Control of position of interface between separated blood components using artificial intelligence

Abstract

Systems and methods are provided for controlling the location of an interface between separated blood components within a centrifuge. An optical monitoring assembly determines the current position of the interface, with the current position and a plurality of input parameters being transmitted from a controller to a neural network. The neural network employs artificial intelligence techniques to analyze the current interface position and at least one of the input parameters to determine a modification to be made to at least one procedural parameter so as to move the interface from its current position to a target position. The neural network transmits a signal that is indicative of the modification to the controller, which implements the modification so as to move the interface toward the target position, with the modification being to an inflow rate, a centrifuge rotation rate, an outflow rate, and / or the location of the target interface position.

Description

[0001] CONTROL OF POSITION OF INTERFACE BETWEEN SEPARATED BLOOD COMPONENTS USING ARTIFICIAL INTELLIGENCE

[0002] RELATED APPLICATIONS

[0003] This application claims the benefit of and priority of U. S. Provisional Patent Application Serial No. 63 / 736,688, filed December 20, 2024, the contents of which are incorporated by reference herein.

[0004] DESCRIPTION TECHNICAL FIELD

[0005] The present disclosure relates to systems and methods for controlling the position of an interface between separated blood components within a centrifuge. More particularly, the present disclosure relates to interface position control using artificial intelligence.

[0006] A wide variety of blood processing systems are presently in practice and

[0007]

[0008] allow for a blood to be fractionated or separated into its constituent parts, thus making it possible to collect particular blood constituents, rather than whole blood, from a blood source. Typically, in such systems, whole blood is drawn from a blood source, the particular blood component or constituent is separated, removed, and collected, and the remaining blood constituents are returned to the blood source. Removing only particular constituents is advantageous when the blood source is a human donor or patient, because potentially less time is needed for the donor's body to return to pre-donation levels, and donations can be made at more frequent intervals than when whole blood is collected. This increases the overall supply of blood constituents, such as plasma and platelets, made available for transfer and / or therapeutic treatment.

[0009] Whole blood is typically separated into its constituents through centrifugation. In continuous processes, this requires that the whole blood be passed through a centrifuge after it is withdrawn from, and before it is returned to, the blood source. To avoid contamination and possible infection (if the blood source is a human donor or patient), the blood is preferably contained within a preassembled, sterile fluid flow circuit or system during the entire centrifugation process. Typical blood processing systems thus include a permanent, reusable module or assembly containing the durable hardware (centrifuge, drive system, pumps, valve actuators, programmable controller, and the like) that controls the processing of the blood and blood components through a disposable, sealed, and sterile flow circuit that includes a centrifugation chamber and is mounted in cooperation on the hardware.

[0010] The hardware engages and spins the disposable centrifugation chamber during a blood separation step. As the flow circuit is spun by the centrifuge, the heavier (greater specific gravity) components of the whole blood in the flow circuit, such as red blood cells, move radially outwardly away from the center of rotation toward the outer or "high-G" wall of the centrifugation chamber. The lighter (lower specific gravity) components, such as plasma, migrate toward the inner or "low-G" wall of the centrifuge. Various ones of these components can be selectively removed from the whole blood by providing appropriately located outlet ports in the flow circuit.

[0011] Detection and control of the location of the interface between separated blood components (typically, between red blood cells and plasma) within the centrifugation chamber is important for efficient collection of the separated components. The interface control systems of conventional blood processing systems apply optical-based methods and proportional-integral-derivative (" PID”) control logic to establish and maintain the interface at the target position. More particularly, a PID control algorithm is employed to minimize the difference between the target position of the interface and the actual or current position of the interface by considering the present difference between the two interface positions (“the P term”), how long and how far the interface position has been from the target position in the past (“the I term”), and a rate of change of the interface position between two time points (“the D term”).

[0012] In nearly all known predicate systems, the input of the PID control algorithm is limited to the position of the interface, though U. S. Patent Application Publication No. 2024 / 0016992 (which is hereby incorporated herein by reference) provides an interface position control algorithm employing additional input, including blood characteristics and blood source characteristics, with the parameters of the PID control algorithm being updated periodically during a procedure.

[0013] SUMMARY

[0014] There are several aspects of the present subject matter which may be embodied separately or together in the devices, systems, and methods described and / or claimed below. These aspects may be employed alone or in combination with other aspects of the subject matter described herein, and the description of these aspects together is not intended to preclude the use of these aspects separately or the claiming of such aspects separately or in different combinations as set forth in the claims appended hereto or later amended.

[0015] The following summary is to acquaint the reader generally with various potential aspects of the present subject matter, and is non-limiting and nonexclusive with respect to the various possible aspects or combinations of aspects. Additional aspects and features may be found in the detailed description herein and / or in the accompanying figures.

[0016] In one aspect, a blood separation device includes a pump system, a centrifugal separator, an interface monitoring assembly, a neural network, and a controller coupled to the neural network, with the controller being programmed to execute a blood separation procedure. The blood separation procedure includes the controller actuating the pump system to pump blood into the centrifugal separator at an inflow rate; actuating the centrifugal separator to operate at a rotational rate to separate the blood in the centrifugal separator into two blood components, with an interface therebetween; and actuating the pump system to pump one of the blood components out of the centrifugal separator at an outflow rate. The controller actuates the interface monitoring assembly to determine the current position of the interface within the centrifugal separator and transmits the current position of the interface and a plurality of input parameters to the neural network. The neural network is programmed to employ artificial intelligence techniques to analyze the current position of the interface and at least one of the input parameters and determine a modification to be made to at least one procedural parameter so as to move the interface from the current position to a target interface position, with the modification being based at least in part on the current position of the interface and at least one of the input parameters. The neural network transmits a signal to the controller that is indicative of the modification to be made, with the controller actuating the pump system and / or the centrifugal separator so as to implement the modification and move the interface from the current position toward the target interface position. The procedural parameter to be modified is selected from the group consisting of the inflow rate, the rotational rate, the outflow rate, and a location of the target interface position.

[0017] In another aspect, a controller-implemented method is provided for executing a blood separation procedure. The method includes actuating a pump system to pump blood into a centrifugal separator at an inflow rate; actuating the centrifugal separator to operate at a rotational rate to separate the blood in the centrifugal separator into two blood components, with an interface therebetween; and actuating the pump system to pump one of the blood components out of the centrifugal separator at an outflow rate. An interface monitoring assembly is actuated to determine the current position of the interface within the centrifugal separator, with the current position of the interface being transmitted to a neural network, along with a plurality of input parameters. The neural network employs artificial intelligence techniques to analyze the current position of the interface and at least one of the input parameters and determines a modification to be made to at least one procedural parameter so as to move the interface from the current position to a target interface position, with the modification being based at least in part on the current position of the interface and at least one of the input parameters. The neural network transmits a signal to the controller that is indicative of the modification to be made, with the controller actuating the pump system and / or the centrifugal separator so as to implement the modification and move the interface from the current position toward the target interface position. The procedural parameter to be modified is selected from the group consisting of the inflow rate, the rotational rate, the outflow rate, and a location of the target interface position.

[0018] These and other aspects of the present subject matter are set forth in the following detailed description of the accompanying drawings.

[0019] BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a perspective view of an exemplary blood separation device according to an aspect of the present disclosure;

[0020] Fig. 2 is a schematic view of an exemplary disposable fluid flow circuit that may be mounted to the blood separation device of Fig. 1 to complete a blood separation system;

[0021] Fig. 3 is a perspective view of an exemplary centrifugal separator of the blood separation device of Fig. 1, with the centrifugal separation chamber of a fluid flow circuit mounted therein;

[0022] Fig. 4 is a top plan view of an exemplary cassette of a fluid flow circuit, which can be actuated to perform a variety of different blood processing procedures in association with the blood separation device shown in Fig. 1;

[0023] Fig. 5 is a perspective view of the centrifugal separator of Fig. 3, with selected portions thereof broken away to show a light source of an interface monitoring system;

[0024] Fig. 6 is a perspective view of the centrifugal separator of Fig. 3, with the light source operating to transmit a light beam to a light detector of the interface monitoring system;

[0025] Fig. 7 is a perspective view of the centrifugal separator of Fig. 3, with selected portions thereof broken away to show the light source and light detector of the interface monitoring system;

[0026] Fig. 8 is a perspective view of an exemplary centrifugal separation chamber of a fluid flow circuit;

[0027] Fig. 9 is a front elevational view of the centrifugal separation chamber of Fig. 8;

[0028] Fig. 10 is an enlarged perspective view of a portion of a channel of the centrifugal separation chamber of Figs. 8 and 9, with an interface between separated blood components being positioned at a desired or target location on a ramp defined within the channel;

[0029] Fig. 11 is an enlarged perspective view of the channel of Fig. 10, with the interface being at an undesired high location on the ramp;

[0030] Fig. 12 is an enlarged perspective view of the channel of Fig. 10, with the interface being at an undesired low location on the ramp; Fig. 13 is a perspective view of a prismatic reflector used in combination with the centrifugal separation chamber of Figs. 8 and 9;

[0031] Fig. 14 is a perspective view of the prismatic reflector of Fig. 13, showing light being transmitted therethrough;

[0032] Figs. 15-18 are diagrammatic views of the ramp and prismatic reflector of the centrifugal separation chamber passing through the path of light from the light source during a calibration phase;

