WO1999002993A1

WO1999002993A1 - Fluid handler and method of handling a fluid

Info

Publication number: WO1999002993A1
Application number: PCT/US1998/014117
Authority: WO
Inventors: Kevin C. Warnke
Original assignee: Abbott Laboratories
Priority date: 1997-07-08
Filing date: 1998-07-08
Publication date: 1999-01-21

Abstract

Methods of fluid handling are disclosed. In one method, a pump fluidly associated with a fluid conveying conduit is energized to move fluid with respect to the fluid conveying conduit. A pressure transducer fluidly associated with the fluid conveying conduit is energized to continuously monitor pressure within the fluid conveying conduit. Pressure within the fluid conveying conduit is continuously monitored with the pressure transducer. It is determined whether movement of fluid with respect to the fluid conveying conduit is intended or unintended based on the continuously monitored pressure within the fluid conveying conduit.

Description

FLUID HANDLER AND METHOD OF HANDLING A FLUID

REFERENCE TO RELATED APPLICATIONS

The application is a continuation-in-part application of U.S. Serial No. 08/XXX,XXX filed on May 22, 1997, which is a continuation-in-part application of U.S. Serial No. 08/572,835 filed on December 14, 1995. Those previous applications are assigned to the assignee of the present case and their disclosures are incorporated herein by this reference.

BACKGROUND OF INVENTION

Embodiments described herein relate generally to handling a fluid automatically. More specifically, the embodiments relate to moving a fluid into and out of a nozzle. Fluid handlers or generally mechanisms for moving a fluid have many applications. One such application involves a type of machine referred to as an automated medical analyzer. These analyzers perform medical tests on a sample, such as blood, urine and the like. These tests may require mixing of the sample with a fluid. Once the sample is mixed with the fluid, a •chemical reaction may take place. The chemical reaction can be used to provide a medical professional with medical information about the sample. Because the sample may have come from a human patient, the medical information about the sample may provide the medical professional with information about the patient's medical condition.

For these analyzers to operate as intended, and to give the correct medical information to the medical professional, it is desirable to mix only a specific amount of sample with a specific amount of fluid. If too much or too little sample, or if too much or too little fluid is used during the mixing process, then the chemical reaction between the sample and the fluid may not occur as desired. If this occurs, the test may give inaccurate medical information about the sample and about the patient. It may not be in the patient's best interests for a medical professional to be given inaccurate medical information. Therefore, it is desirable to provide a fluid handler which can alert an automated medical analyzer operator when the wrong amount of sample and/or fluid has been used with a particular medical test.

SUMMARY OF INVENTION

Methods of fluid handling are disclosed. In one method, a pump fluidly associated with a fluid conveying conduit is energized to move fluid with respect to the fluid conveying conduit. A pressure transducer fluidly associated with the fluid conveying conduit is energized to continuously monitor pressure within the fluid conveying conduit. Pressure within the fluid conveying conduit is continuously monitored with the pressure transducer. It is determined whether movement of fluid with respect to the fluid conveying conduit is intended or unintended based on the continuously monitored pressure within the fluid conveying conduit. BRIEF DESCRIPTION OF DRAWINGS

Fig. 1 is a generic diagrammatic view of a fluid handler described herein; Fig. 2 is a block schematic diagram of a portion of the fluid handler shown in Fig. 1;

Fig. 3 is a schematic diagram of a portion of the fluid handler shown in Fig. 2;

Figs . 4A and 4B display data obtained during intended operation of the fluid handler;

Figs . 5A and 5B display data obtained during unintended operation of the fluid handler;

Figs. 6A and 6B display filtered data obtained during operation of the fluid handler; and Fig. 7 is a schematic diagram of a circuit for use with the fluid handler.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

Embodiments described in detail below handle fluids. These embodiments can be used to handle, e.g. aspirate, dispense, mix, etc., any suitable fluid. While, for the sake of clarity of understanding, the embodiments are discussed with respect to their employments in a medical analytical instrument, it is to be understood that the embodiments may be used in any appropriate employment. Also, regarding the methods of handling a fluid described below, it is to be recognized that the steps of one method may be performed in any suitable order and that steps from one method may be combined with steps from another method to yield yet additional methods . To more clearly identify operation of the fluid handler 10, both intended, i.e. correctly performed, and unintended, i.e. incorrectly performed, fluid handling operations of the fluid handler 10 will be discussed. The fluid handler 10 is sufficient to detect unintended fluid handling operations due to factors such as clogs, clots, debris, bubbles, foam, etc. By detecting unintended fluid handling operations as they occur, it is possible to disregard tests involved without having to rely on the possibility of the test result itself being out of range to indicate an unintended fluid handling operation. This can also reduce the probability of operator error.

