Multi-point filtering liquid level detection method and apparatus

Abstract

A method for detecting a liquid level in a well of a container, the method including: determining an expected level of liquid in a well of the container; measuring and recording the measured level of liquid in the well; varying the liquid level in the well based on an expected amount of liquid to be added or removed; and calculating a next expected liquid level based at least in part on multi-point filtering. In another aspect, an apparatus for performing the method is provided.

Description

[Technical field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority to U.S. Provisional Patent Application No. 63 / 088,934, entitled “MULTI-POINT FILTERING LIQUID LEVEL DETECTION METHODS AND APPARATUS,” filed October 7, 2020, the disclosure of which is incorporated by reference in its entirety for all purposes.

[0002] The present invention relates generally to methods and devices configured to determine a liquid level associated with the aspiration of a liquid, such as when a reagent liquid is aspirated from a well of a container. [Background technology]

[0003] In diagnostic laboratory tests to measure various chemical constituents of body fluids obtained from patients, such as whole blood, serum, plasma, interstitial fluid, cerebrospinal fluid, and urine, fully automated analyzers can reduce the number of skilled technicians required to perform the analysis, improve the accuracy of the tests, and reduce the cost per test.

[0004] Typically, an automated analyzer includes an automated aspirator-dispensing device configured to aspirate liquid (e.g., liquid reagent) from a well of a container and dispense the liquid into a reaction vessel (e.g., a cuvette). The aspirator-dispensing device typically includes a pipette (also referred to herein as a "probe") attached to a robotic mechanism, such as a robotic gantry or arm, to enable defined movement of the probe to perform the aspirator-dispensing function and to enable transfer of liquid (e.g., reagent) into the reaction vessel. Separate probes may be used for sample and reagent aspirator-dispensing operations to avoid cross-contamination.

[0005] During an aspiration operation, a robotic mechanism, which may be under the control of a controller, may position a probe above a well of a reagent container and then lower the probe into the well until the probe is partially submerged in the liquid reagent to a desired well depth. The depth to which the probe is lowered below the upper liquid level in the well is referred to as the "well depth (WD)." A pump or other aspiration device is then activated to draw (aspirate) a portion of the reagent liquid from the container into the interior of the probe. The probe is then raised (retracted) out of the container with the reagent liquid in the probe so that the reagent liquid can be transferred to a reaction vessel for testing.

[0006] To aid in the aspiration of liquid reagent in a well, it may be desirable to determine the location of the top level of the reagent liquid contained in the well. For example, knowing the location of the top level of the liquid reagent may help to avoid aspiration of air. Similarly, knowing the location of the top level of the liquid reagent in a well may help to position a probe at the appropriate well depth. Typically, the top level is located by using a probe incorporating a capacitive sensor. Upon contacting the liquid level, a change in an electrical signal is detected by a capacitance measurement circuit. However, in some cases, the measurement may be erroneous. For example, the liquid level detection may be triggered above the top level, or in some cases, may not be triggered at all. Summary of the Invention [Problem to be solved by the invention]

[0007] Therefore, there is a need for a method and apparatus for more accurately determining the position of the upper liquid level. [Means for solving the problem]

[0008] Detection of liquid levels in reagent supply containers is used in many types of diagnostic devices, such as clinical chemistry analyzers and immunoassay instruments (each of which is referred to herein as an "analyzer"). For example, liquid level detection can be used to detect the top level of a reagent liquid in a well. Although the examples provided herein relate to detecting the level of a reagent liquid in a well of a reagent container, such detection methods and devices may also be used to detect the level (top position) of other liquids, such as process water (e.g., deionized water), waste liquids, or even cleaning liquids.

[0009] In some embodiments, a capacitive liquid level detection system is used to determine the liquid level and position the probe in the liquid based on where the surface liquid level is detected by the capacitive liquid level detection system. However, such capacitive liquid level detection systems are prone to errors, such as when the capacitive liquid level detection is falsely activated due to the presence of air bubbles or static charges, or fails to activate when it should. In some embodiments, liquid level detection using a capacitive sensor integrated into a probe that is coupled to and movable by a robotic transfer arm can result in inaccurate triggering. For example, the robot's positioning system that causes the probe movement can malfunction due to position sensor errors, broken wires, or other component malfunctions. This can result in tests being stopped and possibly rescheduled, resulting in wasted time. Furthermore, reagent containers, such as reagent packs containing multiple wells, can be falsely removed due to erroneous liquid level readings, leading to the disposal of reagent packs that still contain usable reagent. Thus, such erroneous liquid level readings can increase reagent costs by initiating premature replacement of reagent packs.

[0010] The accuracy of the detection sensor may be affected by many additional factors, including hardware malfunctions, environmental conditions, or vessel anomalies. Capacitive liquid level detection systems may be suitable for a wide range of applications, but are susceptible to such influences. When used for aspiration, erroneous liquid level detection signals may cause the probe to move out of position for proper aspiration, resulting in inaccurate liquid transfer, such as by aspirating air or bubbles, or may be inserted too deeply into the well, causing excessive contamination of the exterior of the probe with reagent liquid. Such excessive contamination of the exterior of the probe is referred to herein as "carry-over." Carry-over may drip off the probe, but is typically washed off the exterior of the probe in a wash station of the analyzer prior to dispensing of reagent liquid.