[0033] Figs. 19-22 are diagrammatic views of the voltage output or signal transmitted by the light detector during the conditions shown in Figs. 15-18, respectively;

[0034] Figs. 23-26 are diagrammatic views of the ramp and prismatic reflector passing through the path of light from the light source during a separation procedure;

[0035] Figs. 27-30 are diagrammatic views of the voltage output or signal transmitted by the light detector during the conditions shown in Figs. 23-26, respectively;

[0036] Figs. 31 and 32 are diagrammatic views of separated blood components on the ramp and the pulse widths of a signal generated by the light detector for each condition;

[0037] Fig. 33 is a diagrammatic view of saline on the ramp and the pulse width of a signal generated by the light detector for such a condition;

[0038] Fig. 34 is a diagrammatic view of the position of an interface between separated blood components on the ramp compared to a target interface position;

[0039] Fig. 35 is a diagrammatic view of a conventional approach to controlling the position of an interface between separated blood components;

[0040] Fig. 36 is a diagrammatic view of a conventional proportional-integral- derivative control algorithm implemented to control interface position during a blood separation procedure;

[0041] Fig. 37 is a diagrammatic view of an approach to controlling the position of an interface between separated blood components according to an aspect of the present disclosure; and

[0042] Fig. 38 is a diagrammatic view of an exemplary neural network employed in executing the interface control routine of Fig. 37. DETAILED DESCRIPTION OF THE ILLUSTRATED EMBODIMENTS The embodiments disclosed herein are for the purpose of providing a description of the present subject matter, and it is understood that the subject matter may be embodied in various other forms and combinations not shown in detail. Therefore, specific designs and features disclosed herein are not to be interpreted as limiting the subject matter as defined in the accompanying claims.

[0043] Figs. 1 -3 show components of a blood or fluid separation system that embodies various aspects of the present subject matter. Generally speaking, the system includes two principal components, a durable and reusable blood separation device 10 (Fig. 1 ) and a disposable fluid flow circuit 12 (Fig. 2). The blood separation device 10 includes a spinning membrane separator drive unit 14 (Fig. 1), a centrifuge or centrifugal separator 16 (Fig. 3), additional components that control fluid flow through the disposable flow circuit 12, and a controller 18 (Fig. 1), which governs the operation of the other components of the blood separation device 10 to perform a blood processing and collection procedure selected by an operator.

[0044] I. The Durable Blood Separation Device

[0045] The blood separation device 10 (Fig. 1) is configured as a durable item that is capable of long-term use. It should be understood that the blood separation device 10 of Fig. 1 is merely exemplary of one possible configuration and that blood separation devices according to the present disclosure may be differently configured. For example, it is within the scope of the present disclosure for the blood separation device to omit a spinning membrane separator drive unit 14 and to include only a centrifugal separator 16, as in the blood separation device described in PCT Patent Application Publication No. WO 2021 / 194824 A1 (which is hereby incorporated herein by reference)

[0046] In the illustrated embodiment, the blood separation device 10 is embodied in a single housing or case 20. The illustrated case 20 includes a generally horizontal portion 22 (which may include an inclined or angled face or upper surface for enhanced visibility and ergonomics) and a generally vertical portion 24. The spinning membrane separator drive unit 14 and the centrifugal separator 16 are shown as being incorporated into the generally horizontal portion 22 of the case 20, while the controller 18 is shown as being incorporated into the generally vertical portion 24. The configuration and operation of the centrifugal separator 16 and selected other components of the blood separation device 10 will be described in greater detail.

[0047] In the illustrated embodiment, the generally horizontal portion 22 is intended to rest on an elevated, generally horizontal support surface (e.g., a countertop or a tabletop), but it is also within the scope of the present disclosure for the case 20 to include a support base to allow the case 20 to be appropriately positioned and oriented when placed onto a floor or ground surface. It is also within the scope of the present disclosure for the case 20 to be mounted to a generally vertical surface (e.g., a wall), by either fixedly or removably securing the generally vertical portion 24 of the case 20 to the surface.

[0048] The case 20 may be configured to assume only the position or configuration of Fig. 1 or may be configured to move between two or more positions or configurations. For example, in one embodiment, the generally horizontal and vertical portions 22 and 24 are joined by a hinge or pivot, which allows the case 20 to be moved between a functional or open configuration (Fig.

[0049] 1) in which the generally vertical portion 24 is oriented at approximately 90 degrees to the generally horizontal portion 22 and a transport or closed configuration in which the generally vertical portion 24 is rotated about the hinge to approach the generally horizontal portion 22. In such a reconfigurable embodiment, the generally vertical portion 24 may be considered to be the lid of the case 20, while the generally horizontal portion 22 may be considered to be the base. If the case 20 is so reconfigurable, then it may include a latch for releasably locking the case 20 in its closed configuration and / or a handle, which may be grasped for transporting the case 20 in its closed configuration.

[0050] While it may be advantageous for the blood separation device 10 to be embodied in a compact, portable case 20, it is also within the scope of the present disclosure for the blood separation device to be embodied in a larger case or fixture that is intended to be installed in a single location and remain in that location for an extended period of time. If the blood separation device is provided as a fixture, it may be provided with more components and functionality than a more portable version.

[0051] A. Centrifugal Separator

[0052] The centrifugal separator 16 includes a centrifuge compartment 32 that may receive the other components of the centrifugal separator 16 (Fig. 3). The centrifuge compartment 32 may include a lid 34 that is opened to insert and remove a centrifugal separation chamber 36 of the fluid flow circuit 12. During a separation procedure, the lid 34 may be closed with the centrifugal separation chamber 36 positioned within the centrifuge compartment 32, as the centrifugal separation chamber 36 is spun or rotated about an axis 38 under the power of an electric drive motor or rotor 40 of the centrifugal separator 16.

[0053] The particular configuration and operation of the centrifugal separator 16 depends upon the particular configuration of the centrifugal separation chamber 36 of the fluid flow circuit 12. In one embodiment, the centrifugal separator 16 is similar in structure and operation to that of the ALYX® system manufactured by Fenwal, Inc. of Lake Zurich, Illinois, which is an affiliate of Fresenius Kabi AG of Bad Homburg, Germany, as described in greater detail in U. S. Patent No.

[0054] 8,075,468, which is incorporated herein by reference. More particularly, the centrifugal separator 16 may include a carriage or support 42 that holds the centrifugal separation chamber 36 and a yoke member 44. The yoke member 44 engages an umbilicus 46 of the fluid flow circuit 12, which extends between the centrifugal separation chamber 36 and a cassette 48 of the fluid flow circuit 12 (Fig. 4). The yoke member 44 causes the umbilicus 46 to orbit around the centrifugal separation chamber 36 at a one omega rotational speed. The umbilicus 46 twists about its own axis as it orbits around the centrifugal separation chamber 36. The twisting of the umbilicus 46 about its axis as it rotates at one omega with the yoke member 44 imparts a two omega rotation to the centrifugal separation chamber 36, according to known design. The relative rotation of the yoke member 44 at a one omega rotational speed and the centrifugal separation chamber 36 at a two omega rotational speed keeps the umbilicus 46 untwisted, avoiding the need for rotating seals.

[0055] Blood is introduced into the centrifugal separation chamber 36 by the umbilicus 46, with the blood being separated (e.g., into a layer of less dense components, such as platelet-rich plasma, and a layer of more dense components, such as packed red blood cells) within the centrifugal separation chamber 36 as a result of centrifugal forces as it rotates. Components of an Interface monitoring assembly 26 (which will be described in greater detail herein) may be positioned within the centrifuge compartment 32 to oversee separation of blood within the centrifugal separation chamber 36. As shown in Figs. 5-7, the interface monitoring assembly 26 may include a light source 28 and a light detector 30, which is positioned and oriented to receive at least a portion of the light emitted by the light source 28. The light source 28 emits a light beam through the separated blood components within the centrifugal separation chamber 36 (which may be formed of a material that substantially transmits the light or at least a particular wavelength of the light without absorbing it). A portion of the light reaches the light detector 30, which transmits a signal to the controller 18 that is indicative of the location of an interface between the separated blood components. If the controller 18 determines that the interface is not at a target location (which can affect the separation efficiency of the centrifugal separator 16 and / or the quality of the separated blood components), then it can issue commands to the appropriate components of the blood separation device 10 to modify their operation so as to move the interface to the proper location.

[0056] B. Other Components Of The Blood Separation Device In addition to the spinning membrane separator drive unit 14 and the centrifugal separator 16, the blood separation device 10 may include other components compactly arranged to aid blood processing.

[0057] The generally horizontal portion 22 of the case 20 of the illustrated blood separation device 10 includes a cassette station 54, which accommodates a cassette 48 of the fluid flow circuit 12 (Fig. 4). In one embodiment, the cassette station 54 is similarly configured to the cassette station of U. S. Patent No.

[0058] 5,868,696 (which is incorporated herein by reference), but is adapted to include additional components and functionality. The illustrated cassette station 54 includes a plurality of clamps or valves V1 -V9 (Fig. 1 ), which move between a plurality of positions (e.g., between a retracted or lowered position and an actuated or raised position) to selectively contact or otherwise interact with corresponding valve stations C1 -C9 of the cassette 48 of the fluid flow circuit 12 (Fig. 4). Depending on the configuration of the fluid flow circuit 12, its cassette 48 may not include a valve station C1-C9 for each valve V1-V9 of the cassette station 54, in which case fewer than all of the valves V1 -V9 will be used in a separation procedure.