Fig. 1 illustrates one embodiment of a fluid handler 10. This illustrated embodiment 10 comprises a nozzle 12 having a tip 14 connected via a conduit 16 to a pump 18 operated by a prime mover 20. A pressure transducer 22 having a sensor 24 is fluidly connected with the conduit 16 between the nozzle 12 and the pump 18. The nozzle 12 may be movably supported by a suitably constructed gantry, not shown for clarity, so that the nozzle 12 may move toward and away from a container 26 retaining a fluid 28 having a fluid surface 30. The nozzle 12, the conduit 16 and the pump 18 contain a liquid 32, such as distilled water, a buffer, and the like, which facilitates aspiration and dispense of fluid 28. Thus, the pressure transducer 22 is "in line" with the liquid 32. To control and to monitor operation of the fluid handler 10, the pressure transducer 22 is electrically connected with operational electronics 34 which are, in turn, electrically connected with a controller 36. The controller 36 is electrically connected with the prime mover 20 of the pump 18 so that the pump 18 and the operational electronics 34 can operate in unison.

In one particular construction of the fluid handler 10, the nozzle 12 is a rigid aspiration probe. This probe has an inner diameter of about 0.04 inches, a length of about 7 inches, and a tip with an inner diameter of about 0.014 inches and a length of about 0.278 inches. In this construction, the conduit 16 is made of a polymer of low compliance, such as TEFZEL tubing (available from Du Pont Co., Wilmington, DE) , with an inner diameter of about 0.063 inches. If it is desirable to minimize damping of transient pressure variations, then the axial length of the conduit 16, particularly between the nozzle 12 and the pressure transducer 22, should be minimized or rendered as short as practical. However, it is to be remembered that the pressure transducer 22 may be located at any place so long as the pressure transducer 22 is able to substantially continuously monitor pressure, including but not limited to transient pressure variations, fluidly associated with the nozzle 12.

In one construction, the pressure transducer 22 may be a TransPac IV, manufactured by Abbott Laboratories, Salt Lake City, UT. Generally, the pressure transducer 22 should be capable of sensing rapid transient pressure changes within a range of at least about -2 to about 6 psig, and particularly should have an overpressure capability to about 100 psig to enable effective washing of the nozzle 12. The pressure transducer 22 should have a fast response time, up to about 10 kHz. The pressure transducer 22 monitors pressure within the conduit 16 substantially continuously, thereby monitoring transient conduit 16 pressures, i.e. pressures fluidly associated with the nozzle 12, which allows unintended aspirations and dispenses to be indicated almost immediately. Significant transient variations in the pressure fluidly associated with the nozzle 12 may exist for a limited time period. Thus, it is desirable to monitor the pressure fluidly associated with the nozzle 12 in a substantially continuous manner. In one embodiment, the pressure within the conduit 16 is monitored at a rate of about 1000 times per second. In other embodiments, the pressure within the conduit 16 is monitored at a rate of about 100 times per second, 10 times per second, once per second, or any suitable rate, which may be dictated by such factors as desired sensitivity, available computer memory, etc. In any case, it is to be noted that the pressure within the conduit 16 is monitored substantially continuously so as to identify unintended fluid movements of any degree of interest. In one construction, the pump 18 is a syringe pump, such as a Cavro 3000 (Cavro Scientific Instruments Inc., Sunnyvale, CA) . It is to be recognized that these specific constructions and others below are given for the sake of facilitating understanding .

A description of a particular construction of the operational electronics 34 follows with reference to Figs. 2 and 3. However, it is emphasized that this is only one possible construction for controlling operation of the fluid handler 10. For instance, portions of the operational electronics 34 and/or its functions may be incorporated into a computer 38, provided in hardware or software form, etc.

As shown in Fig. 2, the pressure transducer 22 is electrically connected to an electronic buffer circuit 40 so that an electrical signal generated by the sensor 24 responsive to a pressure in the conduit 16 is transmitted to the buffer circuit 40. A detailed schematic of one embodiment of the buffer circuit 40 is illustrated in Fig. 3. The buffer circuit 40 of Fig. 3 includes a voltage source circuit 42 which supplies about 10 VDC to a bridge circuit 44 associated with the pressure transducer 22. An electrical output of the bridge circuit 44 is electrically connected to a differential amplifier circuit 46, which is also connected to an adjustable offset voltage generator circuit 48. An output of the differential amplifier circuit 46 is the output of the electronic buffer circuit 40. The output voltage of the differential amplifier circuit 46 can be expressed as:

Vout = (__2R3_ + 2⁾(V₂ - V_x) + V_A R₅ + R₆

The output of the differential amplifier circuit 46 is electrically connected to an analog to digital converter 50. The analog to digital converter 50 may be provided as part of the computer 38 or as part of another computer. In another construction, the analog to digital converter 50 may be substantially similar to an ATMIO-16-L9 from National Instruments of Austin, TX . In the illustrated embodiment, an output of the analog to digital converter 50 is electrically connected to a digital filter 52. In one embodiment, the digital filter 52 may be substantially similar to a digital lowpass filter, such as a digital Butterworth filter algorithm and the like. This algorithm may be performed by the computer 38. In a particular embodiment, the filter algorithm of the commercial software package MathCAD (available from MathSoft, Inc., Cambridge, MA) may be used. In other embodiments, filtering may be performed by an analog filter incorporated into the buffer circuit 40 or some other digital filter performed by the computer 38. An output of the digital filter 52 is electrically connected to an error detector 54. The error detector 54 may be of any suitable construction, such as a software routine, a hardware configuration, an electrical circuit etc., that performs fluid handling error detection tasks as detailed below. In some embodiments, the error detector 54 may be part of the computer 38 or may be a separate signal processing device or microcontroller. The computer 38 or the error detector 54 may be electrically connected to the controller 36 to allow operation of the pump 18 to be responsive to a signal from the pressure transducer 22.

