Method for classifying identification tag on sample tube containing sample and automated laboratory system

Abstract

To provide a method for classifying an identification tag on a sample tube including a sample to be processed in an automatic laboratory system, which provides improvement in processing efficiency of the sample tube in the automatic laboratory system which can be configured for at least one of pre-analysis and sample analysis, and an automatic laboratory system.SOLUTION: A method includes: a classifying module 8 configured to read an identification tag 9 on a sample tube 5 by a classifying reader device 7 of a classifying device 2, and to predict whether identification tag information can be recognized from measured tag data detected by a tag reader device 10 assigned to a laboratory device 4 and indicative of characteristics of the identification tag 9; and classification data being first classification data indicative of predicting, by the classifying module 8, that the identification tag 9 is readable by the tag reader device 10 of at least one of the plurality of laboratory devices 4, and second classification data indicative of predicting that the identification tag is not readable.SELECTED DRAWING: Figure 1

Description

Technical Field

[0002] , , , ,

[0004] , , , , ,

[0003]

[0001] The present disclosure relates to a method for classifying identification tags on a sample tube containing a sample processed in an automated laboratory system, and an automated laboratory system.

Background Art

[0002] Sample tubes containing samples such as samples of body fluids can be processed or handled in an automated laboratory system for at least one of pre-analysis and sample analysis. In order for the laboratory apparatus of the automated laboratory system to correctly handle or process the sample tube, the sample tube is provided with an identification tag, which may have, for example, a barcode, but may be provided with any other data matrix code such as a QR code that can be read by an optical reader.

[0003] The identification tag provided on the sample tube generally uniquely identifies the sample tube. Through normal daily use, the characteristics or conditions of identification tags such as barcodes may suffer degradation including tearing, peeling, discovery, and other deformations. Such deformations may prevent the automated laboratory system from correctly handling or processing the sample tube.

[0004] U.S. Patent No. 10,325,182 discloses a method for classifying the barcode tag state on a sample tube from a plan view image in order to rationalize the handling of sample tubes in an advanced clinical laboratory automation system. The classification of the barcode tag state results in the automatic detection of problematic barcode tags and enables the user to take the steps necessary to correct the problematic barcode tags. A vision system is utilized to perform the automatic classification of the barcode tag state of the sample tube from the plan view image. The classification of the barcode tag state on the sample tube is based on the following factors: region of interest (ROI) extraction and correction method based on sample tube detection; barcode tag state classification method based on overall features uniformly sampled from the corrected ROI; and problematic barcode tag region localization method based on pixel-based feature extraction.

[0005] U.S. Patent Application Publication No. 2018 / 0239936 describes a method for detecting the status of barcode tags on sample tubes using side view images of sample tubes to streamline handling in an automated clinical laboratory system. The status of the tags can be classified into classes, each class being further divided into a list of additional subcategories covering individual characteristics of tag quality. A tube characterization station (TCS) is provided for acquiring side view images. The TCS allows for the simultaneous or near-simultaneous acquisition of three images for each tube, resulting in a 360-degree side view of each tube. Two parallel low-level queues are utilized for status recognition in combination with a tube model extraction queue. Semantic scene information is then integrated into an intermediate-level representation for a final decision on one of the status classes.

[0006] U.S. Patent No. 9,230,140 discloses a barcode verifier comprising: an imaging module for capturing images of a field of view; a memory communicatively coupled to the imaging module and configured to store images and a barcode quality verification program; and a processor communicatively coupled to the memory, configured by the barcode quality verification program to perform the following steps: (i) retrieve an image stored in the memory; (ii) locate a barcode symbol in the stored image; (iii) identify one or more unprinted lines in the barcode symbol; (iv) use the one or more unprinted lines to determine a printer malfunction; and (v) generate a printer malfunction report.

[0007] International Publication No. 2017 / 157784 discloses a receptacle having at least one interior for receiving at least one contents, having a fully or partially cylindrical side that at least partially encloses the at least one interior, and a plurality of sequentially arranged identical machine-readable unique identifier codes enclosing the side. A method is provided for applying a plurality of identical machine-readable unique identifier codes to a receptacle having a fully or partially cylindrical side, wherein the identical identifier codes are printed sequentially so as to enclose the side. Furthermore, a method is disclosed for reading at least one of a plurality of machine-readable unique identifier codes on at least one fully or partially cylindrical side of a receptacle, wherein the receptacle has a plurality of sequentially arranged identical machine-readable unique identifier codes enclosing the side, and at least one unique identifier code is read by a code-reading camera without rotating the receptacle. [Overview of the project]

[0008] The objective is to provide a method for classifying identification tags on sample tubes containing samples to be processed in an automated laboratory system, and an automated laboratory system, which can be configured for at least one of pre-analysis and sample analysis, thereby improving the processing efficiency of sample tubes in the automated laboratory system.