[0011] In some cases, the liquid level has no predictability, i.e., is essentially random. However, in many cases, such as the liquid level in a reagent container, the liquid level has some predictability based on the previous history of the surface liquid level position. In these cases, an improved method of liquid level detection is provided. The improved liquid level detection method uses historical knowledge of the liquid level in the well and an estimate of the amount removed by each aspiration over a period of time to determine the expected liquid level that will occur before each aspiration. In particular, the method uses the actual measured surface position value and a multipoint filtering technique to determine the current liquid level. For example, the multipoint filtering technique can correct the expected liquid level based on the actual measured liquid level measurement of the liquid surface and a correction factor based on the gain value, using the proportional control of the liquid level sensor and the normalized liquid level output as feedback to generate the next expected liquid level with a specific correction.

[0012] In some embodiments, this allows filtering of the actual level output readings of the level sensor so that erroneous readings can be detected and minimized, thereby minimizing or preventing erroneous responses. In some embodiments, resilience to inaccurate level sensor readings can be further improved by providing a pre-filter that removes obviously erroneous level values ​​(outliers) before the multi-point filter is applied.

[0013] If the method determines from time to time that an erroneous signal or unexpected change in liquid level has occurred (as determined by monitoring the actual liquid level), the next predicted liquid level predicted by the probe is adjusted (e.g., downward) so that the next aspiration may occur within the liquid, allowing normal aspiration operation to continue. In some embodiments, a correction limit can be used to limit the adjustment (deviation) of the next liquid level from the theoretical liquid level.

[0014] Although embodiments of the methods and devices disclosed herein can be used to aspirate reagent liquid from a reagent container, the methods and devices described herein may also be used to detect and / or track liquid level in any type of container where changes in liquid level are semi-predictable, such as when changes in liquid level are based on the amount of liquid used in the test.

[0015] Thus, in a first embodiment, a method of detecting a liquid level in a well is provided, the method comprising determining an expected level of liquid in the well, lowering a probe into the well containing the liquid to a well depth, measuring and recording the measured level of the liquid from the descent, and calculating a next expected level based at least in part on multi-point filtering.

[0016] According to another embodiment, a liquid level detection device is provided, the liquid level detection device including a liquid level sensor configured to obtain a liquid level measurement of a liquid in a well, and a processor configured to receive the liquid level measurement and calculate a next expected liquid level according to a multi-point filtering module. In some embodiments, the probe is configured to aspirate the liquid.

[0017] In yet another method embodiment, a method of detecting a liquid level in a well is provided, the method including determining an expected level of liquid in a well of a container, measuring and recording the measured level of liquid in the well, varying the level of liquid in the well based on an expected amount of liquid to be added or removed, e.g., based on testing, and calculating a next expected level based at least in part on multi-point filtering.

[0018] Further aspects, configurations, and advantages of the present disclosure may become readily apparent from the following detailed description, which illustrates several exemplary embodiments, including the best mode contemplated for carrying out the present invention. The present disclosure may be capable of other different embodiments, and its several details may be modified in various respects, all without departing from the scope of the present disclosure. Thus, the drawings and description are to be regarded as illustrative and not restrictive. The present disclosure is intended to cover all modifications, equivalents, and alternatives that fall within the scope of the claims and their equivalents.

[0019] The present disclosure will be better understood by reference to the detailed description in conjunction with the following drawings, in which: The drawings are not necessarily to scale. Like numerals are used herein to refer to like elements. [Brief description of the drawings]

[0020] [Figure 1] 1 is a schematic diagram of a suction device including liquid level measurement according to an embodiment of the present disclosure. [Diagram 2] 1 is a decision flow diagram illustrating a method for detecting a liquid level in a well of a container according to an embodiment of the present disclosure. [Figure 3A] 1 is a decision flow diagram illustrating a method for calculating a next predicted level using multi-point filtering and error correction according to an embodiment of the present disclosure. [Figure 3B] 1 is a flow chart illustrating a method for multi-point filtering according to an embodiment of the present disclosure. [Figure 4A] 1A-1C are plots of data illustrating a method of multi-point filtering according to an embodiment of the present disclosure. [Figure 4B] 1A-1C are plots of data illustrating a method of multi-point filtering according to an embodiment of the present disclosure. [Figure 4C] 1A-1C are plots of data illustrating a method of multi-point filtering according to an embodiment of the present disclosure. [Figure 4D] 1A-1C are plots of data illustrating a method of multi-point filtering according to an embodiment of the present disclosure. [Figure 5A] FIG. 13 illustrates a graphical plot of liquid level over successive aspirations, according to an embodiment of the present disclosure. [Figure 5B] 1 is a schematic diagram of a portion of an aspiration device including a liquid level measurement portion inserted into a well of a bulk reagent container according to an embodiment of the present disclosure. [Figure 6] 1 is a flow chart illustrating a method for detecting the position of an upper liquid level of a liquid in a well, according to an embodiment. [Figure 7] 13 is another flow chart illustrating a method for detecting a level of liquid in a well of a container according to an embodiment of the present disclosure. [Figure 8] 13 is another flow chart illustrating a method for detecting a liquid level in a well, according to an embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0021] In view of the above problems, a need remains to accurately determine the location of an upper liquid level, such as in liquid aspiration of liquid reagents used as part of the analysis of a biological liquid by an analyzer to determine the presence and concentration of an analyte or other component of interest in the biological liquid. In particular, the devices and methods described herein find utility for use in analyzers configured to perform analyte measurements, assays, immunoassays, or other tests in which biological liquids and liquid reagents are aspirated and dispensed and where it is desirable to have an accurate metric of liquid level.