Further appreciation of the construction of the fluid handler 10 may be obtained with a discussion of the operation of the fluid handler 10. This discussion is provided for clarity of understanding.

According to one method of operation, referring to Figs. 1 through 3, liquid 32 fills the nozzle 12, the conduit 16, the pressure transducer 22 and the pump 18. In some cases, the pump 18 may move the liquid 32 within the fluid handler 10 such that a volume, about 5 to 10 μl, of ambient air may be located within the nozzle 12 adjacent the tip 14. This volume of air may be used to separate fluid 28 aspirated from the container 26 into the nozzle 12 and the liquid 32 already present in the nozzle 12. This procedure can also be used to "stack" different fluids 28 within the nozzle 12. In essence, a first fluid 28, such as sample, is aspirated (discussed below) into the nozzle 12 and is separated from the liquid 32 by the first air volume. When a second fluid 28, such as a reagent, is aspirated into the nozzle 12, a second air volume separates the second fluid 28 from the first fluid 28. Then, the first and second fluids 28 can be dispensed from the nozzle 12 into the same of a different container 26, such as a reaction vessel. This procedure can be used to increase throughput . To handle fluid 28, the nozzle 12 is moved by a suitable prime mover, not shown for clarity, towards the surface 30 of the fluid 28 to be handled. The nozzle 12 is moved so that the tip 14 of the nozzle 12 is offset sufficiently below the surface 30 of the fluid 28 within the container 26. An appropriate level sense mechanism, such as an R-F level sense, a capacitive level sense, a pneumatic level sense and the like, may be operatively associated with the nozzle 12 to facilitate automated detection of the surface 30 of the fluid 28. The distance of the offset between the surface 30 of the fluid 28 and the tip 14 of the nozzle 12 may vary depending upon fluid characteristics, level sense mechanism employed, ambient conditions, etc.

Once the tip 14 of the nozzle 12 is sufficiently offset below the surface 30 of the fluid 28 within the container 26, the controller 36 energizes the prime mover 20 associated with the pump 18. Substantially simultaneously, the controller 36 also energizes the pressure transducer 22. It is to be noted that, in some embodiments, the pressure transducer 22 may be energized, and thereby substantially continuously monitor pressure fluidly associated with the nozzle 12 at any time and for any time period extending between a first time prior to energizing of the prime mover 20 and a second time after completion of operation of the prime mover 20.

The pump 18 moves or aspirates a desired amount of fluid 28 from the container 26 and into the interior of the nozzle 12. As the fluid 28 passes from the container 26 into the interior of the nozzle 12, the sensor 24 of the pressure transducer 22 constantly measures pressure within the conduit 16 and detects a transient pressure change caused by flow of fluid 28 and liquid 32 within the nozzle 12 and the conduit 16 influenced by operation of the pump 18. The signals energizing the prime mover 20 sent from the controller 36 to the prime mover 20 correspond to a signal expected to be received from the sensor 24. This facilitates analysis of the signal from the sensor 24. Responsive to the pressure sensed by the sensor 24 of the pressure transducer 22, the bridge circuit 44 generates an electrical signal. The electrical signal generated by the bridge circuit 44 is sent to the differential amplifier circuit 46 of the buffer circuit 40. The buffer circuit 40 produces an amplified electrical signal representative of an instantaneous pressure sensed by the sensor 24. The amplified electrical signal is fed into the computer 38. Once the amplified electrical signal is sent to the computer 38, the analog to digital converter 50 generates a digital signal corresponding to a voltage of the amplified electrical signal emitted by the ^' buffer circuit 40. Thus, the digital signal is indicative of the pressure sensed by the sensor 24. The digital signal may be filtered by the digital filter 52. The filtered signal is processed by the error detector 54 which monitors a transient pressure signal to determine whether an intended aspiration has occurred. In one embodiment, the error detector 54 is enabled by a signal from the controller 36, indicating that aspiration has commenced and that the signal from the sensor 24 should be monitored for indication of errors or unintended aspirations. The illustrated embodiment of the fluid handler 10 may also be used to monitor dispense of fluid 28 from the nozzle 12 into the container 26 or some other fluid receiver, such as a reaction vessel and the like. Substantially similarly to the above-described steps, a fluid 28 dispense is initiated by operation of the pump 18. The error detector 54 is enabled by the controller 36 substantially simultaneously with the pump 18. The error detector 54 monitors the signal, which is indicative of the pressure sensed by the sensor 24, sent by the sensor 24 and processed by the operational electronics 34 to determine whether an intended dispense has occurred.

Several methods exist for using the signal, indicative of pressure sensed by the sensor 24, sent from the sensor 24 to determine whether an intended aspiration and/or dispense has occurred based on transient pressure signals sent by the sensor 24. Some of these methods are described in the following examples . It is to be noted that these examples are provided for clarity of understanding.