[0009] To solve the problem, a method is provided for classifying identification tags on a sample tube containing a sample to be processed in the automated laboratory system described in claim 1. Furthermore, an automated laboratory system is provided for processing a sample tube containing a sample for at least one of pre-analysis and sample analysis as described in independent claim 12. Further embodiments are disclosed in dependent claims.

[0010] According to one embodiment, a method for classifying identification tags on sample tubes containing samples to be processed in an automated laboratory system, comprising providing sample tubes containing identification tags and samples to be analyzed in an automated laboratory system having multiple laboratory devices, wherein each of the multiple laboratory devices is assigned a tag reader device configured to read the identification tags in order to recognize the identification tag information, further comprising providing a classification module in a classification device that, for each of the multiple laboratory devices, is assigned to the laboratory device and is configured to predict whether the identification tag information can be recognized from measurement tag data indicating the characteristics of the identification tags detected by the tag reader device, and reading the identification tags on the sample tubes with the classification reader device of the classification device, and Therefore, a method is provided which includes: providing measurement tag data of an identification tag that shows the tag characteristics of an identification tag on a sample tube; determining the tag characteristics of an identification tag from the measurement tag data; a classification module receiving the tag characteristics and predicting whether at least one tag reader device among a plurality of laboratory devices can read the identification tag; and providing classification data, wherein the classification data consists of first classification data indicating that the classification module predicts the identification tag is readable by at least one tag reader device among a plurality of laboratory devices, and second classification data different from the first classification data indicating that the classification module predicts the identification tag is not readable by at least one tag reader device among a plurality of laboratory devices.

[0011] In another embodiment, an automated laboratory system is provided for processing sample tubes containing a sample for at least one of pre-analysis and sample analysis, comprising: a plurality of laboratory devices, each assigned a tag reader device configured to read the identification tag to recognize the identification tag information; and a classification device. The classification device is configured to provide classification data in which, for each of the multiple laboratory devices, it predicts whether an identification tag can be recognized from measurement tag data detected by a tag reader device assigned to the laboratory device and indicating the characteristics of the identification tag, reads the identification tag on the sample tube with the classification reader device, thereby providing measurement tag data of the identification tag indicating the tag characteristics of the identification tag on the sample tube, determines the tag characteristics of the identification tag from the measurement tag data, and in the classification module, receives the tag characteristics, predicts whether at least one of the multiple laboratory devices' tag reader devices can read the identification tag, and provides classification data, wherein the classification data is provided in which first classification data indicating that the classification module predicts the identification tag is readable by at least one of the multiple laboratory devices' tag reader devices, and second classification data different from the first classification data indicating that the classification module predicts the identification tag is unreadable by at least one of the multiple laboratory devices' tag reader devices.

[0012] Based on the proposed technology, sample tubes processed or handled in an automated laboratory system are classified as sample tubes having an identification tag that can be read by a tag reader of at least one of the multiple laboratory devices in the automated laboratory system, such that the identification tag information provided by the identification tag can be recognized by at least one of the multiple laboratory devices, or cannot be recognized. Otherwise, the identification tag on the sample tube may be completely unreadable, or unreadable in such a way that the identification tag information can be recognized by the automated laboratory system as necessary for the proper handling or processing of the sample tube. The identification tag may be completely unreadable by at least one of the tag readers assigned to each of the at least one of the laboratory devices. Alternatively, the identification tag may be partially readable by the tag reader device, but the identification tag information may not be at least partially recognizable from the measurement data read for the identification tag. Therefore, even in such a case, the identification may be classified as unreadable.

[0013] The identification tag information that can be recognized from reading the identification tag may include at least one of the following: the sample information of the sample accepted into the sample tube and the sample tube information of the sample tube containing the sample.

[0014] Therefore, before a sample tube is handled or processed in the automated laboratory system, it can be avoided or prevented from being moved to or provided to the laboratory equipment of the automated laboratory system, and the laboratory equipment of the automated laboratory system is also called the equipment of the automated laboratory system if a classification device predicts that the identification tag of such a sample tube cannot be read in a manner that recognizes the identification tag information by a tag reader device assigned to such laboratory equipment (instrument). Before providing a sample tube to the automated laboratory system, or in the initial stages of processing a sample tube in the automated system, the sample tube is classified as to whether or not it has an identification tag that can recognize the identification tag information by reading the identification tag. Such classification is based on the (classification) prediction of a tag reader device provided in the automated laboratory system, each of which is assigned to at least one of the laboratory equipment (instruments) of the automated laboratory system. A classification device having a classification module is configured to provide each tag reader device with a prediction of whether or not the identification tag of the sample tube can be properly read in order to recognize the identification tag information.