[0022] In one aspect, embodiments of the present disclosure provide methods and apparatus that can substantially reduce errors in determining the top liquid level in a container of liquid, such as a reagent liquid. In particular, embodiments of the present disclosure can accurately determine the liquid level by filtering normalized liquid level readings with a multi-point filter. Although the present disclosure uses the example of detecting the position of the top liquid level of a liquid reagent, the present disclosure may also be used and is applicable to accurately determine the position of other liquids used in analyzers, such as process water, waste liquid in a waste receptacle that must be emptied from time to time, and cleaning fluids used to wash pipettes and / or cuvettes.

[0023] In particular, the inventors have found that in an aspiration device, including an aspiration pump, a probe, and a robotic device for moving the probe, erroneous measurements may be taken that then adversely affect the next liquid level, resulting in improper calculation of the top liquid level. For example, it is desirable to position the probe a few millimeters (e.g., 2 mm to 6 mm) below the top surface of the reagent liquid to minimize carryout. However, if the measured liquid level is incorrect, e.g., calculated to be too low compared to the actual position, the probe may be instructed to dive too deep into the well, resulting in an excessive amount of carryout. Furthermore, the controller may be instructed to remove the liquid reagent too quickly. If the calculated liquid level is too high compared to the actual position, the probe may aspirate air or bubbles. It is therefore desirable to obtain accurate liquid level accounting to ensure proper operation over multiple aspirations and complete use of the reagent in the well.

[0024] Thus, the methods and apparatus herein, according to embodiments, improve the accuracy of liquid level detection. These and other aspects and configurations of the embodiments of the methods and apparatus are described with reference to Figures 1-8 herein.

[0025] Referring to FIG. 1, a first embodiment of an aspirating device 100 including a liquid level detection device 101 according to the present disclosure is shown. The aspirating device 100 may include any suitable robot 102 configured to move a probe 104 as needed to aspirate a liquid, such as a liquid reagent 107R. The liquid reagent 107R may also be dispensed by the aspirating device 100. The robot 102 is adapted to effect movement of the probe 104 in one or more coordinate directions, e.g., X, Y (into and out of the plane of the paper) and / or Z. The robot 102 may include a robotic member 105 (e.g., a robotic arm, boom, frame, etc.) to which the probe 104 may be attached for movement. The robotic member 105, which may be a robotic arm as shown, may be rotated about a fixed axis 112A by the action of a rotational actuator 112 to provide horizontal movement capability, e.g., in the XY plane. Optionally or additionally, vertical motion of the probe 104 along a vertical Z-axis can be imparted by operation of a suitable vertical actuator 109, such as a linear actuator of the robot 102, which can be coupled to the probe 104. The vertical actuator 109 can be operable to lower the probe 104 into and raise the probe 104 from a container 107 (e.g., a reagent container) having one or more wells 107W containing a quantity of liquid reagent 107R, at least a portion of which is aspirated and transferred and dispensed into the reaction vessel 113 together with a sample 108S of biological liquid contained in a sample container 108. The sample 108S can be blood, serum, plasma, cerebrospinal fluid, spinal fluid, interstitial fluid, urine, etc. Other liquids can also be aspirated and dispensed. Additionally or optionally, an actuator 106, such as a linear actuator, can be provided to impart motion along the X-axis. Each of the actuators 106, 109, 112 may be suitably actuated under the control of a position control module 114 of a controller 115 to impart a desired motion to the probe 104 in one, two, or three dimensional space.

[0026] Each of the actuators 106, 109, 112 may be configured and operable to move the probe 104 from the container 107 to the reaction vessel 113 (e.g., a cuvette). Using another probe (not shown) or the same probe with intervening cleaning, the sample 108S may be aspirated from the sample container 108 and dispensed into the reaction vessel 113. An aspirating device 100 may be provided to aspirate a desired liquid (e.g., liquid reagent 107R) into the interior of the probe 104. The amount aspirated may be up to about 100 μL, but in some embodiments may be less than about 25 μL.

[0027] The aspirating device 100 may include a pump 122, such as a piston-type pump that may be driven by a suitable motor (not shown). Other types of pumps may be used. The pump 122 is configured to aspirate the liquid reagent 107R into the probe 104 at an aspiration rate for a predetermined period of time to generate a pressure signal in the line 134 from the pressure sensor 130. The aspiration rate achievable by the aspirating device 100 ranges from about 20 microliters per second to about 500 microliters per second. The pump 122 may be fluidly coupled to the probe 104 by a flexible tube 125, such as a section of hollow Teflon tubing or other suitably flexible conduit. The flexible tube 125 and the probe 104 may be filled with a supporting liquid 126 (e.g., purified water) to provide suitable liquid support and generate a suitable vacuum pressure to aspirate a liquid (e.g., liquid reagent 107R) into the internal passage of the probe 104 and transfer it to the reaction vessel 113.