[0059] In the actuated position, a valve V1 -V9 engages the associated valve station C1 -C9 to prevent fluid flow through that valve station C1 -C9 (e.g., by closing one or more ports associated with the valve station C1-C9, thereby preventing fluid flow through that port or ports). In the retracted position, a valve V1-V9 is disengaged from the associated valve station C1-C9 (or less forcefully contacts the associated valve station C1-C9 than when in the actuated position) to allow fluid flow through that valve station C1-C9 (e.g., by opening one or more ports associated with the valve station C1-C9, thereby allowing fluid flow through that port or ports). Additional clamps or valves V10 and V11 may be positioned outside of the cassette station 54 to interact with portions or valve stations C10 and C11 (which may be lengths of tubing) of the fluid flow circuit 12 to selectively allow and prevent fluid flow therethrough. The valves V1 -V9 and corresponding valve stations C1-C9 of the cassette station 54 and cassette 48 may be differently configured and operate differently from the valves V10 and V11 and valve stations C10 and C11 that are spaced away from the cassette station 54.

[0060] The cassette station 54 may be provided with additional components, such as pressure sensors A1-A4, which interact with sensor stations S1-S4 of the cassette 48 to monitor the pressure at various locations of the fluid flow circuit 12. For example, if the blood source is a human patient, one or more of the pressure sensors A1 -A4 may be configured to monitor the pressure of the patient’s vein during blood draw and return. Other pressure sensors A1 -A4 may monitor the pressure of a spinning membrane separator received by the spinning membrane separator drive unit 14 and a centrifugal separation chamber 36 received by the centrifugal separator 16. The controller 18 may receive signals from the pressure sensors A1 -A4 that are indicative of the pressure within the fluid flow circuit 12 and, if a signal indicates a low- or high-pressure condition, the controller 18 may initiate an alarm or error condition to alert an operator to the condition and / or to attempt to bring the pressure to an acceptable level without operator intervention. The blood separation device 10 may also include a plurality of pumps P1-P6 (which may be collectively referred to as a pump assembly or pump system) to cause fluid to flow through the fluid flow circuit 12. The pumps P1 -P6 may be differently or simliarly configured and / or function similarly or differently from each other. In the illustrated embodiment the pumps P1 -P6 are configured as peristaltic pumps, which may be generally configured as described in U. S. Patent No. 5,868,696. Each pump P1 -P6 engages a different tubing loop T1-T6 extending from a side surface of the cassette 48 (Fig. 4) and may be selectively operated under command of the controller 18 to cause fluid to flow through a portion of the fluid flow circuit 12. In one embodiment, all or a portion of the cassette station 54 may be capable of translational motion in and out of the case 20 to allow for automatic loading of the tubing loops T1-T6 into the associated pump P1-P6.

[0061] The illustrated blood separation device 10 also includes a centrifugal separator sensor M1 for determining one or more properties of fluids flowing out of and / or into the centrifugal separator 16. If the fluid flowing out of the centrifugal separator 16 includes red blood cells, the centrifugal separator sensor M1 may be configured to determine the hematocrit of the fluid, if the fluid flowing out of the centrifugal separator 16 is platelet-rich plasma, the centrifugal separator sensor M1 may be configured to determine the platelet concentration of the platelet-rich plasma. The centrifugal separator sensor M1 may detect the one or more properties of a fluid by optically monitoring the fluid as it flows through tubing of the fluid flow circuit 12 or by any other suitable approach. The controller 18 may receive signals from the centrifugal separator sensor M1 that are indicative of the one or more properties of fluid flowing out of the centrifugal separator 16 and use the signals to optimize the separation procedure based upon that property or properties. If the property or properties is / are outside of an acceptable range, then the controller 18 may initiate an alarm or error condition to alert an operator to the condition. A suitable device and method for monitoring hematocrit and / or platelet concentration is described in U. S. Patent No. 6,419,822 (which is incorporated herein by reference), but it should be understood that a different approach may also be employed for monitoring hematocrit and / or platelet concentration of fluid flowing out of the centrifugal separator 16. The illustrated blood separation device 10 further includes a spinner outlet sensor M2, which accommodates tubing of a fluid flow circuit that flows a separated blood component out of a spinning membrane separator received by the spinning membrane separator drive unit 14. The fluid flow circuit 1 of Fig. 2 does not employ a spinning membrane separator, so the spinner outlet sensor M2 is not described in detail herein.

[0062] The illustrated blood separation device 10 also includes an air detector M3 (e.g., an ultrasonic bubble detector), which accommodates tubing of the fluid flow circuit 12 that flows fluid to a recipient. It may be advantageous to prevent air from reaching the recipient, so the air detector M3 may transmit signals to the controller 18 that are indicative of the presence or absence of air in the tubing. If the signal is indicative of air being present in the tubing, the controller 18 may initiate an alarm or error condition to alert an operator to the condition and / or to take corrective action to prevent the air from reaching the recipient (e.g., by reversing the flow of fluid through the tubing or diverting flow to a vent location).

[0063] The generally vertical portion 24 of the case 20 may include a plurality of weight scales W1-W6 (six are shown, but more or fewer may be provided), each of which may support one or more fluid containers F1 -F5 of the fluid flow circuit 12 (Fig. 2). The containers F1-F5 receive fluid components or waste products separated during processing or intravenous fluids or additive fluids. Each weight scale W1-W6 transmits to the controller 18 a signal that is indicative of the weight of the fluid within the associated container F1 -F5 to track the change of weight during the course of a procedure. This allows the controller 18 to process the incremental weight changes to derive fluid processing volumes and flow rates and subsequently generate signals to control processing events based, at least in part, upon the derived processing volumes. For example, the controller 18 may diagnose leaks and obstructions in the fluid flow circuit 12 and alert an operator.

[0064] The illustrated case 20 is also provided with a plurality of hooks or supports H1 and H2 that may support various components of the fluid flow circuit 12 or other suitably sized and configured objects.

[0065] C. Controller

[0066] According to an aspect of the present disclosure, the blood separation device 10 includes a controller 18, which is suitably configured and / or programmed to control operation of the blood separation device 10. In one embodiment, the controller 18 comprises a main processing unit (MPU), which can comprise, e.g., a Pentium™ type microprocessor made by Intel Corporation, although other types of conventional microprocessors can be used. In one embodiment, the controller 18 may be mounted inside the generally vertical portion 24 of the case 20, adjacent to or incorporated into an operator interface station (e.g., a touchscreen). In other embodiments, the controller 18 and operator interface station may be associated with the generally horizontal portion 22 or may be incorporated into a separate device that is connected (either physically, by a cable or the like, or wirelessly) to the blood separation device 10.

[0067] The controller 18 is configured and / or programmed to execute at least one blood processing application but, more advantageously, is configured and / or programmed to execute a variety of different blood processing applications. For example, the controller 18 may be configured and / or programmed to carry out one or more of the following: a double unit red blood cell collection procedure, a plasma collection procedure, a plasma / red blood cell collection procedure, a red blood cell / plateiet / plasma collection procedure, a platelet collection procedure, a platelet / plasma collection procedure, and a mononuclear cell collection procedure. Additional or alternative procedure applications can be included without departing from the scope of the present disclosure.

[0068] More particularly, in carrying out any one of these blood processing applications, the controller 18 is configured and / or programmed to control one or more of the following tasks: drawing blood into a fluid flow circuit 12 mounted to the blood separation device 10, conveying blood through the fluid flow circuit 12 to a location for separation (i.e., into a spinning membrane separator or centrifugal separation chamber 36 of the fluid flow circuit 12), separating the blood into two or more components as desired, and conveying the separated components into a storage container or to a recipient (which may be the source from which the fluid was originally drawn).

[0069] This may include instructing the spinning membrane separator drive unit 14 or the centrifugal separator 16 to operate at a particular rotational speed and instructing a pump P1-P6 to convey fluid through a portion of the fluid flow circuit 12 at a particular flow rate. Hence, while it may be described herein that a particular component of the blood separation device 10 (e.g., the spinning membrane separator drive unit 14 or the centrifugal separator 16) performs a particular function, it should be understood that that component is being controlled by the controller 18 to perform that function.

[0070] Before, during, and after a procedure, the controller 18 may receive signals from various components of the blood separation device 10 (e.g., the pressure sensors A1 -A4) to monitor various aspects of the operation of the blood separation device 10 and characteristics of the blood and separated blood components as they flow through the fluid flow circuit 12. If the operation of any of the components and / or one or more characteristics of the blood or separated blood components is outside of an acceptable range, then the controller 18 may initiate an alarm or error condition to alert the operator and / or take action to attempt to correct the condition. The appropriate corrective action will depend upon the particular error condition and may include action that is carried out with or without the involvement of an operator.

[0071] For example, the controller 18 may include an interface control module, which receives signals from the light detector 30 of the interface monitoring assembly 26. The signals that the controller 18 receives from the light detector 30 are indicative of the location of an interface between the separated blood components within the centrifugal separation chamber 36. If the controller 18 determines that the interface is not at a target location, then it can issue commands to the appropriate components of the blood separation device 10 to modify their operation so as to move the interface to the proper location. For example, the controller 18 may instruct one of the pumps P1 -P6 to cause blood to flow into the centrifugal separation chamber 36 at a different rate and / or for a separated blood component to be removed from the centrifugal separation chamber 36 at a different rate and / or for the centrifugal separation chamber 36 to be spun at a different speed by the centrifugal separator 16. A particular protocol carried out by the interface control module in adjusting the position of the interface within the centrifugal separation chamber 36 will be described in greater detail.