[0015] Based on such classification, any further handling or processing of the sample tube can be performed or controlled depending on the result of the classification. In embodiments, it may even be provided that the sample tube is not transported to or provided to the automated laboratory system because it is known that the identification tag on the sample tube is not readable by at least one of the tag reader devices provided in the automated laboratory system. In some cases, the identification tag cannot be read at all by the tag reader device, or the identification tag information cannot be recognized from the (incorrect) reading result (measurement tag data).

[0016] Classification data can be output via an output device that is part of the classification device or connected to the classification device. The output device may include, for example, a display device and / or speaker device that outputs video data and audio data, respectively. Alternatively, the output data may be provided to a data interface that provides output data to, for example, a data communication line connected to a control device of an automated laboratory system.

[0017] The tag reader device of the automated laboratory system may include, for example, an optical reader such as a scanner.

[0018] The laboratory devices of an automated laboratory system may be configured to perform one or more functions provided by the automated laboratory system. For example, a laboratory device may be configured to perform pre-analysis. Alternatively, a laboratory device may be configured to perform sample analysis of a sample received in a sample tube. Another laboratory device may be configured to transfer sample tubes between different laboratory devices configured for at least one of pre-analysis and sample analysis. Tag reader devices may be assigned to all or some of the multiple laboratory devices. A single tag reader device may be assigned to two or more laboratory devices.

[0019] Regarding the second classification data, it can be shown that the classification module predicts that the identification tag is not readable by all tag reader devices forming multiple laboratory devices.

[0020] If second classification data is provided indicating that the classification module predicts that the identification tag is unreadable by at least one tag reader device among multiple laboratory devices, the steps of reading the identification tag, determining the tag characteristics, and predicting may be repeated at least once. In such repeated classifications, modified thresholds (readable vs. unreadable) for the classification parameters may be applied, which differ from the original thresholds for the classification parameters applied in the first classification.

[0021] The method may further include determining control data configured to control the handling of sample tubes by an automated laboratory system according to a handling mode. First control data indicating a first handling mode is determined in response to the provision of first classification data, and second control data indicating a second handling mode is determined in response to the provision of second classification data, wherein the second handling mode is different from the first handling mode. In addition to the first or second classification data, control data indicating a first or second mode of handling or processing of sample tubes in the automated laboratory system is provided. For example, the control data may indicate that a sample tube whose identification tag has been read by a classification reader device should be transported or provided only to a specific laboratory device (and not to other laboratory devices in the system) in the process of handling the sample tube by the automated laboratory system. Alternatively, the control data may indicate that a sample tube should not be provided to another laboratory device where it is predicted that the identification tag will not be read so that a tag reader device assigned to such other laboratory device can recognize the identification tag information.

[0022] Different control data may be processed by a control unit of the automated laboratory system, which may be configured to control the processing or handling of sample tubes in the automated laboratory system. The control unit can control the workflow or transfer workflow of the sample device in the automated laboratory system in accordance with the control data. The control data may be generated or determined at least partially by the classification device. Alternatively, the control data may be determined or generated by the control unit of the automated laboratory system in response to the reception of at least one of the first and second classification data provided by the classification device.

[0023] Determining the control data may further include providing control data indicating that the sample tube is assigned to at least one first handling mode selected from the following group: transferring the sample tube to one or more laboratory devices by a transfer device of an automated laboratory system; performing workflow processing of the sample tube within the automated laboratory system; applying pre-analysis to the sample tube; and performing sample analysis of the sample contained in the sample tube by one of the laboratory devices of multiple laboratory devices. According to the second mode of processing or handling the sample tube, it may be limited to laboratory devices in which the tag reader device is expected to be able to read the identification tag.

[0024] Determining the control data may further include providing control data indicating that the sample tube is assigned to at least one second handling mode selected from the following groups: the sample tube is refused handling by the automated laboratory system; the sample tube is replaced; and the sample tube is relabeled. Depending on either the first classification data or the first control data, the sample tube may be refused by the automated laboratory system and its handling may be prevented. Alternatively or additionally, depending on the classification indicating that the identification tag information is not readable by at least one tag reader device among multiple laboratory devices, the sample tube may be replaced and / or relabeled. If the first control data is provided or generated in accordance with the first classification data, the sample tube is made available for handling or processing in the automated laboratory system. For example, the control data may indicate that the sample tube is to be provided to one or more selected laboratory devices in the automated laboratory system during handling or processing of the sample tube.