[0028] The aspiration control module 128 of the controller 115 may be configured and operable to control the pump 122 to draw (e.g., aspirate) a desired amount of liquid (e.g., liquid reagent 107R) into the interior of the probe 104. The aspiration control module 128 of the controller 115 may also control dispensing operations performed by the aspiration device 100. The aspiration device 100 may include other conventional components, such as one or more valves, accumulators, dispensers, or other hydraulic components (not shown) that provide for the aspiration and dispensing of liquid. Any suitable device for aspirating liquid into the probe 104 may be used. For example, aspiration dispensing systems that may be used are described, for example, in U.S. Pat. Nos. 7,867,769; 7,634,378; 7,477,997; 7,186,378; 7,150,190; and 6,370,942.

[0029] According to embodiments, a full cycle of aspiration of a liquid (e.g., liquid reagent 107R) can be completed in less than about 300 ms, or even less than about 200 ms, or in some embodiments, even less than about 100 ms. In some embodiments, the aspiration cycle can be completed in about 40 ms to about 200 ms. Other aspiration durations may be used.

[0030] According to one or more embodiments, the aspiration verification module 129 can also be used to verify the validity of liquid aspiration. For example, the aspiration verification module 129 can determine that air is present in the probe 104 during aspiration by determining whether a low viscosity error is present. A low viscosity error is logged when aspiration is performed and a predefined parameter of a certain pressure is not met. A low viscosity error indicates that air is aspirated rather than liquid. In another example, the aspiration verification module 129 can determine that a clog or other obstruction is present in the probe 104.

[0031] More particularly, the suction pressure in the flexible tube 125 may be measured by a pressure sensor 130. The pressure sensor 130 is operable and configured to provide a raw pressure signal in a line 134. The pressure sensor 130 may be positioned at any suitable location along the flexible tube 125 or adjacent the pump 122 or the probe 104. As is conventional, a sensor conditioning unit may be provided to condition the raw pressure signal in the line 134. The sensor conditioning unit may include, for example, a suitable amplifier, an A / D converter, and a suitable anti-aliasing filter.

[0032] According to an embodiment of the present disclosure, as described with reference to FIG. 1 and FIG. 2, the aspirating device 100 further includes a liquid level detection module 131. The liquid level detection module 131 is configured to accurately determine the liquid level of the upper liquid level 107T of the liquid reagent 107R in the well 107W. The liquid level detection module 131 may include a liquid level sensor. The liquid level sensor 133 may be, for example, a capacitive sensor integrated with the probe 104, and when the probe tip contacts the upper liquid level 107T of the liquid reagent 107R, an electrical signal is generated, which is conveyed by the line 135 and differentiated by the liquid level circuit 136 (e.g., a capacitive liquid level circuit) of the liquid level detection module 131. Capacitive liquid level sensors and capacitive liquid level circuits are described, for example, in U.S. Pat. Nos. 4,977,786, 5,550,059, 7,150,190, and 9,863,905. However, it should be understood that other types of level sensors and level circuits may be used, such as optical, ultrasonic, conductivity, or resistive sensors.

[0033] The level detection module 131 further includes a multipoint level filter module 138. The multipoint level filter module 138 may include a method 200 for retrieving multiple level readings from the memory 140 and processing them, such as in the processor 142 of the controller 115. For example, in some embodiments, a currently measured level value and two or more previously obtained level measurements may be used. In particular, the multipoint level filter module 138 may include a digital filter where filtering is performed in software within the controller 115.

[0034] The operations and filtering performed by the multipoint level filter module 138 will now be described with reference to FIG. 2. FIG. 2 shows a flow chart of the various operations performed by the method 200 when determining the level of a well 107W containing a liquid reagent 107R. The method 200 begins when the controller 115 determines that an aspiration from the well 107W is to occur, for example when a sequenced test is initiated in an analyzer that has received a sample for testing. The method 200 can begin with the start of an aspiration sequence in block 202 and determine whether the aspiration is the first aspiration in block 204. If the answer is yes, the well is assumed to be full and the expected well level (synonymous with the upper level position) is assigned a theoretical well level in block 206. The theoretical level is the level that the particular reagent 107R contains from the factory, i.e. the factory fill level. This value is usually quite tightly controlled.

[0035] According to method 200, if the answer in block 204 is no, then in block 208 the predicted well level is assigned the last predicted well level stored in memory 140 after filtering according to method 200. Thus, in either case, yes or no, the next predicted well level is provided at either the theoretical well level or the last predicted well level. Next, in block 210, the probe 104 is lowered to the well depth WD by the movement of the robot 102. In operation, the robot 102 can position the probe 104 above the vessel 107 during aspiration and lower the probe 104. The lowering into the vessel 107 can be caused by the movement of the vertical actuator 109 under the control of the position control module 114 until the probe 104 reaches the desired well depth WD in the vessel 107.

[0036] When at well depth WD, method 200 can optionally check to see if well 107W is empty. Empty is determined to be, for example, an expected liquid level of a few millimeters above the bottom of vessel 107. Thus, method 200 determines whether well 107W is empty at block 212. If well 107 is empty (yes), an empty error can be issued to the operator at block 214 so that aspiration from a new well 107W can be initiated or vessel 107 can be replaced. If well 107 is not empty (no), the actual liquid level is measured and recorded at block 216. The actual liquid level is the liquid level measured when probe 104 is lowered downward. For example, when probe 104 touches upper liquid level 107S, liquid level circuit 136 detects a change, such as a change in capacitance from liquid level sensor 133. This liquid level is recorded in memory 140 as the actual liquid level.