[0072] If provided, an operator interface station associated with the controller 18 allows the operator to view on a screen or display (in alpha-numeric format and / or as graphical images) information regarding the operation of the system. The operator interface station also allows the operator to select applications to be executed by the controller 18, as well as to change certain functions and performance criteria of the system. If configured as a touchscreen, the screen of the operator interface station can receive input from an operator via touchactivation. Otherwise, if the screen is not a touchscreen, then the operator interface station may receive input from an operator via a separate input device, such as a computer mouse or keyboard. It is also within the scope of the present disclosure for the operator interface station to receive input from both a touchscreen and a separate input device, such as a keypad.

[0073] II. The Disposable Fluid Flow Circuit

[0074] A. Overview

[0075] As for the fluid flow circuit or flow set 12 (Fig. 2), it is intended to be a sterile, single use, disposable item. Before beginning a given blood separation procedure, the operator loads various components of the fluid flow circuit 12 in the case 20 in association with the blood separation device 10. The controller 18 implements the procedure based upon preset protocols, taking into account other input from the operator. Upon completing the procedure, the operator removes the fluid flow circuit 12 from association with the blood separation device 10. The portions of the fluid flow circuit 1 holding the collected blood component or components (e.g., collection containers or bags) are removed from the case 20 and retained for storage, transfusion, or further processing. The remainder of the fluid flow circuit 12 is removed from the case 20 and discarded.

[0076] A variety of different disposable fluid flow circuits may be used in combination with the biood separation device 10, with the appropriate fluid flow circuit depending on the separation procedure to be carried out using the system. Generally speaking, though, the fluid flow circuit 12 includes a cassette 48 (Fig.

[0077] 4), to which the other components of the fluid flow circuit 12 are connected by flexible tubing. The other components may include a plurality of fluid containers F1-F5 (for holding blood, a separated blood component, an intravenous fluid, or an additive solution, for example), one or more fluid source access devices (e.g., a connector for accessing blood within a fluid container), and a spinning membrane separator (not illustrated) and / or a centrifugal separation chamber 36 (Figs. 8 and 9).

[0078] B. Cassette And Tubing

[0079] The cassette 48 (Fig. 4) provides a centralized, programmable, integrated platform for all the pumping and many of the valving functions required for a given blood processing procedure. In one embodiment, the cassette 48 is similarly configured to the cassette of U. S. Patent No. 5,868,696, but is adapted to include additional components (e.g., more tubing loops T1-T6) and functionality.

[0080] In use, the cassette 48 is mounted to the cassette station 54 of the blood separation device 10, with a flexible diaphragm of the cassette 48 placed into contact with the cassette station 54. The flexible diaphragm overlays an array of interior cavities formed by the body of the cassette 48. The different interior cavities define sensor stations S1-S4, valve stations C1-C9, and a plurality of flow paths. The side of the cassette 48 opposite the flexible diaphragm may be sealed by another flexible diaphragm or a rigid cover, thereby sealing fluid flow through the cassette 48 from the outside environment.

[0081] Each sensor station S1 -S4 is aligned with an associated pressure sensor A1-A4 of the cassette station 54, with each pressure sensor A1-A4 capable of monitoring the pressure within the associated sensor station S1 -S4. Each valve station C1-C9 is aligned with an associated valve V1-V9, and may define one or more ports that allow fluid communication between the valve station C1-C9 and another interior cavity of the cassette 48 (e.g., a flow path). As described above, each valve V1-V9 is movable under command of the controller 18 to move between a plurality of positions (e.g., between a retracted or lowered position and an actuated or raised position) to selectively contact the valve stations C1-C9 of the cassette 48. In the actuated position, a valve V1 -V9 engages the associated valve station C1-C9 to close one or more of its ports to prevent fluid flow therethrough. In the retracted position, a valve V1-V9 is disengaged from the associated valve station C1-C9 (or less forcefully contacts the associated valve station C1-C9 than when in the actuated position) to open one or more ports associated with the valve station C1-C9, thereby allowing fluid flow therethrough.

[0082] As described, a plurality of tubing loops T 1 -T6 extend from the side surface of the cassette 48 to interact with pumps P1-P6 of the blood separation device 10. In the illustrated embodiment, six tubing loops T1 -T6 extend from the cassette 48 to be received by a different one of six pumps P1 -P6, but in other embodiments, a procedure may not require use of all of the pumps P1-P6, in which case the cassette 48 may include fewer than six tubing loops. The different pumps P1-P6 may interact with the tubing loops T 1 -T6 of the cassette 48 to perform different tasks during a separation procedure. Certain procedures require fewer than all of the sensor stations, valve stations, and / or tubing loops illustrated in the exemplary cassette 48 of Fig. 4, such that it should be understood that the cassettes of different fluid flow circuits 12 may be differently configured (e.g., with fewer sensor stations, valve stations, and / or tubing loops) without departing from the scope of the present disclosure.

[0083] Additional tubing extends from the side surface of the cassette 48 to connect to the other components of the fluid flow circuit 12, such as the various fluid containers F1-F5, a spinning membrane separator (if provided), and the centrifugal separation chamber 36. The tubing connected to the centrifugal separation chamber 36 (which includes one inlet tube and two outlet tubes) may be aggregated into an umbilicus 46 (Fig. 3) that is engaged by the yoke member 44 of the centrifugal separator 16 (as described above) to cause the umbilicus 46 to orbit around and spin or rotate the centrifugal separation chamber 36 during a separation procedure.

[0084] Various additional components may be incorporated into the tubing leading out of the cassette 48 or into one of the cavities of the cassette 48. For example, as shown in Fig. 2, a manual clamp 56 may be associated with a line or lines leading to the blood / fluid source and / or fluid recipient, a return line filter 58 (e.g., a microaggregate filter) may be associated with a line leading to a fluid recipient, filters may be positioned upstream of one or more of the fluid containers to remove a substance (e.g., leukocytes) from a separated component (e.g., red blood cells or platelets) flowing into the fluid container, and / or an air trap 62 may be positioned on a line upstream of the centrifugal separation chamber 36.

[0085] C. Centrifugal Separation Chamber

[0086] The fluid flow circuit 1 is provided with a centrifugal separation chamber 36, with Figs. 8 and 9 illustrating an exemplary centrifugal separation chamber 36. In the illustrated embodiment, the body of the centrifugal separation chamber 36 is pre-formed in a desired shape and configuration (e.g., by injection molding) from a rigid, biocompatible plastic material, such as a non-plasticized medical grade acrylonitrile-butadiene-styrene (ABS). All contours, ports, channels, and walls that affect the fluid separation process are preformed in a single, injection molded operation. Alternatively, the centrifugal separation chamber 36 can be formed by separate molded parts, either by nesting cup-shaped subassemblies or two symmetric halves.

[0087] The underside of the centrifugal separation chamber 36 includes a shaped receptacle 86 (Fig. 8) that is suitable for receiving an end of the umbilicus 46 of the fluid flow circuit 12 (Fig. 3). A suitable receptacle 86 and the manner in which the umbilicus 46 may cooperate with the receptacle 86 to deliver fluid to and remove fluid from the centrifugal separation chamber 36 are described in greater detail in U. S. Patent No. 8,075,468.

[0088] The illustrated centrifugal separation chamber 36 has radially spaced apart inner (low-g) and outer (high-g) side wall portions 88 and 90, a bottom or first end wail portion 92, and a cover or second end wall portion 93. The cover 93 comprises a simple flat part that can be easily welded or otherwise secured to the body of the centrifugal separation chamber 36. Because all features that affect the separation process are incorporated into one injection molded component, any tolerance differences between the cover 93 and the body of the centrifugal separation chamber 36 will not affect the separation efficiencies of the centrifugal separation chamber 36. The wall portions 88 and 90, the bottom 92, and the cover 93 together define an enclosed, generally annular channel 94.

[0089] The (whole blood) inlet 96 communicating with the channel 94 is defined between opposing interior radial walls 98 and 100. One of the interior walls 98 joins the outer (high-g) wall portion 90 and separates the upstream and downstream ends of the channel 94. The interior walls 98 and 100 define the inlet passageway 96 of the centrifugal separation chamber 36 which, in one flow configuration, allows fluid to flow from the umbilicus 46 to the upstream end of the channel 94.

[0090] The illustrated centrifugal separation chamber 36 further includes first and second outlets 102 and 104, respectively, which may be defined by opposing surfaces of interior radial walls. Both the first and second outlets 102 and 104 extend radially inward from the channel 94. The first (plasma) outlet 102 extends radially inward from an opening which, in the illustrated embodiment, is located at the inner side wall portion 88, while the second (red blood cell) outlet 104 extends radially inward from an opening that is associated with the outer side wall portion 90. The illustrated first outlet 102 is positioned adjacent to the inlet 96 (near the upstream end of the channel 94), while the second outlet 104 may be positioned at the opposite, downstream end of the channel 94.