[0025] Regarding the refusal to process sample tubes by automated laboratory systems, if approximately 90% or less of the tag reader devices assigned to multiple laboratory devices can read the identification tags, the sample tubes may be offered not to be processed (rejected). Alternatively, if approximately 80%, 70%, or 60% or less of the tag reader devices can read the identification tags, the sample tubes may not be processed.

[0026] Furthermore, regarding the refusal to handle a sample tube by an automated laboratory system, even if a tag reader device of a laboratory device providing some first measurement or analysis can read the identification tag, but a tag reader device of another laboratory device providing some second measurement or analysis related to the first measurement or analysis is determined to be unable to read the identification tag, the sample tube still does not need to be processed. For example, the first measurement or analysis may need to be performed before a second measurement or analysis that requires the measurement results of the first measurement or analysis as input. Such knowledge can also prevent two laboratory devices from becoming part of the workflow for processing sample tubes.

[0027] Determining control data may further include providing workflow data indicating the workflow processing of sample tubes within an automated laboratory system, and performing the workflow processing of sample tubes according to the workflow data. The workflow may be determined to exclude some or all laboratory devices in which the tag reader device of such laboratory devices is expected not to be able to read the identification tags. For example, an automated laboratory system may comprise two or more laboratory devices that provide the same or similar functions with respect to at least one of a sample pre-analysis laboratory and sample analysis. In the process of determining the workflow (data) of such multiple laboratory devices, laboratory devices can be selected such that they are part of a workflow in which the identification tags are expected to be read by the tag reader device assigned to such laboratory devices. Thus, workflow management optimized according to classification data can be performed.

[0028] The method can further include: (i) providing first laboratory device data for identifying at least one first laboratory device from a plurality of laboratory devices predicted to be readable by a tag reader device of the at least one first laboratory device with an identification tag; and (ii) providing second laboratory device data for identifying at least one second laboratory device from a plurality of laboratory devices predicted not to be readable by a tag reader device of the at least one second laboratory device with an identification tag. In addition to the general classification indicated by the first or second classification data, first and / or second laboratory device data for identifying at least the first and / or second laboratory devices for which it is predicted whether the identification tag is readable are provided. There can be a list of first laboratory devices indicated by the first laboratory device data that provides a list of laboratory devices for which it has been determined that the identification tag is readable by a tag reader device assigned to such a first laboratory device. Similarly, a list of second laboratory devices can be provided by second laboratory device data indicating a plurality of second laboratory devices predicted to be unreadable by a tag reader device assigned to such a second laboratory device. In the process of determining a plurality of first or second laboratory devices, the result of the prediction (whether the identification tag information is recognizable by reading) can be applied to another laboratory device assigned the same tag reader device or another laboratory device assigned a tag reader device that provides the same reading procedure as the tag reader device assigned to another laboratory device. The prediction of whether identification tag information can be recognized from the reading by the tag reader device is made based on the same tag reader characteristics, whereby such a tag reader device provides similar results with respect to the reading of the identification tag.

[0029] The method may further include at least one of the following: (i) receiving first update information in the classification device indicating that the identification tag is not readable by a tag reader device in a first laboratory while the sample tube is being handled in the automated laboratory system, and updating the classification model in accordance with the first update information; and (ii) receiving second update information in the classification device indicating that the identification tag is readable by a tag reader device in a second laboratory while the sample tube is being handled in the automated laboratory system, and updating the classification model in accordance with the second update information. If it is found that the predictions made for some laboratory devices determined during the classification process are incorrect, the classification device is updated by updating the classification model because the opposite result exists when the sample tube is actually processed or handled in the automated laboratory system. For example, different sets of tag characteristics required to read an identification tag can be defined, and such tag characteristics can be assigned to the laboratory device and provide a result opposite to the prediction previously made for the sample tube. By updating the classification module, a continuously learned or improved classification device can be provided.

[0030] The classification module can include a machine - learned classifier. Providing the classification module can further include providing training data for at least a subset of the laboratory devices of a plurality of laboratory devices, where the training data is for one or more tag reader devices assigned to the subset of laboratory devices and includes (i) first tag characteristics for an identification tag on a first sample tube recognized from a reading of the identification tag on the first sample tube; and (ii) second tag characteristics for an identification tag on a second sample tube not recognized from a reading of the identification tag on the second sample tube. Providing the classification module can further include performing a training procedure for the classification module using the training data, thereby generating a machine - learned classifier. The machine - learned classifier is trained based on training data indicating actual reading events performed to recognize identification tag information by at least a subset of the laboratory devices of a plurality of laboratory devices of an automated laboratory system. For example, if tag reader devices provide similar reading characteristics for reading identification tags of sample tubes, the subset of laboratory devices need not include historical data (representing reading events) for all laboratory devices of the automated laboratory system. Update information regarding actual reading events performed by one or more of the tag reader devices assigned to a laboratory device can be applied to (additional) training after the classification device has been in use for some time.