[0037] Next, at block 218, a quantity of liquid reagent 107R is aspirated into the probe 104 to perform the test. The aspirator 100 may be operated by signals from the aspiration control module 128 to draw a predetermined quantity of liquid reagent 107R for a particular test into the internal passage of the probe 104. As the pump 122 operates, it attempts to draw down (aspirate) the level of liquid reagent 107R in the container 107. Once it is determined that the desired quantity of liquid 107R, as measured by the aspiration control module 128, has been received into the probe 104, the pump 122 may be stopped to prevent further aspiration of liquid reagent 107R. This may be determined by a suitable feedback sensor (not shown) on the motor driving the pump 122, and thus providing feedback regarding the position of the pump 122. Any suitable position feedback may be provided.

[0038] In some cases, it may be desirable to further determine that the aspiration was successful, such as by performing an aspiration quality verification at block 220. During the aspiration process, a representative raw aspiration pressure in line 134 may be measured by pressure sensor 130 (e.g., measured aspiration pressure). This raw measured pressure in line 134 may be converted (A / D), conditioned, and filtered by aspiration verification module 129 to provide a conditioned pressure signal. By examining this conditioned pressure signal, the aspiration quality, in particular whether air was aspirated, may be accurately determined. Aspiration verification may be accomplished by any known method, such as those described in U.S. Pat. No. 7,477,997, U.S. Pat. No. 7,867,789, U.S. Pat. No. 7,926,325, U.S. Patent Application Publication No. 2015 / 0276534, and U.S. Patent Application Publication No. 2016 / 0258972.

[0039] For example, in some embodiments, to verify the effectiveness (e.g., quality) of aspiration at block 220, the conditioned pressure signal is compared at evaluation points of the conditioned pressure trace throughout the aspiration. At the desired evaluation points, the conditioned pressure signal is compared to a predetermined threshold. If the conditioned pressure signal is higher than the predetermined threshold, the aspiration may be considered successful. If the conditioned pressure signal has a value lower than the predetermined threshold, the aspiration may be considered unsuccessful or incomplete. This may indicate that some air has been aspirated.

[0040] 2, after aspiration in block 218, the aspiration device 100 looks for a liquid level sense transition (LLS transition) in block 222. Determining whether there is an LLS transition in block 222 involves examining the signal in line 135 by the level circuit 136 of the level detection module 131 of FIG. 1. For example, as the probe 104 is lowered, a voltage spike higher than a predetermined voltage value detected by the level circuit 136 can indicate a positive confirmation that an LLS transition is present (yes). If an LLS transition is not detected in block 222 (no), the aspiration quality verification from block 220 is used to determine whether low viscosity occurred in block 224. Low viscosity in block 224 means that the viscosity was much lower than expected, such as a pressure reading in line 134 that is lower than an expected value, such as below a predetermined threshold. An indication of low viscosity (yes) in block 224 indicates not only that an LLS transition was not detected, but also that air may have been aspirated. In this case, the test is stopped at block 226 and the next predicted level is assigned the current predicted level stored in memory 140 -ΔH at block 230, where ΔH is the theoretical change in height of the level based on the theoretical amount of reagent fluid 107 L needed for the test currently being performed.

[0041] If no LLS transition is found in block 222 and no low viscosity is detected in block 224, it is assumed that an aspiration has occurred but the level sensor 133 or the level circuit 136 or line 135 has malfunctioned. In this case, the LL detection event may be logged as an error in block 228. In this case, however, the correction is zeroed and the next expected level is calculated based on multipoint filtering in block 232, as will be more fully described in FIG. 3. Following the calculation of the next expected level based on multipoint filtering in block 232, the next expected level is recorded in memory 140 in block 234. Method 200 then continues from block 208 for the next aspiration for a new test.

[0042] The details of the multi-point filtering of block 232 used by method 200 will now be described with reference to Figure 3. First, the actual measured liquid level is obtained from memory 140 in block 336. This measured liquid level was recorded in block 216 (Figure 2) from the descent to well depth WD in block 210. The position of the probe tip of probe 104 from position control module 114 was previously recorded when the upper liquid level 107S of liquid reagent 107R was sensed by level sensor 133 and level circuit 136 of level detection module 131.

[0043] Pre-filtering may then optionally be applied to determine whether the actual measured liquid level makes sense. Pre-filtering in block 338 includes pre-filtering to remove outliers before the application of multi-point filtering. In particular, in block 338 it is determined whether the measured liquid level is within a predetermined pre-filter window set in block 340. As best shown in FIG. 5A, in block 340 the pre-filter window is set to the measured LL. nThe pre-filter window may be established and set as a preset +Y value higher than the measured liquid level at block 338. Optionally, the preset value may be provided by, for example, summing all theoretical changes in ΔH over previous aspirations for well 107W and allowing for a + / - tolerance therefor. Other means for pre-setting the pre-filter window at block 340 may be used. n If it is found not to be within the pre-filter window set in block 340 (No), then a zero offset is assigned in block 342 as will be described later in this specification. n If it is determined in block 340 that the measured liquid level is within the pre-filter window (yes), then in block 344, multi-point filtering is used to filter the measured liquid level. n is filtered.