In the examples, data were collected using an embodiment of the fluid handler 10 which is substantially similar to that described above. Fluid 28 volumes of about 50 μl were aspirated into the nozzle 12, with fluid 28 flow accelerated at a substantially constant rate of about 1563 μl/sec/sec until steady state flow rate of about 83 μl/sec was reached. Flow deceleration was also substantially constant at about 1563 μl/sec/sec. Dispenses of about 40 μl were also measured, with a steady flow rate of about 391 μl/sec and flow acceleration and deceleration rates of about 26,040 μl/sec/sec. The operational electronics 34 were adjusted (gain, threshold, etc.) so that an electrical signal presented to the analog to digital converter 50 is about 0 volts prior to aspiration and does not saturate the operational electronics 34 during an intended aspirate/dispense cycle, which comprises aspiration of the desired amount of fluid 28 into the nozzle 12 and dispense of a desired amount of fluid 28 from the nozzle 12. A common gain of the operational electronics 34 is about 9.6 V / psi. A representation of MathCAD software used to collect and to analyze the data in these examples is presented in Appendix A. The following examples use parameters defined in Appendix A, section 1. Example 1 - Intended Aspiration and Dispense profiles

About 50 μl aspiration and about 40 μl dispense of pig serum were performed with the flow parameters as listed above. The analog pressure signal from the sensor 24 was sampled substantially continuously at a rate of about 1000 samples/second and stored as unfiltered numerical data representing the instantaneous pressure profiles sensed by the sensor 24 during both aspiration and dispense of fluid 28. The results of 5 repetitions of this experiment are presented in Fig. 4A (5 aspirations) and Fig. 4B (5 dispenses) . These Figures demonstrate repeatability of the pressure sensing method. The pressure spikes immediately preceding and following the aspiration and dispense were artificially imposed on the data to facilitate extraction and display of the pressures during actual aspiration and dispense. A representation of MathCAD software used to extract the pressures during aspiration and dispense is found in Appendix A, section 2. From this, it is apparent that an instantaneous pressure profile derived from data obtained by the sensor 24 can be compared to a predetermined pressure profile to conclude whether the fluid handling operation involved was intended or unintended.

Example 2 - Unintended Aspiration and Dispense

Interrupted aspiration of pig serum was performed by removing the container 26 from the nozzle 12 during an about 50 μl aspiration, resμlting in partial aspiration of air. The dispense portion of the aspirate/dispense cycle consequently also included partial dispense of air. Instantaneous pressure profiles of 5 unintended aspirations and dispenses were produced as in Example 1 and are illustrated in Fig. 5A (5 aspirations) and Fig. 5B (5 dispenses) . Varying amounts of air were aspired during the unintended aspirations. A representation of the MathCAD software used to extract the pressures during the unintended aspiration and dispense is found in Appendix A, section 3.

Example 3 - Digital Filtering of Pressure Profiles

The pressure profile data acquired as described in examples 1 and 2 was filtered with a MathCAD digital Butterworth filter described above. This is a 31 coefficient digital lowpass filter with a cutoff frequency of about .01 times the sampling frequency (or about 10 Hz) . A representation of the MathCAD software used to execute this filtering is found in Appendix A, section 4. The results of this filtering are illustrated in Fig. 6A (5 intended and 5 unintended aspirations) and Fig. 6B (5 intended and 5 unintended dispenses) . The unintended aspirations and dispenses are clearly distinguishable from the intended aspiration and dispense profiles.

Example 4 - Error Detection bv Integration

Intended and unintended filtered and unfiltered pressure profile data was integrated as an exemplary error detection method. The integration was mathematically performed with

MathCAD software (represented in Appendix A, sections 2, 3, and 5) . The results of the integration are presented in Table 1. The integration results for the intended aspirations and dispenses are statistically distinguishable from those of the unintended aspirations and dispenses. This method is, however, sensitive to fluid viscosity (which may range from about 1 to about 14 centipoise) and accordingly is less desirable for small aspiration and dispense volumes (on the order of about 10 μl) .

Example 5 - Error Detection bv Mean Pressure Difference

Five consecutive data points were sampled from each of the filtered intended and unintended aspiration pressure profiles immediately prior to syringe deceleration, so that they represent pressure sensed by the sensor 24 during steady-state. Five consecutive data points were also sampled following completion of aspiration. The difference between the mean pressure during steady state aspiration and the mean pressure following completion of the aspiration was calculated and compared to a tolerance band. In this example, executed by the software represented in Appendix A, section 6, the tolerance band was from about 0.35 to about 0.55. As shown in Table 2, the intended aspirations yielded mean pressure differences within the tolerance band, while the unintended aspirations yielded pressure differences outside the band. Example 6 - Error Detection bv Variation of Pressure Difference from Mean

Data points from each of the filtered intended and unintended aspiration pressure profiles that fell within the steady state aspiration region were used to calculate the instantaneous difference between the pressure sensed by the sensor 24 at any time and the pressure sensed following completion of the aspiration. Each of these values was compared to the mean difference between the pressure during steady state aspiration and the pressure following aspiration (calculated as in Example 5) . The number of times that the instantaneous pressure difference differed from the mean pressure difference by more than a specified tolerance was counted. In the case of the software represented in Appendix A, section 7, the tolerance band was about 0.1 on either side of the mean difference. As shown in Table 3, the intended aspiration profiles had uniformly zero deviations outside of the tolerance band, while the unintended aspirations had numerous such deviations.