[0031] The method may further include providing classification data to the control unit of an automated laboratory system and controlling the handling of sample tubes by the automated laboratory system according to the classification data. Upon receiving the classification data by the control unit of the automated laboratory system, control data can be generated to control the handling of sample tubes according to the classification data. For example, according to the control data, sample tubes can be provided only to one or more laboratory devices of the automated laboratory system where the identification tag information is expected to be recognizable (readable) by a tag reader device assigned to one or more such laboratory devices. According to the control data, it can be prevented from transferring or providing sample tubes to other laboratory devices where the identification tag is expected to be unreadable.

[0032] This method may further include providing the classification device in at least one of the following: a handheld device, a mobile communication device, and an input device for an automated laboratory system. The handheld device may be provided by, for example, a handheld scanner device such as a barcode or QR code scanner. The mobile communication device may be, for example, a mobile phone.

[0033] The sorting device may be located at a sample collection point separate from the automated laboratory system. Alternatively, the sample collection point may be provided by an input device of the automated laboratory system. In different embodiments, the sorting device may be communicably connected to the automated laboratory system, for example, the control device of the automated laboratory system. The sorting device may be implemented in, or by, an automated pre-analysis laboratory system or automated analysis laboratory system configured to perform analysis of samples contained in sample tubes.

[0034] The provision of sample tubes may include providing sample tubes to be processed in an automated pre-analysis laboratory system, and providing sample tubes to be processed in an automated analysis laboratory system. The automated laboratory system may be provided as an automated connected laboratory system. In a connected laboratory system, the data processor assigned to the laboratory device may be located in a cloud data system that enables data sharing and exchange between multiple laboratory devices.

[0035] With regard to automated laboratory systems, the above-described embodiments can be applied mutatis mutandis to methods for classifying identification tags on sample tubes. [Brief explanation of the drawing]

[0036] Further embodiments are described in more detail below with reference to the figures.

[0037] [Figure 1] This is a schematic diagram of an automated laboratory system configured to process or handle sample tubes containing samples for the application of at least one of pre-analysis and sample analysis. [Figure 2] Figure 1 is a schematic block diagram of a method for classifying identification tags on sample tubes containing samples processed or handled in the automated laboratory system. [Modes for carrying out the invention]

[0038] Further Description of Embodiments Figure 1 shows a schematic diagram of a configuration comprising an automated laboratory system 1 and a classification device 2 that can be communicatively connected to at least one control device 3 of the automated laboratory system 1 for data communication. Alternatively, the classification device 2 may be implemented by equipment of the automated laboratory system 1, for example, by equipment such as an input laboratory device or a sample collection point.

[0039] The automated laboratory system 1 comprises a plurality of laboratory devices 4 configured to process sample tubes 5 containing samples 6 for at least one of pre-analysis and sample analysis. The automated laboratory system 1 can be configured to process biological or chemical samples. For example, the automated laboratory system 1 can be configured to analyze samples of one or more body fluids.

[0040] The classification device 2 may be provided by a separate module in the automated laboratory system 1. Alternatively, the classification device 2 may be part of the pre-analysis device of the automated laboratory system 1. For example, the classification device 2 may be provided in a sample checking module (not shown). Such a sample checking module is known, for example, from European Patent Application Publication No. 2246689.

[0041] Multiple laboratory devices 4, together with a classification device 2, can be installed in connected laboratories, for example, to provide data exchange and sharing between the multiple laboratory devices 4 and the classification device 2. To perform such data communication and data processing, each of the multiple laboratory devices 4 and the classification device 2 comprises one or more data processors, data memory, and data communication devices, as is known in itself. Data communication may be performed by at least one of wireless data communication and wired data communication.

[0042] As is known, the sample tubes 5 can be transferred between laboratory devices 4 of a plurality of laboratory devices by a transfer system (not shown). One or more sample tubes can be placed in a rack for transfer.

[0043] The classification device 2 comprises a classification reader device 7 and a classification module 8. The classification reader device 7 is configured to read identification tags 9 on the sample tube 5, for example, by optical reading. The identification tags 9 may be provided with, for example, a barcode or a QR code. The identification tags 9 are configured to uniquely identify the sample tube 5 containing the sample 6.