[0044] The multi-point filtering in block 344 is further described with respect to Figures 4A-4D. A liquid level is abbreviated herein as LL. Similarly, a plurality of liquid levels is abbreviated herein as LLs. In particular, Figure 4A shows the theoretical LL value of well 107W in dotted trace 440 and the actual measured LL value of well 107W across a series of tests in solid trace 442. Figure 4B shows a data set for the normalized theoretical LL value in well 107W in dotted trace 440N. Similarly, the normalized measured LL data of well 107W across multiple tests is shown as solid trace 442N. The entire data set is shown. However, the method calculates the normalized LL for each new test data point point-wise. The LL value is normalized for each test n according to a normalization formula as follows: Normalized LL n = measurement LL n -Theory LL n

[0045] Figure 4C shows a dotted trace 440NF showing the normalized and filtered theoretical liquid levels over the series of tests. Figure 4C also shows a solid trace 442NF of the normalized and filtered data set of the normalized and filtered actual measurements over the series of tests.

[0046] More specifically, the multi-point filtering of the inspection data is best illustrated in Figure 3B. As shown in Figure 3B, the multi-point filtering of block 344 involves filtering the measurement LL according to the above equations in block 346. n A running record of the theoretical LL for each test can be maintained in memory 140. Next, at block 348, the normalized and filtered LL is n is calculated. The normalized and filtered LL n can be based on a normalized data point n and a plurality of previous m-1 normalized data points, such that m is 3 or more, 10 or more, 15 or more normalized measured LL data points. The m data points are obtained consecutively with point n. In some embodiments, the multi-point filter is a median filter. In particular, the median filter does not use the n value directly, but uses the median value of all previous m normalized liquid level measurements, including the most recent n value, as the new median value. This new median value is the normalized filtered LL value, as shown in FIG. 4C. Thus, filtering is effective to minimize the effect of erroneous LL measurements. As shown, data point n is at the edge of a set of median data points from previous tests. The median of m data points including the data point is shown as a set of a fixed number m of data points, where m can be set by experience. In some embodiments, m can be, for example, 3 to 40, or even 5 to 20. In some embodiments, rather than a median filter, a mode filter using the mode value of the m data points may be used, or a mean filter using the average of the m data points may be used, although a median filter is most desirable.

[0047] Next, in block 350, the normalized and filtered LL n is denormalized and filtered LL n The denormalization is performed according to the following formula: Filtered LL n = filtered and normalized LL n +Theory LL n

[0048] 4D shows the non-normalized data by adding or subtracting theoretical values. Thus, dotted trace 440F shows the non-normalized theoretical LL for well 107W across a series of tests. Similarly, solid trace 442F shows the non-normalized filtered measured LL for well 107W across a series of tests.

[0049] 3A, after the multi-point filtering is completed in block 344, the method may determine the error in block 352. The error determination may be calculated as follows: Error=FLL n - Prediction LL n

[0050] Once an error has been determined, the error correction to be applied may be determined in block 354. The error correction may be determined as follows. Error Compensation = Gain x Error

[0051] The gain may be a constant value determined empirically. The gain may have a value of, for example, 0.05 to 1.5, or even 0.05 to 0.5. When applied to significantly different vessel systems, such as when monitoring liquid level in a bulk vessel, both the gain and the number of points in the central filter may vary outside the above ranges, since they are functions of the number of data points relative to the rate of change of the liquid level.

[0052] Next, the next predicted LL n+1 The value is calculated. The next expected LL n+1 can be calculated as follows:

[0053] Next Prediction LL n+1 = Expected LL n - (Change in Inspection LL + Change in Carry-Out LL) + Error Correction Therefore, the next prediction LL n+1 The value may be based at least in part on multi-point filtering, but may additionally be based on error correction. The change in test LL is based on the drop in LL based on the amount of liquid reagent 107R being aspirated for test n. Carryout in the general sense is the accounting for the expected loss of LL due to reagent carryout on the probe 104, or even loss of LL due to evaporation. In the case of probe carryout, the change in carryout LL (carryout) is based on the amount of change in LL due to carryout of liquid reagent 107R outside the probe 104, which may be an empirically determined set value based on moving the probe 104 to the well depth WD (see FIG. 1), drawing and washing multiple times, and then determining the average carryout per aspiration. The evaporative carryout may also be set experimentally. Thus, in some embodiments, the carryout may include both reagent carryout and / or evaporative carryout. Thus, the next expected LL n+1 The calculation of the value can take into account carry-outs.

[0054] Next Prediction LL n+1 Once the value is calculated, at block 358 the next expected LL n+1 The value can be checked against the accommodation limit. As shown in FIG. 4D, the accommodation limit 440L can be set to a value for n that is X mm lower (e.g., 5-10 mm lower) than the theoretical value 440F. Thus, in block 358, the next predicted L n+1 If it is less than the adjustment limit of 440L (yes), then the next predicted LL n+1 is set equal to the n+1 adjustment limit. This limit is then recorded in block 234 of FIG. 2. Alternatively, the next expected LL is set in block 358. n+1If the adjustment limit is equal to or greater than 440L (NO), the next predicted LL n+1 is recorded. Block 234 is reproduced in Figure 3A for illustrative purposes.

[0055] FIG. 5A graphically illustrates a plot 570 of LL over successive aspirations and what happens when certain anomalies occur. As can be seen, actual values ​​571 are shown as diamonds. The normalized filtered central trace 542 is shown as a solid line. An upper prefilter 572 is shown, which has a prefilter limit that can be set based on experience, for example, at a value +Y higher than the normalized filtered LL value for testing. Thus, points such as LL value 574 are ignored by the central filter, and the expected LL is simply adjusted downward by ΔH+carryout. The upper prefilter 572 mitigates large LL detection noise or air bubbles due to drawing the expected LL from the liquid reagent 107R.