The error detection methods presented in Examples 4 through 6 are just four of the many possible methods to detect unintended aspirations and dispenses from the pressure profiles measured by the sensor 24. In some embodiments, several error detection methods may be combined. The Examples demonstrate that the pressure data collected by embodiments of the fluid handler 10 may be used in an aspiration and dispense error detection method.

The pressure profiles of intended and unintended aspirations and dispenses are directly affected by the acceleration and deceleration profiles of the pump 18. For instance, a uniform ramp acceleration and deceleration of the pump 18 will result in pressure values that do not reach steady state. It is possible to modify the acceleration and deceleration profiles of the pump 18 to yield intended and unintended aspiration and dispense pressure profiles that may be more easily or reliably distinguished by these or other error detection schemes. In other embodiments, feedback from elements, other than the pressure transducer 22, of the fluid handler 10, such as movement of the pump 18, fluid surface 30 level sense information, etc., alone or in combination with each other or with the pressure transducer 22 information, can be used to provide an indication of intended or unintended fluid handling operations.

Table 1 Error Detection by Integration of Pressure

Table 2: Error Detection by Mean Pressure

Difference

Mean pressure difference

Intended Unintended

0.496 Within 0.141 Outside tolerance tolerance

0.528 Within 0.169 Outside tolerance tolerance

0.517 Within 0.225 Outside tolerance tolerance

0.501 Within 0.185 Outside tolerance tolerance

0.509 Within 0.230 Outside tolerance tolerance Table 3: Error Detection by Variation of Pressure Difference from Mean

Example 7 -- Error detection bv Integration with Temperature Compensation

Some methods of distinguishing intended from unintended fluid movements may be affected by the viscosity of the fluid 32 within the nozzle 12 and the conduit 16. In particular, integrating the pressure substantially continuously monitored by the transducer 22 yields a result which may vary with viscosity of the fluid 32. The viscosity of the fluid 32 may be affected by temperature of the fluid 32.

If the fluid 32 were water, for example, the viscosity of the fluid may vary by about thirty percent over a temperature range of about 15° to about 30° Celsius. The resulting integrated pressure can vary by about 22 percent, as shown in Table 4. With no knowledge of the temperature of the fluid 32, this variation may reduce the ability to distinguish an intended fluid movement (in this case, an aspiration of fluid) from an unintended fluid movement (in this case, an aspiration of only air) into the nozzle 12.

Table 4

Fluid temperature Integrated Pressure

(° C) Flu d Aspiration Air Aspiration 15 197 135

22.5 179 118

30 156 100

It may be desirable, therefore, to utilize a method of reducing the effect of variability in the integrated pressure value due to the temperature of the fluid 32. One such method may be to measure the temperature of the fluid 32, or of a location that is indicative of the temperature of the fluid 32 to provide direct or indirect fluid 32 temperature measurement, and use that measured temperature in determining whether the aspiration was intended or unintended.

In one embodiment of this method, a relationship between the temperature of the fluid 32, the volume of the fluid aspirated, and the integrated pressure value is determined to be:

P dt = a*V + b*T + c*V*T + d

where V is the volume of the aspiration in microliters, T is the temperature of the fluid 32 in degrees Celsius, and a, b, c, and d are constants empirically determined to be 9.99, -0.852, -0.106, and 44.2, respectively. The pressure which is integrated is the electrical signal from the transducer 22 prior to fluid motion minus that signal at each time point during the aspiration. This relationship allows the calculation of an integrated pressure ratio equal to:

Integrated Pressure Ratio = P dt / (a*V + b*T + c*V*T + d)

which will be close to 1.0 for an intended fluid movement. Table 5, sample values for one embodiment of this method applied to an about 20 μl aspiration, illustrates that the Integrated Pressure Ratio is essentially constant as the temperature of the fluid 32 varies, allowing one to distinguish the intended from unintended fluid movement .

Table 5

Fluid temperature Integrated Pressure Ratio

(° C) Fluid Aspiration Air Aspiration

15 0.99 0.68

22.5 1.01 0.67

30 1.01 0.64 Example 8 - Error detection bv Integration with Constant Fluid Flow Compensation

Another method of reducing the effect of variation in the viscosity of the fluid 32 is to store a value of the signal indicative of the fluid pressure as sensed by the transducer 22 during a period of known constant flow of the fluid 32 from the pump 18 or some other pump through and out of the nozzle 12. This constant pressure is, similar to the integrated pressure during a fluid aspiration, a function of the viscosity of the fluid 32 and may be used in distinguishing an intended from an unintended fluid movement .

In one embodiment of this method, a relationship between the pressure during an about 1000 microliter/second outward flow of the fluid 32, the volume of the fluid aspiration, and the integrated pressure value was determined to be:

P dt a*V + b*W + c*V*W + d

where V is the volume of the aspiration in microliters, and W is the value of the pressure signal sensed prior to any fluid motion minus that signal measured during an about 1000 microliter/second flow of the fluid 32 through the conduit 16 and out the nozzle 12. The parameters a, b, c, and d are constants empirically determined to be 2.04, -47.03, -3.25, and -79.80, respectively. This relationship allows the calculation of an integrated pressure ratio equal to:

Integrated Pressure Ratio = P dt / (a*V + b*W + c*V*W + d)

which will be close to 1.0 for an intended fluid movement. The pressure which is integrated is the electrical signal from the transducer 22 prior to fluid motion minus that signal at each time point, such as about once per second, about 10 times per second, about 100 times per second, about 1000 times per second, etc., during the aspiration. Table 6 illustrates that the Integrated Pressure Ratio is essentially constant as the temperature of the fluid 32 varies, allowing distinguishing intended from unintended fluid movement . The data given below is for aspiration of about 50 μl .