[0044] Under normal, everyday use, the condition of the identification tag 9 may deteriorate, including tearing, peeling, discoloration, and other deformations. Due to such deterioration, the identification tag information provided by the identification tag 9 may not be recognizable by reading the identification tag 9. The classification device 2 is for predicting whether a tag reader device 10, which is provided in the automated laboratory system 1 and assigned to one of the multiple laboratory devices 4, can recognize the identification tag information by reading the identification tag 9.

[0045] During operation, when a sample tube 5 arrives at one laboratory device from multiple laboratory devices 4, the identification tag 9 is read by the tag reader device 10 of the laboratory device 4, for example, by optical reading or scanning. Depending on whether the tag reader device 10 can recognize the identification tag information from such reading, the sample tube 5 can be properly processed or handled by the laboratory device to which the tag reader device is assigned.

[0046] The tag reader device 10 may be different for each laboratory device 4. Alternatively, the tag reader devices 10 for multiple or all laboratory devices 4 may be the same. For example, the tag reader device 10 may use different reading technologies such as LED light sensors, laser light sensors, and / or image-based devices. The tag reader device 10 may be implemented differently in the laboratory device 4 with respect to at least one of the following: illumination of the identification tag 9 for reading, angle for tag reading, and reading distance between the identification tag 9 and the tag reader device 10. Also, the tag reader devices 10 may come from different suppliers, thereby providing different types of tag reader devices.

[0047] If such identification tag information cannot be recognized from the reading, for example, if identification tag 9 cannot be read at all, the handling of the sample tube 5 containing the sample 6 by such laboratory equipment may fail or be performed incorrectly. To avoid such a situation, the identification tag 9 on the sample tube 5 is read in advance by the classification reader device 7 of the classification device 2, for example, before the sample tube 5 is handled by one of the multiple laboratory equipment devices 4. This can be done at the sample collection point. In response to such a reading, the characteristics of the identification tag 9 are provided to the measurement tag classification device 2. For example, an image of the identification tag 9 can be detected by the classification reader device 7. The image can then be processed and analyzed by image data analysis or processing performed by one or more data processors provided in the classification device 2. Through such analysis, the identification tag 9 is classified as described in more detail with reference to Figure 2.

[0048] Referring to Figure 2, the method for classifying the identification tag 9 on the sample tube 5 is described. In step 20, the identification tag 9 is read by the classification reader device 7, thereby generating measured tag data indicating the tag characteristics of the identification tag 9. In step 21, the tag characteristics, sometimes also called tag features, are determined by analyzing the measured tag data. For example, tag characteristics can indicate the level of degradation of the identification tag 9. The classification module 8 makes a prediction as to whether at least one of the tag reader devices 10 among the multiple laboratory devices 4 can read the identification tag 9 and recognize the identification tag information from such reading. The identification tag information includes at least one of the sample information of the sample 6 received in the sample tube 5 and the sample tube information of the sample tube 5. The sample information may indicate the type of sample, such as a blood sample or a sample of another body fluid. The sample tube information may indicate, for example, the type of sample tube and / or the size of the sample tube 5.

[0049] The classification device 2 predicts whether such identification tag information can be recognized by the tag reader devices 10 of multiple laboratory devices 4. Separate prediction results may be provided for each laboratory device from the multiple laboratory devices 4. Alternatively, such predictions may be provided only for a subset of laboratory devices from the multiple laboratory devices 4. For example, information regarding at least one of pre-analysis and sample analysis may be provided to the classification device 2, and such information indicates one or more laboratory devices (a subset of laboratory devices) involved in the processing or handling of sample tubes 5 in the automated laboratory system 1. In such a case, the prediction regarding the recognizability of the identification tag information from reading the identification tag 9 may be made only for such subset of laboratory devices. The prediction may be made per device (for each of the laboratory devices related to or selected to the prediction process performed by the classification device 2).

[0050] In step 23, classification data is generated indicating at least whether there is at least one tag reader device 10 among a plurality of laboratory devices 4 that is predicted to be readable by such a tag reader device. Depending on such classification data, the sample tube 5 containing the sample 6 is processed or handled in the automated laboratory system 1. If the classification data indicates that there is no tag reader device 10 among a plurality of laboratory devices 4 that is predicted to be recognizable by the identification tag information, the sample tube 5 may be rejected from any (further) processing in the automated laboratory system 1. Replacement of the sample tube 5 is possible. Alternatively or additionally, relabeling may be applied to the sample tube 5, thereby replacing the identification tag 9. The sample tube 5 with the new identification tag may then be classified again by the classification device 2.