[0091] It should be understood that the centrifugal separation chamber 36 illustrated in Figs. 8 and 9 is merely exemplary and that the centrifugal separation chamber 36 may be differently configured without departing from the scope of the present disclosure.

[0092] 1. Centrifugal Separation And Interface Detection Principles Blood flowed into the channel 94 separates into an optically dense layer RBC and a less optically dense layer PLS (Figs. 10-12) as the centrifugal separation chamber 36 is rotated about the rotational axis 38. The optically dense layer RBC forms as larger and / or heavier blood particles move under the influence of centrifugal force toward the outer (high-g) wall portion 90. The optically dense layer RBC will typically include red blood cells (and, hence, may be referred to herein as the “RBC layer”) but, depending on the speed at which the centrifugal separation chamber 36 is rotated, other cellular components (e.g., larger white blood cells) may also be present in the optically dense layer RBC.

[0093] The less optically dense layer PLS typically includes a plasma constituent, such as platelet-rich plasma (and, hence, may be referred to herein as the “PLS layer”). Depending on the speed at which the centrifugal separation chamber 36 is rotated and the length of time that the blood is resident therein, other components (e.g., smaller white blood cells and anticoagulant) may also be present in the less optically dense layer PLS.

[0094] In one embodiment, blood Introduced into the channel 94 via the inlet 96 will travel in a generally clockwise direction (in the orientation of Fig. 8) as the optically dense layer RBC separates from the less optically dense layer PLS. The optically dense layer RBC continues moving in the clockwise direction as it travels the length of the channel 94 along the outer side wall portion 90, from the upstream end to the downstream end, where it exits the channel 94 via the second outlet 104. The less optically dense layer PLS separated from the optically dense layer RBC reverses direction, moving counterclockwise along the inner side wall portion 88 to the first outlet 102, adjacent to the inlet 96. The inner side wall portion 88 may be tapered inward as it approaches the second outlet 104 to force the plasma liberated at or adjacent to the downstream end of the channel 94 to drag the interface back towards the upstream end of the channel 94, where the lower surface hematocrit will re-suspend any platelets settled on the interface.

[0095] As described above, the transition between the optically dense layer RBC and the less optically dense layer PLS may be referred to as the interface INT. The location of the interface INT within the channel 94 of the centrifugal separation chamber 36 can dynamically shift during blood processing, as Figs. 10-12 show. If the location of the interface INT is too high (that is, if it is too close to the inner side wall portion 88 and the first outlet 102, as in Fig. 11 ), red blood cells can flow into the first outlet 102, potentially adversely affecting the quality of the low-density components (platelet-rich plasma or platelet-poor plasma). On the other hand, if the location of the interface INT is too low (that is, if it resides too far away from the inner wall portion 88, as Fig. 12 shows), the collection efficiency of the system may be impaired. The ideal or target location of the interface INT may be experimentally determined, which may vary depending on any of a number of factors (e.g., the configuration of the centrifugal separation chamber 36, the rate at which the centrifugal separation chamber 36 is rotated about the rotational axis 38, etc.).

[0096] As described above, the blood separation device 10 may include an interface monitoring assembly 26 and a controller 18 with an interface control module to monitor and, as necessary, correct the location of the interface INT. In one embodiment, the centrifugal separation chamber 36 is formed with a ramp 106 extending from the high-g wall portion 90 at an angle a across at least a portion of the channel 94 (Figs. 8-12). The angle a, measured with respect to the rotational axis 38 is about 25° in one embodiment. Figs. 10-12 show the orientation of the ramp 106 when viewed from the low-g side wall portion 88 of the centrifugal separation chamber 36. Although it describes a flexible separation chamber, the general structure and function of the ramp 106 may be better understood with reference to U. S. Patent No. 5,632,893, which is incorporated herein by reference. The ramp 106 may be positioned at any of a number of locations between the upstream and downstream ends of the channel 94, but in one embodiment, the ramp 106 may be positioned generally adjacent to the first outlet 102, in the path of fluid and / or a fluid component moving from the inlet 96 to the first outlet 102.

[0097] The ramp 106 makes the interface INT between the optically dense layer RBC and the less optically dense layer PLS more discernible for detection, displaying the optically dense layer RBC, less optically dense layer PLS, and interface INT for viewing through a light-transmissive portion of the centrifugal separation chamber 36. To that end, the ramp 106 and at least the portion of the centrifugal separation chamber 36 angularly aligned with the ramp 106 may be formed of a light-transmissive material, although it may be advantageous for the entire centrifugal separation chamber 36 to be formed of the same light- transmissive material.

[0098] In the illustrated embodiment, the light source 28 of the interface monitoring assembly 26 is secured to a fixture or wall of the centrifuge compartment 32 and oriented to emit a light that is directed toward the rotational axis 38 of the centrifugal separator 16, as shown in Figs. 5-7. If the light detector 30 is positioned at an angle with respect to the light source 28 (as in the illustrated embodiment), the light L emitted by the light source 28 must be redirected from its initial path before it will reach the light detector 30. In the illustrated embodiment, the light L is redirected by a reflector that is associated with a light-transmissive portion of the inner side wall portion 88, as shown in Figs. 5-7. The reflector may be a separate piece that is secured to the inner side wall portion 88 (e.g., by being bonded thereto) or may be integrally formed with the body of the centrifugal separation chamber 36.

[0099] In one embodiment, the reflector may be a reflective surface, such as a mirror, that is oriented (e.g., at a 45° angle) to direct light L emitted by the light source 28 to the light detector 30. In another embodiment, the reflector is provided as a prismatic reflector 108 (Figs. 13 and 14), which is formed of a light-transmissive material (e.g., a clear plastic material) and has inner and outer walls 110 and 112 and first and second end walls 114 and 116. The inner wall 110 is positioned against the inner side wall portion 88 of the centrifugal separation chamber 36 and is oriented substantially perpendicular to the initial path of the light L from the light source 28. This allows light L from the light source 28 to enter into the prismatic reflector 108 via the inner wall 110 while continuing along its initial path. The light L continues through the prismatic reflector 108 along its initial path until it encounters the first end wall 114. The first end wall 114 is oriented at an angle (e.g., an approximately 45° angle) with respect to the first surface 110 and the second end wall 116, causing the light to be redirected within the prismatic reflector 108, rather than exiting the prismatic reflector 108 via the first end wall 114.

[0100] The first end wall 114 directs the light L at an angle to its initial path (which may be an approximately 90° angle, directing it from a path toward the rotational axis 38 to a path that is generally parallel to the rotational axis 38) toward the second end wall 116 (Fig. 14). The first end wall 114 and the inner and outer walls 110 and 112 of the prismatic reflector 108 may be configured to transmit the redirected light L from the first end wall 114 to the second end wall 116 by total internal reflection. The second end wall 116 is oriented substantially perpendicular to the redirected path of the light L through the prismatic reflector 108, such that the light L will exit the prismatic reflector 108 via the second end wall 116, continuing along its redirected path. In one embodiment, the second end wall 116 is roughened or textured or otherwise treated or conditioned to diffuse the light L as it exits the prismatic reflector 108, which may better ensure that the light L reaches the light detector 30 (Fig. 7).

[0101] The prismatic reflector 108 may be angularly aligned with the ramp 106, such that the light L from the light source 28 will only enter into the prismatic reflector 108 when the ramp 106 has been rotated into the path of the light L. At all other times (when the ramp 106 is not in the path of the light L), the light L will not reach the prismatic reflector 108 and, thus, will not reach the light detector 30. This is illustrated in Figs. 15-18, which show the ramp 106 and prismatic reflector 108 as the centrifugal separation chamber 36 is rotated about the rotational axis 38 (while the light source 28 remains in a fixed location). In Fig. 15, the ramp 106 and prismatic reflector 108 have not yet been rotated into the initial path of the light L from the light source 28. At this time, no light is transmitted to the light detector 30, such that the output voltage of the light detector 30 (i.e., the signal transmitted from the light detector 30 to the controller 18) is in a low- or zero-state (Fig. 19).

[0102] Upon the ramp 106 first being rotated into the initial path of the light L from the light source 28 (Fig. 16), the light L will begin to reach the prismatic reflector 108, which directs the light L to the light detector 30. This causes the voltage output of the light detector 30 (i.e., the signal transmitted from the light detector 30 to the controller 18) to increase to a non-zero value or state, as shown in Fig. 20.

[0103] During a calibration phase, the channel 94 is filled with a fluid that will transmit the light L rather than absorbing or reflecting the light or otherwise preventing the light L from reaching the prismatic reflector 108, such that the voltage output of the light detector 30 will remain generally constant as the ramp 106 and prismatic reflector 108 are rotated through the initial path of the light L from the light source 28 (Figs. 17 and 21). Such a calibration phase may coincide with a priming phase during which saline is pumped through the fluid flow circuit 12 to prime the fluid flow circuit 12 or may comprise a separate phase. A calibration phase may be useful in ensuring the proper operation of the light source 28 and the light detector 30, standardizing the readings taken during a separation procedure in case of any irregularities or imperfections of the centrifugal separation chamber 36, and establishing a baseline value for the signal transmitted from the light detector 30 to the controller 18 when the ramp 106 and prismatic reflector 108 are aligned with the light source 28. As will be described in greater detail, the voltage output of the light detector 30 will typically not remain constant as the ramp 106 and prismatic reflector 108 are rotated through the initial path of the light L from the light source 28 because the different fluid layers displayed on the ramp 106 will allow different amounts of light L to reach the prismatic reflector 108.