[0056] Although a single upper level of the prefilter limit set at +Y is shown in FIG. 5A, the upper prefilter 572 can have multiple prefilter limit levels, for example, allowing a wider range of +Y prefilter limits for multiple aspirations after initiating access to the well 107W at 580 to allow for larger fill changes in the well 107W. Similarly, a wider +Y prefilter limit value can be allowed for multiple aspirations after an adjustment is made by air aspiration, such as at point 582. Thus, the upper prefilter 572 includes a prefilter limit that can be adjusted based on the state of the vessel 107. For example, when the vessel 107 is first accessed, the state of the vessel can be full, or the state of the vessel can be that air has just been aspirated, as after point 582. An additional prefilter limit (not shown) can be set at a preset position above the bottom of the well 107W.

[0057] Additionally, a normalized lower accommodation limit 578 is shown that prevents erroneous pressure errors from pushing the probe 104 too deep. The normalized expected LL value cannot cross this normalized accommodation limit 578. The accommodation limit can be set to any desired level X, where X may be, for example, 5 mm to 10 mm.

[0058] As can be seen at the start of the test at point 580, there is no previously measured LL value, so the expected LL and the theoretical normalized value are the same. In some embodiments, the central filter can be pre-populated with enough zeros to fill about half of the filter (e.g., 8 zeros for a 15-point central filter). The central filter then does not react until the 9th point, after which the central filter can start to operate and adjust the filtered expected LL value 542. Of course, the central filter can be loaded with fewer zeros to make it start to react after fewer draws. However, this comes at the expense of less noise removal. Optionally, more zeros can be pre-populated into the central filter to make it react after more draws and better at removing noise.

[0059] An event is also shown indicating air aspiration in the LL measurement 582. Once air aspiration is determined, the probe 104 adjusts the next aspiration downward by X / 2 or other appropriate value to a new position so that the tip of the probe 104 is approximately X / 2 below the LL. Also, the prefilter value 572 is reset to a value such as Y higher than the normalized filtered LL. Again, the expected LL and theoretical LL are identical until three data points (LL measurements) are included in the median filter, after which the median filtering can begin to adjust the next expected LL.

[0060] FIG. 5B is a schematic diagram of a portion of the aspirating device 100 including the liquid level measuring portion and liquid level sensor 133 inserted into a well 507W of a bulk container 507. The method 200 described herein is equally applicable, with minor modifications, to the accurate measurement of the liquid level of the bulk liquid 507B contained in the bulk container 507. The method 800 can eliminate errors due to waves or splashing of the bulk liquid 507B due to liquid movement or vibration. Furthermore, the method 800 described herein in FIG. 8 is equally applicable to the accurate measurement of the liquid level of a process liquid (e.g., deionized water) in a process water container, a cleaning liquid in a cleaning liquid container, a waste liquid in a waste liquid container, or any other container whose liquid level changes (e.g., increases or decreases), regardless of whether the liquid is accessed and aspirated or dispensed by the probe 104. In these examples, the method 800 can eliminate errors due to waves or splashing of the bulk liquid 507B due to the inflow or outflow of liquid, such as when the valve 509 is opened or closed or when waste liquid is injected.

[0061] 6, a broad method of the present disclosure is shown. Method 600 includes, at 602, determining an expected level of liquid (e.g., liquid reagent 107R) in a well (e.g., well 107W, 507W). If the test is a first test, the liquid level is a theoretical LL. After at least three readings have been taken, the expected LL is calculated based on the filtered LL stored in memory 140. n+1 The method 600 further includes lowering the probe (e.g., probe 104) into a well (e.g., well 107W, 507W) containing the liquid (e.g., liquid reagent 107R) to a well depth (WD) at block 604. The method 600 further includes lowering the probe (e.g., probe 104) into a well (e.g., well 107W, 507W) containing the liquid (e.g., liquid reagent 107R) to a measured level (e.g., measured LL) at block 606. n ) from the downward movement; and at block 608, determining a next predicted liquid level (e.g., a next predicted LL n+1) for the measurement. However, the calculation may also include error correction, i.e., a combination of multipoint filtering and error correction. Multipoint filtering is described in FIG. 3B. A combination of multipoint filtering and error correction is shown and described with reference to FIG. 3A. In some embodiments, the method 600 includes aspirating a volume of liquid for testing at block 610, and dispensing the volume of liquid into a reaction vessel (e.g., reaction vessel 113) at block 612.

[0062] In some embodiments, bulk liquid 507B can be added to or removed from bulk container 507 without the use of probe 104, for example in the example using bulk container 507 as shown in FIG. 5B. For example, bulk liquid 507B can be added or removed through fixed ports or spouts formed on the side or bottom of bulk container 507. Using fixed ports having, for example, a valve 509 as shown in dotted lines, bulk liquid 507B can be withdrawn from bulk container 507 for use in testing or the like. Similarly, waste liquid from a just completed test can be added. Additionally, wash liquid can be withdrawn, such as by valve 509, for use in testing, for example. The liquid level in such bulk container 507 can be more accurately recorded using methods including multi-point filtering.