Table 6

Fluid temperature Integrated Pressure Ratio f° C ) Fluid Aspiration Air Aspiration

15 1.02 0.48

22.5 1.01 0.42

30 0.95 0.29

This method provides an additional advantage of reducing an effect of variation in the sensitivity of the sensor 24 of the pressure transducer 22. Such variation in sensitivity will affect both the pressure signal measured during the fluid movement and the pressure signal measured during constant outward fluid flow through the nozzle 12 (although the gain of the operational electronics 34 may be different for the two measurements) . As illustrated by Table 7, a variation in sensor sensitivity of 40 percent yields a variation in the Integrated Pressure Ratio of only 15 percent, which increases the sensitivity of this method of distinguishing an intended from an unintended fluid movement .

Table 7 Sensor Sensitivity Integrated Pressure Fluid Aspiration Integrated

Pressure Ratio Low (8.4 mV/psi) 557 0.85

High (14.0 mV/psi) 846 1.00 As this example shows, this method comprises three steps. To illustrate further by example, upon starting of the pump 18, pressure fluidly associated with the nozzle 12 is substantially continuously measured. The pressure may be substantially continuously monitored at any appropriate rate, such as about once per second, about 10 times per second, about 100 times per second, about 1000 times per second, etc. Pressure monitoring continues during movement of a first fluid with respect to the nozzle 12, e.g. a fluid dispense, in some cases at a known, substantially steady fluid flow rate. In one embodiment, the fluid flow rate may be about 1000 μl per second. This fluid may be the liquid 32 and the flow of first fluid may be directed out of the nozzle 12, thereby allowing the pressure to be obtained while washing the nozzle 12 and the conduit 16 with a flow of liquid 32.

An average of that pressure substantially continuously monitored during movement of the first fluid is calculated. That average pressure is used to calculate or determine a threshold, as determined above to give a ratio of about 0.5, for a second fluid movement, e.g. fluid aspiration, of a certain volume with respect to the nozzle 12.

Then, during movement of the second fluid with respect to the nozzle 12, pressure fluidly associated with the nozzle 12 is again substantially continuously monitored by the transducer 22. This movement of second fluid may be directed into the nozzle 12 and the second fluid may be moved by the pump 32 or another pump. The substantially continuously monitored pressure, for instance during and shortly after, such as about 20μsec after, such a second fluid movement, e.g. aspiration, is integrated and the integrated pressure, or more specifically the integrated pressure ratio, is compared to the threshold calculated or determined for that volume of fluid to determine if the aspiration was intended or unintended. The threshold may be fixed or may change dependent on such factors as volume of second fluid, particular characteristics of the second fluid, etc. This method may be used to reduce an impact of changes in viscosity of the liquid 32 in the conduit 16 and resulting changes in measured pressures fluidly associated with the nozzle 12.

Example 9 - Error detection bv Frequency Analysis

Intended and unintended pressure profile data may be analyzed with Fourier Transform techniques to yield a frequency spectrum of one portion of- an aspiration. The lowest frequency components may then be analyzed using data entropy methods (such as described in Mathematics in Medicine and the Life Sciences by Hoppensteadt & Peskin, Springer-Verlag, 1992, pages 23 - 25) . In one embodiment of this method, 128 msec of pressure data was analyzed, from which the lowest 15 frequency values were analyzed using data entropy techniques. The mean entropy for intended fluid movement (fluid aspirations) was 0.024, while the mean entropy for unintended fluid movement (air aspirations) was 0.130, more than five times the value for fluid.

Example 10 - Error Detection Bv Continuous Integration

The examples of methods of distinguishing intended from unintended fluid motion by integrating pressure data, elucidated in examples 4, 7, and 8, utilized numerical integration techniques. Numerical integration utilizes substantially continuous, yet discrete, sampling of the fluid pressure monitored by the pressure transducer 22. As stated in example 8, this sampling rate can span a wide range, for example, from about less than 1 sample per second to about more than 1000 samples per second.

It is also possible to utilize continuous (analog) integration techniques in which pressure at the transducer 22 is sampled continuously. The result of such integration is substantially similar to that resulting from numerical integration techniques. A circuit 56 capable of performing continuous integration is illustrated in Fig. 7. At conductor 58, an electrical signal representing the pressure at the pressure transducer 22 is fed into an input of an integrator 60. This integrator 60 may be an analog device such as that described in Circuits, Devices, and Systems by Ralph J. Smith (John Wiley & Sons, New York, 1976, p. 512) . The disclosure of that reference is incorporated herein in its entirety.

