on a display unit coupled to the processor 110 to indicate the quality of the reading 150 to a user. The user can then make a decision regarding an intervention based on the quality of the reading 150.

[0085] In some embodiments, the processor further comprises an alarm unit 158. The alarm unit 158 ​​may inform the user when a component of the blood flow velocity vector falls outside a predefined range and / or when the comparison unit 154 finds a large discrepancy between the model 152 and the sensed readings 150. The alarm may be an audio alarm, and / or a tactile alarm, and / or a visual alarm. If the alarm is a visual alarm, it may be displayed on a display unit coupled to the processor 110.

[0086] FIG. 12 shows a block diagram of a method for measuring flow velocity in a blood vessel according to embodiments described herein. The method 200 for measuring flow velocity in a blood vessel 100 consists of four main steps.

[0087] First, the velocity of the blood flow in the blood vessel 100 is sensed by each of the multiple flow velocity sensors 106 (S210). This allows the flow velocity in the blood vessel 100 to be accurately determined. The velocities sensed by each of the flow sensors 106 are then transmitted (S220) to the sensor network 108. This may allow the sensed velocities to be collated into a single reading.

[0088] The output of the sensor network 108 is then input (S230) to a mathematical model 152 stored in the processor 100. This allows the sensed velocity 150 to be corrected for fluid dynamic irregularities and / or to more accurately define the blood flow velocity within the blood vessel 100. The flow velocity in the blood vessel in which the catheter 104 is located is then calculated (S240) by the mathematical model 152. This may allow for an accurate reading of the blood flow velocity.

[0089] Many other effective alternatives will no doubt occur to those skilled in the art, and it can be understood that the invention is not limited to the described embodiments, but encompasses modifications that are obvious to those skilled in the art and fall within the scope of the appended claims.

Claims

1. An apparatus for measuring the flow velocity within a blood vessel (100), comprising: a catheter (104) configured to be inserted into the blood vessel (100); a plurality of flow velocity sensors (106) coupled to the catheter (104); a sensor network (108) coupled to the plurality of flow velocity sensors (106); a processor (110) coupled to the sensor network (108); wherein each of the plurality of flow velocity sensors (106) is configured to sense the velocity of the blood flow; the output of the sensor network (108) is configured to be input into a mathematical model (152) stored in the processor (110); and the mathematical model (152) is configured to calculate the flow velocity within the blood vessel (100) in which the catheter (104) is located. An apparatus.

2. The apparatus according to claim 1, wherein the sensor network (108) further comprises a pressure sensor configured to sense the pressure within the blood vessel (100).

3. The apparatus according to claim 1 or 2, wherein the mathematical model (152) includes or is a function, a polynomial function, a regression model, a lumped parameter model, a decision tree, a random forest, a neural network, or a numerical model, and the mathematical model (152) is configured to output a velocity vector of the flow velocity.

4. The apparatus according to claim 3, wherein the velocity vector is independent of the orientation of the catheter.

5. The apparatus according to claim 1 or 2, wherein the quality of the sensed parameter (150) is configured to be evaluated by at least one of a regression coefficient, a correlation coefficient, or a fitting coefficient.

6. The apparatus according to claim 1 or 2, wherein the mathematical model (152) includes information regarding the geometric shape of the catheter (152) and / or the influence of the catheter (104) on the flow velocity, and the information is configured such that the mathematical model (152) can compensate for the influence of the geometric shape of the catheter (104) and / or the flow velocity on the catheter (104).

7. The apparatus according to claim 1 or 2, wherein the output of the mathematical model (152) is communicated to the user as a signal.

8. The apparatus according to claim 1 or 2, wherein the mathematical model (152) is configured to be adjusted to be specific to the geometric shapes and flow states of different blood vessels (100).

9. ​ The device according to claim 1 or 2, wherein the mathematical model (152) is adjusted to identify a laminar flow state and / or a transitional flow state and / or a turbulent flow state in the blood vessel (100).

10. The device according to claim 1 or 2, wherein the plurality of flow velocity sensors (106) are hot-wire anemometer sensors.

11. The device according to claim 10, wherein each of the plurality of flow velocity sensors (106) is configured to thermally affect at least one other flow velocity sensor (106) among the plurality of flow velocity sensors (106).

12. The device according to claim 3, wherein the mathematical model (152) is a numerical model, the mathematical model (152) includes the Navier-Stokes equation, the output of the Navier-Stokes equation is compared with the sensed blood flow velocity, and a merit index is calculated based on the comparison.

13. The mathematical model (152) is a lumped parameter model, and the lumped parameter model includes or consists of discrete entities configured to approximate the behavior of the outputs of the plurality of flow velocity sensors. The lumped parameter model is defined by where Q is thermal energy in joules, h is the heat transfer coefficient between the catheter (104) and the blood flow, A is the surface area of heat transfer, T is the temperature of the surface of the catheter (104), Tenv is the environmental temperature, and ΔT(t) is the time-dependent thermal gradient between the environment and the catheter (104). The device according to claim 3.

14. The device according to claim 1 or 2, further comprising an alarm, wherein the alarm notifies the user whether the blood flow velocity is outside a predetermined range.

15. A method (200) for measuring the flow velocity in a blood vessel by a device, wherein the device comprises a catheter configured to be inserted into a blood vessel, a plurality of flow velocity sensors coupled to the catheter, a sensor network coupled to the plurality of flow velocity sensors, a processor coupled to the sensor network, and the method comprises a step (S210) of sensing the velocity of the blood flow in the blood vessel by each of the plurality of flow velocity sensors, a step (S220) of transmitting the velocity sensed by each of the flow velocity sensors to the sensor network, a step (S230) of inputting the output of the sensor network into a mathematical model stored in the processor. ​ A step (S240) of calculating a flow velocity in a blood vessel in which the catheter is positioned by the mathematical model; A method including the above.

Images