[0063] In this example, a LL sensor is included to detect the liquid level LL of the upper liquid level. Any suitable LL sensor can be used. The LL sensor may be a stationary sensing device such as a float on the sensor shaft, a capacitive sensor, or even a precision scale that measures weight, where weight can be measured and converted to obtain the measured LL. In this case, the weight scale is also considered to be a LL sensor. The method 700, 800 can be used to determine the next expected liquid level (e.g., the next expected LL n+1 ) can be accurately determined, thus providing a more accurate estimate of liquid level over time.

[0064] 7, a method 700 for detecting the level of a liquid (e.g., a process liquid, a wash liquid, or a waste liquid) in a well 507W of a container (e.g., bulk container 507) is shown. Method 700 is identical to method 200 described in FIG. 2, except that non-essential aspects have been removed and liquid is removed or added by dispensing into a spout or by adding or removing through a port located on the side or bottom of bulk container 507. In comparison to FIG. 2, identical numbers are used for identical blocks, except for block 718, which denotes a change in the liquid level in a well with an expected change in the liquid level based on testing, etc., as opposed to aspiration by probe 104.

[0065] Thus, broadly, as shown in FIG. 8, the method includes, at block 802, determining an expected level (e.g., expected LL) of a liquid (e.g., bulk liquid 507B) in a well (e.g., well 507W) of a container (e.g., bulk container 507). Method 800 includes, at block 804, measuring and recording the measured level of the liquid (e.g., bulk liquid 507B) in the well (e.g., well 507W). According to method 800, block 806 includes varying the level in the well based on an expected amount of liquid to be added or removed. The expected amount of liquid to be added or removed can be based on the requirements of the test. For example, the test may require a set amount of processing liquid for the test. Similarly, if the liquid is a wash liquid, a set amount of wash liquid can be used for the wash operation. If the liquid is a waste liquid, there may be a set amount of waste liquid for each test, which is the expected amount of waste liquid to be added to well 507W. The test may be a test of a biological liquid as described herein, and a prescribed volume of bulk liquid 507B is used in the test. Thus, if bulk liquid 507B is a processing or cleaning liquid, a prescribed target volume must be used for each test. As in the previous method, method 800 may include, at block 808, determining a next predicted liquid level (e.g., a next predicted LL) based at least in part on the multipoint filtering. n+1), which is illustrated in Figure 3B. Error correction may be used in addition to multipoint filtering, as shown in blocks 352-356 of Figure 3A. Pre-filter windows and adjustment limits may be used, as shown in blocks 340, 338, 342 and blocks 358, 360, respectively.

[0066] While certain exemplary embodiments have been shown, those skilled in the art will recognize that many variations are possible within the scope of this disclosure. It is therefore intended that this disclosure be limited only as indicated by the claims and the equivalents thereof.

Claims

1. 1. A method for detecting a liquid level in a well, comprising: checking in the memory an expected level of liquid in the well; lowering a probe into a well containing liquid to a well depth; measuring and recording a measured level of the liquid from the lowering of the probe; pre-filtering to remove outliers in the measured and recorded liquid levels, including pre-filter limits that are adjusted based on the condition of the container containing the liquid; calculating a next predicted liquid level based at least in part on the multipoint filtering; and The method comprising:

2. The method of claim 1 , wherein calculating the next expected liquid level is further based on applying an error correction.

3. The method of claim 2 , wherein the error correction is equal to the error times the gain.

4. The method of claim 1 , wherein calculating the next expected liquid level further comprises considering carryout.

5. The method of claim 1 , wherein the multipoint filtering is achieved using a median filter.

6. 6. The method of claim 5, wherein the median filter uses a median value derived from three or more measured liquid level values.

7. The method of claim 6, wherein the median filter uses the median of 10 or more consecutively obtained values.

8. 7. The method of claim 6, wherein the median filter uses the median of 15 or more consecutively obtained values.

9. The method of claim 1 , wherein the multipoint filtering is achieved using a mode filter or an average filter.

10. 10. The method of claim 1, including using an adjustment limit to limit adjustment from the next theoretical level for the next predicted level.

11. 10. The method of claim 1, further comprising aspirating a volume of liquid designated for testing prior to the calculation.

12. The method of claim 1 , wherein the liquid in the well is a reagent liquid.

13. The method of claim 1 , wherein the liquid in the well is water, waste liquid, or a wash liquid.

14. The method of claim 1 , comprising performing an aspiration quality verification after aspiration.

15. 15. The method of claim 14, wherein the adjustment comprises adjusting the probe position downward when the aspiration quality verification detects that air is aspirated.

16. 2. The method according to claim 1, wherein the measurement of the measured liquid level is performed during descent.

17. 1. A method for detecting a liquid level in a well, comprising: checking in the memory an expected level of liquid in the well of the container; Measuring and recording a measured level of liquid in the well; pre-filtering to remove outliers in the measured and recorded liquid levels, including a pre-filter limit that is adjusted based on the condition of the container containing the liquid; Varying the liquid level in the well based on an anticipated amount of liquid to be added or removed; calculating a next predicted liquid level based at least in part on the multipoint filtering; and The method comprising:

18. 1. A liquid level sensing device comprising: a liquid level sensor configured to obtain a level measurement of liquid in the well; a processor configured to receive the liquid level measurements, pre-filter to remove outliers based on pre-filter limits that are adjusted based on a condition of the vessel containing the liquid, and calculate a next predicted liquid level according to a multi-point filtering module; The liquid level detection device.

19. 20. The liquid level detection device of claim 18, comprising a probe configured to aspirate liquid.

Images