The electrical signal representing the pressure at the pressure transducer 22 at conductor 58 is also fed through a low pass filter 62 and into a sample-and-hold circuit 64. The sample-and-hold circuit 64, such as that described in Digital Control of Dynamic Systems by Franklin and Powell (Addison-

Wesley, Reading, Mass, 1980, p. 44), the entirety of which is incorporate by reference, operates by fixing its output value at conductor 66 equal to the value of its input voltage at conductor 68 at the time of receipt of a sampling signal from conductor 70. The sampling signal at conductor 70 is provided by an indication of motion of the pump 18.

By virtue of the low pass filter 62 and the sample-and-hold circuit 64, the signal at conductor 68 is therefore equal to the average value of the pressure signal at conductor 58 during the period immediately prior to the motion of the pump 18, termed the "initial pressure." The signal at conductor 68, indicative of the initial pressure is fed through an inverter 72, which provides an output voltage equal to the negative value of the input signal at conductor 68, which is in turn fed into a second (summing) input of the integrator 60.

The initial condition of the integrator 60 is set to zero by a reset circuit 74 (such as described in Smith, p. 526) activated by the sampling signal at conductor 70, the same signal indicative of pump motion that triggers the sample-and- hold circuit 64. An output of the integrator 60 at conductor 76, therefore, is indicative of the integrated value of the pressure during pump motion minus the pressure immediately prior to pump motion. This value is substantially equal to the integrated pressure value described in examples 7 and 8 and may be analyzed in a similar fashion to distinguish intended from unintended fluid motion. Accordingly, it can be appreciated that it is possible to monitor pressure with the pressure transducer 22 in a substantially continuous, albeit discrete, method or in a truly continuous method. Electrical signals obtained by either method may be analyzed in substantially similar fashions.

APPENDIX A

Section 1

27

Test Run Tube - { Asp Volume - 5000 Disp Volume -4000 Tert Cycle Count— J

Theta (X) VeL - 6000 Tbeta (X) Act - 32000

Disp CAΓ.0 Vet - 20000 Disp CΛΓ.COAJΛ - 50000

Z Vel - 25513 Z Act - 985L 10

Σ Vet (liquid sense} - 25518 Z Act (liquid sense) - 1500000

Asp. Theta. Position * 20 Asp. Liquid Sense Over "0 Asp MaxZdαwα— 5000

Liquid Seαse Gain t OFF

Liquid Sense Gain 2 O fq id Sense Gain 3 OK

Aspirate Cosine Ramp OFF

Disp. Theta Position - -2486 Disp. Liq. Sense Over - I Disp. Mix Z down -4000

T st Were Run with SCALE Mode

AND taking weight measurements — 2nd weight Delay — 5

Dispense Cσsirve Ramp OFF

Wash Theta Position - -1327 Wish Max Z down - 3600

Priming Syr. Steps = 0

Frrrtfrg Vet. Draw = 25000 Priming AcL Dra « 200000

Priming VeL DLs ease = 60000 Priming AcL Dispense =* 200000

Pruning Syr. Overdrive - Q

Syr. Valve lc Step? <= 1030 S' r. Valve Bypass Steps - 2150 Syr. Valve Out Step* -0

Syr. Valve VeL = 20000 Syr. Valve AcL « 100000

Cycles to run prlm-e «* 5 Syr. V L Draw = L5CC0 Syr. AcL. Draw « 3COO00 Syr. VeL Dispense - 75000 Syr. AcL Dispense » 5000000 Syringe Resolution « 43000 Syringe. Size - 250 Syringe ovrr riv- ^β 2950

Syr. Ease VeL Draw - 0 Syr. Base Vet Dispense - 0. Int. Probe Wish VeL = 40000 InL Probe With Act « 500000 L- L Pr be Wish Cycles - 27500

Gibed ish V-L « 23000 Gϋsoc W sh AcL - 500000 QDSβc ash Cycles = 10C00

DeUy to s- t-mvil W «Q = 50 Delay la iπfw »I wish «= 0 Delay at wash bαttnra - 900 S-avnc Data Asp. ? situs = ICOOO Save Dais Dis. Points - 0

Asp. Dela I = 100 Asp. Delay 2 - 100 Disp. Delay t = C0 Diip. Delay 2 ^• 100 Pτ£-A_rp A T Gβ = Ϊ 52C ?c-r:_As Air Gu ^β 0 Full Cy ; i ^■» Ti rr ^'~, g QJCTMSCC) -Lasx Run

T c t-3 Ajpv acc = 5 ϊ 5

Thβ βα Aspirate "515

Z down Aspirate - 697 Syringe Aspirate - 832 Z Up Aspuatβ-851 Theta to Dispense - 669

Z Down Dispense - 205 Syringe Dispense - 551 Z Up Dispense - 301 Theta to aih, - 332

Wash, Down Cycle - 190 Wash. Up Cycle - 1088 To End of Cycle - 1 Total Cycle Time - 6282

A??qNPI£

Section 2

Asp Vβf - tβ.OOO AaptAøe- 300.000

Disp Vet - 75,000 Dfcp Ace « 5.000,000

Aspirate Vofumβ ■ 5Quf of Pig Serum FkΛd Qte enβe Volume ■ 40tJ of Pfg Serum Serum Fi*

Press 9 £AΣ3PRNC«vtO 0OO) Prβs2 a> RBADPHN nΛ OOI)

Prt .*■ Pre* Fr ^Presι2"^<

NI UscPrI> N2 :- Iιst(Pr2)