[0051] If the classification data indicates that at least one tag reader device 5 among multiple laboratory devices 4 is capable of recognizing the identification tag information from reading the identification tag 9, the sample tube 5 is provided for further processing by the automated laboratory system 1. For example, the sample tube 5 is provided to one or more laboratory devices among multiple laboratory devices 4. In the course of such processing, at least one of the following can be performed: pre-analysis and sample analysis.

[0052] The classification device 2 may include one of the following: a handheld device such as a scanner or optical reader, a mobile device such as a mobile phone, and an input device for the automated laboratory system 1 such as a fixed input device. The classification of the identification tags 9 on the sample tubes 5 may be performed at a sample tube collection point, which is located away from or near the automated laboratory system 1.

[0053] The classification module 8 of the classification device 2 may include a machine learning classifier trained on a set of training data. The training data may include historical data indicating the success or failure of read events performed by a tag reader device 10 provided in at least one of the automated laboratory system 1 or some other automated laboratory system (not shown). The training data indicates read events that enable the recognition of identification tag information from the reads. Based on such historical data, a machine learning classifier can be trained, and following the machine learning classifier can be applied to the classification module 8. The training data may provide device-specific historical data for multiple laboratory devices 5. Thus, in this embodiment, the machine learning classifier can be trained individually for multiple laboratory devices 4.

[0054] A machine learning-based classifier provides predictive modeling to approximate a so-called mapping function from an input variable (identifier tag characteristics or features) to a discrete output variable (identifier tag readability).

[0055] The disclosed technology allows the automated laboratory system 1 to implement different operating modes. The first mode is sometimes called the "passive mode." This operating mode can be used when the system starts up, and the accuracy of the classification module 8 (predictor) can be measured by recording the performance of the classification device 2. The second mode is sometimes called the "gatekeeper mode." If the classification device 2 receives a signal from the classification data indicating that a sample tube cannot be read by one of the instruments of the automated laboratory system 1, the sample tube is rejected for relabeling. In the third mode, sometimes called the "adaptive mode," some sample tubes that have been flagged as unreadable are allowed in the automated laboratory system 1 to verify the prediction (or adjust the predictor if the sample can be read).

[0056] For classification, the classification device 2 may apply regression. In the embodiment, the tag reader device 10 of the laboratory device 4 may provide one parameter (e.g., code readability index). Each tag reader device 10 may apply its own individual threshold to determine whether an identification tag 9 is readable by the tag reader device 10. If the classification device 2 provides multiple parameters (or variables), an SVM (Support Vector Machine) or a CNN (Convolutional Neural Network) may be applied to the classification. Image recognition can be applied to multi-parameter classification. As for other parameters (or variables) that can be applied additionally or alternatively, they can be selected from barcode contrast, brightness, gradation level, multicolor image, and integrity, taking into account the imaging device.

[0057] In different embodiments, the disclosed technology can offer several advantages. The determination (classification) of the readability of the identification tag 9 on the sample tube 5 is based on the actual performance of the automated laboratory system 1, and not on a reference level. If such an early detection and determination mechanism were implemented using only a reference level, a margin would need to be incorporated into the system. For example, if barcodes of quality levels 3 to 5 (highest) can be read, the quality levels that can be accepted in the input module are 4 to 5. If intended, this input quality control system can be configured to take in only sample tubes that can be reliably read by all instruments.

[0058] Improved flexibility is provided. The best selection of equipment for identification tag 9 is defined. Even similar equipment (laboratory equipment) may respond differently to different identification tag defaults, and classification module 8 learns this.

[0059] The reading technique is adaptable. The proposed technique is technique-independent. The only important information used from the different instruments of automated laboratory system 1 is the reading result (readable or unreadable).

[0060] Ultimately, this technology provides value to the operator of the automated laboratory system 1, which can be implemented as a connected laboratory, because early detection ultimately leads to faster results for sample analysis. Furthermore, rework can be reduced because the sorting device 2 provides selection based on the actual reading performance of each tag reader device 10 in the automated laboratory system 1, and does not provide selection to any detection level. Even less rework is achieved if sample tubes can be sent to different instruments.

[0061] This technology can be used in unconnected automated laboratory systems compared to connected laboratory systems, the only difference being that the technician or user may need to follow instructions partially provided by the sample tube classification.