[0104] The ramp 106 and prismatic reflector 108 are eventually rotated out of alignment with the light source 28 (Fig. 18), at which time no light L will reach the prismatic reflector 108 and the voltage output of the light detector 30 will return to its low- or zero-state (Fig. 22).

[0105] 2. Detection Of Interface Position

[0106] During separation of blood within the channel 94, the light L from the light source 28 travels through a light-transmissive portion of the outer side wall portion 90 and the ramp 106 to intersect the separated blood components thereon when the ramp 106 has been rotated into the initial path of the light L. After passing through the ramp 106, the light continues through the channel 94 and the fluids in the channel 94. At least a portion of the light L (i.e., the portion not absorbed or reflected by the fluids) exits the channel 94 by striking and entering a light- transmissive portion of the inner side wall portion 88. The light L passes through the inner side wall portion 88 and enters the prismatic reflector 108, which redirects the light L from its initial path to the light detector 30, as described above.

[0107] The light detector 30 generates a signal that is transmitted to the interface control module of the controller 18, which can determine the location of the interface INT on the ramp 106. In one embodiment, the location of the interface INT is associated with a change in the amount of light L that is transmitted through the less optically dense layer PLS and the optically dense layer RBC. For example, the light source 28 may be configured to emit a light L that is more readily transmitted by platelet-rich plasma or platelet-poor plasma than by red blood cells, such as red visible light (from a laser or a differently configured light source 28), which is substantially absorbed by red blood cells. The less optically dense layer PLS and the optically dense layer RBC each occupy a certain portion of the ramp 106, with the light detector 30 receiving different amounts of light L depending on whether the light L travels through the less optically dense layer PLS on the ramp 106 or the optically dense layer RBC on the ramp 106. The percentage of the ramp 106 occupied by each layer is related to the location of the interface INT in the channel 94. Thus, by measuring the amount of time that the voltage output or signal from the light detector 30 is relatively high (corresponding to the time during which the light L is passing through only the less optically dense layer PLS on the ramp 106), the controller 18 may determine the location of the interface INT and take steps to correct the location of the interface INT, if necessary.

[0108] Figs. 23-26 show a portion of the ramp 106 being rotated into and through the initial path of the light L from the light source 28. Four specific events are shown: just before the ramp 106 is rotated into the path of the light L (Fig. 23), the ramp 106 first being rotated into the path of the light L (Fig. 24), just before the interface INT displayed on the ramp 106 is rotated into the path of the light L (Fig.

[0109] 25), and just after the interface INT is rotated into the path of the light L (Fig. 26). Figs. 27-30 respectively illustrate the voltage output of the light detector 30 (corresponding to the signal that it transmits to the controller 18) during each of these events.

[0110] As described above, the light detector 30 will receive no light L from the light source 28 when the prismatic reflector 108 is out of alignment with the initial path of the light L from the light source 28, as shown in Fig. 23. Fig. 27 shows that the output voltage of the light detector 30 (i.e., the signal transmitted from the light detector 30 to the controller 18) at this time is in a low- or zero-state.

[0111] When the ramp 106 is first rotated into the path of light L from the light source 28 (Fig. 24), the light detector 30 may begin receiving light L. The amount of light L received by the light detector 30 depends upon the fluid on the ramp 106 encountered by the light L (i.e., the fluid in the channel 94 between the ramp 106 and the inner side wall portion 88 that the light L must traverse before being directed to the light detector 30). As described above, the less optically dense layer PLS occupies a certain percentage of the channel 94 adjacent to the inner side wall portion 88, while the optically dense layer RBC occupies a certain percentage of the channel 94 adjacent to the outer side wall portion 90 (with the interface INT positioned at the transition between the two separated blood component layers). The illustrated ramp 106 is closest to the inner side wall portion 88 at its left end (in the orientation of Figs. 23-26), while being farther spaced from the inner side wall portion 88 at its right end. At and adjacent to its left end, the ramp 106 will display only the fluid positioned closest to the inner side wall portion 88 (i.e., the less optically dense layer PLS), while the ramp 106 will display only the fluid positioned closest to the outer side wall portion 90 (i.e., the optically dense layer RBC) at and adjacent to its right end, as shown in Figs. 23-

[0112]

[0113] At some point between its ends, the angled ramp 106 will be at a radial position where it will display the transition between the less optically dense layer PLS and the optically dense layer RBC (i.e., the interface INT). Hence, the location of the interface INT on the ramp 106 is dependent upon the percentage of the width of the ramp 106 that displays the less optically dense layer PLS (which is indicative of the percentage of the channel 94 occupied by the less optically dense layer PLS) and the percentage of the width of the ramp 106 that displays the optically dense layer RBC (which is indicative of the percentage of the channel 94 occupied by the optically dense layer RBC). It should be understood that the percentage of the ramp 106 occupied by the less optically dense layer PLS and by the optically dense layer RBC is not necessarily equal to the percentage of the channel 94 occupied by the less optically dense layer PLS and by the optically dense layer RBC, but that the percentage of the ramp 106 occupied by a separated blood component layer may be merely indicative of the percentage of the channel 94 occupied by that separated blood component layer.

[0114] In such an embodiment, as the ramp 106 is rotated into the path of the light L from the light source 28, the light L will first encounter the portion of the ramp 106 that is positioned closest to the inner side wall portion 88 (i.e., the section of the ramp 106 that most restricts the channel 94), as shown in Fig. 24. As described above, the less optically dense layer PLS will be positioned adjacent to the inner side wall portion 88 as it separates from the optically dense layer RBC, such that the fluid displayed on this radially innermost section of the ramp 106 (i.e., the fluid present in the channel 94 between the ramp 106 and the inner side wall portion 88) will be the less optically dense layer PLS. The light is substantially transmitted through the less optically dense layer PLS to the inner side wail portion 88, and through the light-transmissive inner side wall portion 88 to the prismatic reflector 108, which redirects the light L to the light detector 30. This causes the voltage output of the light detector 30 (i.e., the signal transmitted from the light detector 30 to the controller 18) to increase to a non-zero value or state, as shown in Fig. 28. Depending on the nature of the light L, the amount of light L received by the light detector 30 (and, hence, the magnitude of the voltage output) after the light L has passed through the less optically dense layer PLS may be greater than, less than, or equal to the amount of light L received by the light detector 30 after passing through saline during the calibration phase described above.

[0115] Further rotation of the ramp 106 through the path of light L from the light source 28 exposes the light L to portions of the ramp 106 that are increasingly spaced from the inner side wall portion 88 (i.e., the light L travels through portions of the channel 94 that are less restricted by the ramp 106 as the ramp 106 is rotated through the path of the light L). Up until the time that the interface INT on the ramp 106 is rotated into the path of the light L (as shown in Fig. 25), the only fluid in the channel 94 that the light L will have passed through will be the less optically dense layer PLS, such that a generally uniform level of light reaches the light detector 30 between the conditions shown in Figs. 24 and 25. Accordingly, the voltage output of the light detector 30 will be generally uniform (at an elevated level) the whole time that the ramp 106 passes through the path of the light L before being exposed to the interface INT, as shown in Fig. 29. The controiler 18 may be programmed and / or configured to consider a signal that deviates from a maximum signal level (e.g., a 10% decrease) to be part of the elevated signal for purposes of calculating the pulse width of the signal. The controller 18 will treat a greater deviation (i.e., a greater decrease in the magnitude of the signal) as the end of the elevated signal for purposes of calculating the pulse width of the signal.

[0116] Just after the interface INT has been rotated into the path of light L from the light source 28, the light L will begin to encounter the optically dense layer RBC in the channel 94, as shown in Fig. 26). As described above, the optically dense layer RBC will be positioned adjacent to the outer side wall portion 90 as it separates from the less optically dense layer PLS, such that the optically dense layer RBC will not be displayed on the ramp 106 until the ramp 106 is spaced a greater distance away from the inner side wall portion 88 (i.e., toward the right end of the ramp 106 in the orientation of Figs. 23-26). Less light L is transmitted through the optically dense layer RBC than through the less optically dense layer PLS (which may include all or substantially all of the light L being absorbed by the optically dense layer RBC), such that the amount of light L that reaches the light detector 30 will decrease compared to the amount of light L that reaches the light detector 30 while traveling through only the less optically dense layer PLS in the channel 94 (Figs. 24 and 25). When receiving less light L, the voltage output or signal from the light detector 30 will decrease to a lower level than when the light L was passing through only the less optically dense layer PLS in the channel 94, as shown in Fig.

[0117] 30. When the light L encounters the optically dense layer RBC in the channel 94, the light detector 30 may be generating a signal or voltage output that is approximately equal to its zero-state (as in Fig. 27, when the light detector 30 is receiving no light L) or a signal or voltage output that is some degree less than the magnitude of the signal or voltage output generated while the light L encounters only the less optically dense layer PLS in the channel 94. The controller 18 may be programmed and / or configured to recognize this lower level signal as representing the presence of the optically dense layer RBC on the ramp 106 (and in the portion of the channel 94 being traversed by the light L) and treat this lower level signal as the end point of the elevated signal generated by the light detector 30 while light L passes through only the less optically dense layer PLS in the channel 94.