Nf-5.24-10³ N2 -5^2' IO³ i:-O..Nt j:*0„N2

Weight! :» 40.77 Welghtt:_* 40.73

Press* :- READPΛ

Press3 :- READPRNCwOla002)

~ »<o> Pr4 :_*.Press

Pr3 y» Press3

N := Ias(Pr )

N3 :• Last<Pr3)

._^3 N4 -6.33 *^• 10

K -6.326'tO

I:»0.. 4. k-0-..N3

Weight* := 40.72

Weightf := 40.76 w^{^}

PressJ :»

prf-Ptesxi 5 :« Us ?r5 N5 -6.3AS-10⁵ cn^O.-W

PΓICB ^:aPrIi2 l3- 30

JJl •^•= ; 712 - JU

JJl * 50

Prld:= ^■ Σ c> - Prlb)._α2

L22-51

Prld = •1.212- tO³

II lb := JU- 13

JJlb * 50

?rle: = Σ ^W*02

L22-51

Pτts = = 773.774

* too*"

Pr2b- :=Pr2_i2+._nι.₅₀

JJl := J22 - J2I JJ3:= J32-J31

JJ2<-50 JJ3 <-50 Pτ2d:« 2 d-Pώb^ Pr3d:. ∑ (2-Pr3b)_I21

122-51 Ϊ22-51

Pr2d « 1.24M0⁵ Pr3d- 1.233 «10³

JJIb :a J24 - J23 JJ3b:' J34- J33

Prle =735.531 Fr3e =777.303

^Pr5c

Pr5« -771.073

A?ggNPI% A

sect ion — 1

Aβp Vβl - 10,000 Atp. Ac « 300.000

Ofsp Ve( - 75,000 Qfsp Ace - 5.000,000

Aspfracβ Vαlυm a - 5Cd of Pta Sβram wfth a breaic In fbdi contactft-ring m≠r Sβn Oiβpensβ Volume - 40d of Ptø Seiυm avalabfo

Pressfi ^ READPRNCwOldOOO)

<0> Press7 :» READPHNCVWOldOOl)

Fro :*PΓSSJ6 Α7 sm^'Bseni** N7 » Isst(Pt7) 8 -5.96I^«I0³ N7 -&02»I0³ i=»0..NS iv:»20-Nδ j*0-N7

WetghtfS :- .05

We£^Etf7:»-.Q4

PressS » READPKNCwtQ I fl02 Presi9 ϊ» P^ADPRN< «Q1<2Q 3)

Pr9 :*Press9 ,<o> N3 :* .ast(PcS) N9 := Lsst<Pr9) N3«5.U2-10³ N9 -6.135-10³ l:=0..N9 Weigh-S :* .23

Weight* :=» 23.63

PresslO *-- EADPRNCw0tdD04)

PrlO^Preslo"⁵³* NIO := Lxr^PrlO) N10 -6. 8-10³ cχ:»0..NLO W g llO :* 10.75

6:= J62- J<51

JJ6 *> 50 PrS := ^ (l-PtΛ)^

Ϊ22-51

PrSd =737.253

JJ6b:- J64- J63

JJob * 50

Pτ5« •^■* ^ ^nz

122=51

Pτ*<- =233.231

Pr9d -913.611

119b π 194 -19$

119b + 50 Pr9β:» 2 PtScjjj

Ϊ22-51

Prrød- 1.033*10*

JllO :» 110*1 - X103 JTlGo + SO JJlOb - 133 PrlOβ:= ^ Ptiqβ_jjj i22-5t PrlOβ -399.219

APPKNDTX A

Section i

c_βtd - 1-263-10* p^ _{β 1}.2S .t<j^J?

U3+50 JSt+S^Q

i22»5l 122-51

Pre3d - 1.232* tO³ P»4d =* t.272^«t0³

_Ptβ d-I.07*t^Q3 ?«^•«^•" ∑ (l-ftβO a,

£22-51

Prc8d-I.2I2*t0³

119 •_*• 50

Prrfd^:. J -Pr^βSh}^ J7₁₀.5₀

¹²² "^5t PratCd:* ^ (2-PrN*l0b _l22

?r:9 - 1.099- lO³ Ϊ22-5I

PrslOd-I.tδNS-IQ³

APPENDIX A

Section 5

APPENDIX A

Section S.

λPPSNDIX A

Section 7

Claims

WHAT IS CLAIMED IS;

1. A method of handling a fluid, the method comprising the steps of:

(a) energizing a pump fluidly associated with a fluid conveying conduit to move fluid with respect to the fluid conveying conduit; (b) energizing a pressure transducer fluidly associated with the fluid conveying conduit to continuously monitor pressure within the fluid conveying conduit;

(c) continuously monitoring pressure within the fluid conveying conduit with the pressure transducer; and (d) determining whether movement of fluid with respect to the fluid conveying conduit is intended or unintended based on the continuously monitored pressure within the fluid conveying conduit .

2. A method of handling a fluid as defined in claim 1 wherein the continuously monitoring step (c) occurs at least during operation of the pump.

3. A method of handling a fluid as defined in claim 1 wherein the continuously monitoring step (c) occurs during at least one of before operation of the pump, during operation of the pump and after operation of the pump.