[0118] Thus, the pulse width of the elevated signal from the light detector 30 to the controller 18 (i.e., the time during which light L is traversing only the less optically dense layer PLS In the channel 94) is determined by the percentages of the ramp 106 that are occupied by the less optically dense layer PLS and the optically dense layer RBC. Accordingly, a greater pulse width of the signal from the light detector 30 to the controller 18 is associated with the less optically dense layer PLS occupying a larger portion of the ramp 106 (as shown in Fig. 31 from the point of view of the light source 28, which may correspond to the condition shown in Fig. 12) and will be indicative of a thinner optically dense layer RBC on the ramp 106 (and in the channel 94). Conversely, a signal from the light detector 30 to the controller 18 having a narrower pulse width is associated with the less optically dense layer PLS occupying a smaller portion of the ramp 106 (as shown in Fig. 32) and will be indicative of a thicker optically dense layer RBC on the ramp 106 (and in the channel 94).

[0119] The controller 18 may compare the pulse width of the signal to the pulse width generated during the calibration phase (described above and shown in Fig.

[0120] 33), which corresponds to the pulse width when light L is transmitted to the light detector 30 over the entire width of the ramp 106. The pulse width of the signal generated by the light detector 30 during the calibration phase may be referred to as the saline calibration signal. Comparing these two pulse widths will indicate the percentage of the ramp 106 that is occupied by the less optically dense layer PLS and by the optically dense layer RBC, which information the controller 18 may use to determine the location of the interface INT within the channel 94. In particular, the interface position may be calculated as follows:

[0121] Interface position (%) = ((saline calibration pulse width - current plasma pulse width) / saline calibration pulse width) * 100 [Equation 1]

[0122] It will be seen that Equation 1 effectively calculates the percentage of the ramp 106 that is occupied by the optically dense layer RBC, as the difference between the two pulse widths corresponds to the length of time that the ramp 106 is rotated through the path of the light L without the light detector 30 received an elevated level of light L (i.e., the amount of time that the ramp 106 is rotated through the path of the light L while the optically dense layer RBC is present on the ramp 106).

[0123] 3. Correction Of Interface Position

[0124] According to a conventional approach, when the location of the interface INT on the ramp 106 has been determined, the interface control module compares the actual interface location with a desired interface location, which may be referred to as the setpoint S. The difference between the setpoint S and the calculated interface position may be referred to as the error signal E, which is shown in Fig. 34.

[0125] The interface control module of the controller 18 inputs the error signal into a PID control algorithm to generate a control signal. The controller 18 uses the control signal to adjust an operational parameter accordingly, such as by adjusting the rate at which the plasma constituent is removed through the first outlet 102 under action of a pump of the blood separation device 10. All other things being equal, increasing the rate at which the plasma constituent is removed through the first outlet 102 will increase the thickness of the optically dense layer RBC, whereas decreasing the rate at which the plasma constituent is removed will decrease the thickness of the optically dense layer RBC. So adjusting the plasma outflow rate causes the interface INT to move toward the desired control position (as Fig. 10 shows), where the error signal is zero. Fig. 35 illustrates this conventional approach, with Fig. 36 illustrating a conventional PID control algorithm.

[0126] Fig. 37 illustrates an interface position control approach according to an aspect of the present disclosure that is implemented by the controller 18, with Fig.

[0127] 38 schematically illustrating an exemplary neural network 118 that communicates with the controller 18 in implementing an interface control routine of the type shown in Fig. 37. As shown in Fig. 37, the controller 18 calculates or determines the current position of the interface (as in the conventional approach shown in Fig.

[0128] 35), which may include the controller 18 determining the position of the interface according to the approach described above or according to any suitable approach. When the controller 18 has determined the current position of the interface, it provides the current interface position to a neural network 118 of the type shown in Fig. 38, along with a plurality of other input parameters. The controller 18 may also provide the neural network 118 with the target interface position, if the neural network 118 does not already have that information and / or if the target interface position has changed and become different from the value possessed by the neural network 118.

[0129] The neural network 118 employs artificial intelligence techniques (including, for example, machine learning and deep learning) to analyze the current position of the interface and at least one of the other input parameters to determine a modification to be made to at least one procedural parameter so as to move the interface from its current position to the target interface position or setpoint S. Once it has determined the appropriate modification(s) to be made to the procedural parameter(s), the neural network 118 transmits a return signal to the controller 18 to inform it of the modification(s) to be implemented. The controller 18 implements the modification(s) provided by the neural network 118 so as to move the interface toward (and, ideally, to) the target interface position and then the process may be repeated, with the controller 18 determining the current position of the interface after the modification(s) to the procedural parameter(s) has / have been made.

[0130] As used herein, the term “procedural parameter” refers to an aspect of the blood separation procedure being implemented by the controller 18 that affects the position of the interface. This may include, for example, the rate at which blood is conveyed into the centrifugal separation chamber 36 within the centrifugal separator 16 by the pump system of the blood separation device 10, the rotational rate at which the centrifugal separator 16 rotates the centrifugal separation chamber 36, and the rate(s) at which one or more of the separated blood components are removed from the centrifugal separation chamber 36 during blood separation. Accordingly, a modification to a procedural parameter may include a change in the rate at which blood is conveyed into the centrifugal separation chamber 36, a change in the rotational rate at which the centrifugal separator 16 rotates the centrifugal separation chamber 36, or a change in the rate at which one or more of the separated blood components are removed from the centrifugal separation chamber 36 during blood separation. Thus, depending on the nature of the modification to be made, the controller 18 may actuate the centrifugal separator and / or the pump system to operate differently in order to implement the modification(s).

[0131] It should be understood that the neural network 118 is not limited to providing the controller 18 with modifications to be made so as to move the interface toward a fixed or unchangeable target interface position. For example, if one or more of the input parameters has changed during the course of a blood separation procedure, it may be appropriate for the target interface position to be changed from an initial position to an adjusted or updated or modified position. In view of this possibility, the neural network 118 may be programmed and enabled to determine whether it is appropriate to modify the target interface position and, upon determining that such a modification is appropriate, provide the controller 18 with a modification to the target interface position itself. As such, it should be understood that the target interface position is one of the procedural parameters that may be modified as the result of an analysis performed by the neural network 118. Typically, changing the target interface position will also require changing one of the other procedural parameters so as to cause the interface to move from its current position to the new target position, but it is also possible that changing the target interface position will not require a modification of any of the other procedural parameters (namely, when the interface is currently located at the new target interface position). As for the term “input parameter,” it is used herein to refer to any aspect of the blood source, the blood itself, or the blood separation procedure that may affect the most appropriate modification(s) to be made to one or more of the procedural parameters so as to move the interface from its current position to the target interface position. The exact input parameters provided to the neural network 118 by the controller 18 may vary without departing from the scope of the present disclosure, but may include (for example): the age and / or sex of a blood source, the composition of the blood (e.g., the hematocrit and / or a platelet precount), the temperature of the blood, and the current phase of the blood separation procedure (e.g., a steady state separation phase or a blood component collection phase or a return phase). These input parameters may be determined in any suitable manner, including by being provided to the controller 18 by an operator (e.g., by entering the age of a blood source or the hematocrit of blood contained in a blood bag) and being provided by a sensor of the blood separation device 10 (e.g., by periodically notifying the controller 18 of the hematocrit of the blood or the rate at which the blood is being pumped into the centrifugal separation chamber 36).

[0132] The neural network 118 may be variously configured and variously operational without departing from the scope of the present disclosure. In the embodiment illustrated in Fig. 38, the neural network 118 includes an input layer 120, an output layer 122, and one or more hidden layers 124 and 126, with each layer defining one or more nodes. The nodes of the input layer 120 are connected to the nodes of at least one of the hidden layers 124, with the nodes of one hidden layer 124 being connected to the nodes of another hidden layer 126 (if provided), and with the nodes of at least one of the hidden layers 126 being connected to the nodes of the output layer 122. The input layer 120 receives the various inputs from the controller 18 (i.e., the current interface position, the other input parameters, and (optionally) the target interface position), processes them (if necessary), and passes them along to the hidden layer(s) 124, 126. The hidden layers 124 and 126 analyze the inputs from the input layer 120, further process them (if necessary), and pass them along to the next layer (which may be either another hidden layer or the output layer 122). The output layer 122 provides the final results of the analysis and processing performed by the preceding layers to the controller 18 (i.e., the modification(s) to be made to one or more of the procedural parameters).

[0133] Each node of the hidden layer(s) 124 and 126 has its own weight, which corresponds to the importance of an associated variable or input to the result that is output by the neural network 118. The weights of one or more of the nodes may be updated after the neural network 118 has analyzed the various inputs from the controller 18. After the neural network has analyzed the various inputs from the controller 18 multiple times, it may calculate an error resulting from each analysis to assess whether a change in one or more of the weights has improved the quality of the output produced by the neural network 118 (i.e., if the modification(s) transmitted to and then implemented by the controller 18 after an analysis by the neural network 118 have caused the interface to move closer to the target interface position than the modification(s) transmitted after a previous analysis). The neural network 118 may then employ backpropagation (as shown in Fig. 38) to update the weight of one or more of the hidden layer nodes (and, thus, the relative importance of the corresponding input parameter(s) provided to the neural network 118 by the controller 18) so as to improve the quality of the output produced by the neural network 118 when it next analyzes the various inputs from the controller 18.

[0134] Aspects

[0135] Aspect 1. A blood separation device, comprising: a pump system; a centrifugal separator; an interface monitoring assembly; a neural network; and a controller coupled to the neural network, wherein the controller is programmed to execute a blood separation procedure in which the controller actuates the pump system to pump blood into the centrifugal separator at an inflow rate, actuates the centrifugal separator to operate at a rotational rate to separate the blood in the centrifugal separator into two blood components, with an interface therebetween, actuates the pump system to pump one of said blood components out of the centrifugal separator at an outflow rate, actuates the interface monitoring assembly to determine a current position of the interface within the centrifugal separator, transmits the current position of the interface and a plurality of input parameters to the neural network, receives a signal from the neural network that is indicative of a modification to be made to at least one procedural parameter so as to move the interface from the current position to a target interface position, and actuates the pump system and / or the centrifugal separator so as to implement said modification and move the interface from the current position toward the target interface position, wherein the neural network is programmed to employ artificial intelligence techniques to analyze the current position of the interface and at least one of said plurality of input parameters and determine said modification based at least in part on the current position of the interface and said at least one of said plurality of input parameters, and said at least one procedural parameter is selected from the group consisting of the inflow rate, the rotational rate, the outflow rate, and a location of the target interface position.

[0136] Aspect 2. The blood separation device of Aspect 1, wherein said plurality of input parameters includes the inflow rate, the rotational rate, and / or the outflow rate.

[0137] Aspect 3. The blood separation device of any one of the preceding Aspects, wherein said plurality of input parameters includes at least one characteristic of the blood.

[0138] Aspect 4. The blood separation device of Aspect 3, wherein said at least one characteristic of the blood includes a temperature of the blood.

[0139] Aspect 5. The blood separation device of Aspect 3, wherein said at least one characteristic of the blood includes the composition of the blood.

[0140] Aspect 6. The blood separation device of any one of the preceding Aspects, wherein said plurality of input parameters includes at least one blood source characteristic.

[0141] Aspect 7. The blood separation device of Aspect 6, wherein said at least one blood source characteristic is the age or the sex of the blood source.

[0142] Aspect 8. The blood separation device of any one of the preceding Aspects, wherein said plurality of input parameters includes a current phase of the blood separation procedure.

[0143] Aspect 9. The blood separation device of any one of the preceding Aspects, wherein the neural network is programmed to employ machine learning.

[0144] Aspect 10. The blood separation device of any one of the preceding Aspects, wherein the neural network is programmed to employ deep learning. Aspect 11. A controller-implemented method for executing a blood separation procedure, comprising: actuating a pump system to pump blood into a centrifugal separator at an inflow rate; actuating the centrifugal separator to operate at a rotational rate to separate the blood in the centrifugal separator into two blood components, with an interface therebetween; actuating the pump system to pump one of said blood components out of the centrifugal separator at an outflow rate; actuating an interface monitoring assembly to determine a current position of the interface within the centrifugal separator; transmitting the current position of the interface and a plurality of input parameters to a neural network; receiving a signal from the neural network that is indicative of a modification to be made to at least one procedural parameter so as to move the interface from the current position to a target interface position; and actuating the pump system and / or the centrifugal separator so as to implement said modification and move the interface from the current position toward the target interface position, wherein the neural network is programmed to employ artificial intelligence techniques to analyze the current position of the interface and at least one of said plurality of input parameters and determine said modification based at least in part on the current position of the interface and said at least one of said plurality of input parameters, and said at least one procedural parameter is selected from the group consisting of the inflow rate, the rotational rate, the outflow rate, and a location of the target interface position.

[0145] Aspect 12. The method of Aspect 11, wherein said plurality of input parameters includes the inflow rate, the rotational rate, and / or the outflow rate.

[0146] Aspect 13. The method of any one of Aspects 11-12, wherein said plurality of input parameters includes at least one characteristic of the blood.

[0147] Aspect 14. The method of Aspect 13, wherein said at least one characteristic of the blood includes a temperature of the blood.

[0148] Aspect 15. The method of Aspect 13, wherein said at least one characteristic of the blood includes the composition of the blood.

[0149] Aspect 16. The method of any one of Aspects 11-15, wherein said plurality of input parameters includes at least one blood source characteristic.

[0150] Aspect 17. The method of Aspect 16, wherein said at least one blood source characteristic is the age or the sex of the blood source. Aspect 18. The method of any one of Aspects 11-17, wherein said plurality of input parameters includes a current phase of the blood separation procedure.

[0151] Aspect 19. The method of any one of Aspects 11-18, wherein the neural network is programmed to employ machine learning.

[0152] Aspect 20. The method of any one of Aspects 11-19, wherein the neural network is programmed to employ machine learning.

[0153] It will be understood that the embodiments described above are illustrative of some of the applications of the principles of the present subject matter.

[0154] Numerous modifications may be made by those skilled in the art without departing from the spirit and scope of the claimed subject matter, including those combinations of features that are individually disclosed or claimed herein. For these reasons, the scope hereof is not limited to the above description but is as set forth in the following claims, and it is understood that claims may be directed to the features hereof, including as combinations of features that are individually disclosed or claimed herein.

Claims

CLAIMS1. A blood separation device, comprising:a pump system;a centrifugal separator;an interface monitoring assembly;a neural network; anda controller coupled to the neural network, wherein the controller is programmed to execute a blood separation procedure in which the controller actuates the pump system to pump blood into the centrifugal separator at an inflow rate,actuates the centrifugal separator to operate at a rotational rate to separate the blood in the centrifugal separator into two blood components, with an interface therebetween,actuates the pump system to pump one of said blood components out of the centrifugal separator at an outflow rate,actuates the interface monitoring assembly to determine a current position of the interface within the centrifugal separator,transmits the current position of the interface and a plurality of input parameters to the neural network,receives a signal from the neural network that is indicative of a modification to be made to at least one procedural parameter so as to move the interface from the current position to a target interface position, andactuates the pump system and / or the centrifugal separator so as to implement said modification and move the interface from the current position toward the target interface position, whereinthe neural network is programmed to employ artificial intelligence techniques to analyze the current position of the interface and at least one of said plurality of input parameters and determine said modification based at least in part on the current position of the interface and said at least one of said plurality of input parameters, andsaid at least one procedural parameter is selected from the group consisting of the inflow rate, the rotational rate, the outflow rate, and a location of the target interface position.

2. The blood separation device of claim 1, wherein said plurality of input parameters includes the inflow rate, the rotational rate, and / or the outflow rate.

3. The blood separation device of any one of the preceding claims, wherein said plurality of input parameters includes at least one characteristic of the blood.

4. The blood separation device of claim 3, wherein said at least one characteristic of the blood includes a temperature of the blood.

5. The blood separation device of claim 3, wherein said at least one characteristic of the blood includes the composition of the blood.

6. The blood separation device of any one of the preceding claims, wherein said plurality of input parameters includes at least one blood source characteristic.

7. The blood separation device of claim 6, wherein said at least one blood source characteristic is the age or the sex of the blood source.

8. The blood separation device of any one of the preceding claims, wherein said plurality of input parameters includes a current phase of the blood separation procedure.

9. The blood separation device of any one of the preceding claims, wherein the neural network is programmed to employ machine learning.

10. The blood separation device of any one of the preceding claims, wherein the neural network is programmed to employ deep learning.

11. A controller-implemented method for executing a blood separation procedure, comprising:actuating a pump system to pump blood into a centrifugal separator at an inflow rate;actuating the centrifugal separator to operate at a rotational rate to separate the blood in the centrifugal separator into two blood components, with an interface therebetween;actuating the pump system to pump one of said blood components out of the centrifugal separator at an outflow rate;actuating an interface monitoring assembly to determine a current position of the interface within the centrifugal separator;transmitting the current position of the interface and a plurality of input parameters to a neural network;receiving a signal from the neural network that is indicative of a modification to be made to at least one procedural parameter so as to move the interface from the current position to a target interface position; andactuating the pump system and / or the centrifugal separator so as to implement said modification and move the interface from the current position toward the target interface position, whereinthe neural network is programmed to employ artificial intelligence techniques to analyze the current position of the interface and at least one of said plurality of input parameters and determine said modification based at least in part on the current position of the interface and said at least one of said plurality of input parameters, andsaid at least one procedural parameter is selected from the group consisting of the inflow rate, the rotational rate, the outflow rate, and a location of the target interface position.

12. The method of claim 11, wherein said plurality of input parameters includes the inflow rate, the rotational rate, and / or the outflow rate.

13. The method of any one of claims 11-12, wherein said plurality of input parameters includes at least one characteristic of the blood.

14. The method of claim 13, wherein said at least one characteristic of the blood includes a temperature of the blood.

15. The method of claim 13, wherein said at least one characteristic of the blood includes the composition of the blood.

16. The method of any one of claims 11-15, wherein said plurality of input parameters includes at least one blood source characteristic.

17. The method of claim 16, wherein said at least one blood source characteristic is the age or the sex of the blood source.

18. The method of any one of claims 11-17, wherein said plurality of input parameters includes a current phase of the blood separation procedure.

19. The method of any one of claims 11-18, wherein the neural network is programmed to employ machine learning.

20. The method of any one of claims 11-19, wherein the neural network is programmed to employ machine learning.

Images