Claims

1. A method for classifying identification tags on sample tubes containing samples to be processed in an automated laboratory system, comprising providing a sample tube (5) having an identification tag (9) and containing a sample (6) to be analyzed in an automated laboratory system (1) having a plurality of laboratory devices (4), wherein a tag reader device (10) configured to read the identification tag (9) in order to recognize the identification tag information is assigned to each of the plurality of laboratory devices (4), the method for classifying identification tags on sample tubes (5), the method for classifying identification tags on sample tubes (5), the method for classifying identification tags on sample tubes (6), the method for classifying identification tags on sample tubes (5), the method for classifying identification tags on sample tubes containing samples to be processed in an automated laboratory system (1) having a plurality of laboratory devices (4), the method for classifying identification tags on sample tubes (5), the method for classifying identification tags on sample tubes (5) having an identification tag (9) and containing a sample (6) to be analyzed, the method for classifying identification tags on sample tubes (5) having an identification tag (9) and containing an identification tag (9) a plurality of laboratory devices (4), the method for classifying identification tags on sample tubes (5) having an identification tag (9) and containing a sample (6) to be analyzed, the method for classifying identification tags on sample tubes (5) having a plurality of laboratory devices (4) Furthermore, in the classification device (2), To provide a classification module (8) configured to predict whether the identification tag information can be recognized from measurement tag data that is detected by the tag reader device (10) assigned to each of the plurality of laboratory devices (4), and that indicates the characteristics of the identification tag (9), for each of the plurality of laboratory devices (4), The identification tag (9) on the sample tube (5) is read by the classification reader device (7) of the classification device (2), thereby providing the measurement tag data of the identification tag (9) that indicates the tag characteristics of the identification tag (9) on the sample tube (5). The tag characteristics of the identification tag (9) are determined from the measurement tag data, The classification module (8) receives the tag characteristics and predicts whether at least one of the plurality of laboratory devices (4) can read the identification tag (9), To provide classification data, the classification data is First classification data indicating that the classification module (8) predicts that the identification tag (9) is readable by at least one of the multiple laboratory devices (4) and the tag reader device (10), and To provide classification data which is a second classification data different from the first classification data, indicating that the classification module (8) predicts that the identification tag (9) is unreadable by at least one of the multiple laboratory devices (4) of the tag reader device (10), The classification data is provided to the control device (3) of the automated laboratory system (1), The handling of the sample tube (5) by the automated laboratory system (1) is controlled according to the classification data. Includes, A method for determining that the workflow of the sample tube (5) in the automated laboratory system (1) does not include some or all of the laboratory devices (4) in which the tag reader device (10) of the laboratory device (4) is expected to be unable to read the identification tag (9).

2. The method further includes determining control data configured to control the handling of the sample tube (5) by the automated laboratory system (1) according to the handling mode, A first control data indicating a first handling mode is determined in response to providing the first classification data. The method according to claim 1, wherein second control data indicating a second handling mode is determined in response to providing the second classification data, and the second handling mode is different from the first handling mode.

3. Determining the aforementioned control data is as follows: The sample tube (5) is transferred to one or more laboratory devices by the transfer device of the automated laboratory system (1). The automated laboratory system (1) performs workflow processing of the sample tube (5). Applying sample pretreatment to the sample tube (5), and The sample (6) contained in the sample tube (5) is subjected to sample analysis by one of the multiple laboratory devices (4) described above. To provide control data indicating that the sample tube (5) is assigned to at least one first handling mode selected from the group. The method according to claim 2, further comprising:

4. Determining the aforementioned control data is as follows: Remove the sample tube (5) from handling by the automated laboratory system (1), Replacing the sample tube (5), and Relabeling the sample tube (5) To provide control data indicating that the sample tube (5) is assigned to at least one second handling mode selected from the group. The method according to claim 2 or 3, further comprising:

5. The method according to claim 3, or claim 4 as dependent on claim 3, further comprising determining the control data to provide workflow data indicating the workflow processing of the sample tube (5) in the automated laboratory system (1), and executing the workflow processing of the sample tube (5) in accordance with the workflow data.

6. To provide first laboratory device data that identifies at least one of the plurality of laboratory devices (4) whose identification tag (9) is expected to be readable by the tag reader device (10) of at least one first laboratory device, and To provide second laboratory device data that identifies at least one second laboratory device of the plurality of laboratory devices (4) whose identification tag (9) is predicted to be unreadable by the tag reader device (10) of at least one second laboratory device. The method according to any one of claims 1 to 5, further comprising at least one of the following.

7. In the classification device (2), while the sample tube (5) is being handled in the automated laboratory system (1), the device receives first update information indicating that the identification tag (9) is unreadable by the tag reader device (10) of the first laboratory device, and updates the classification model in response to the first update information, and In the classification device (2), while the sample tube (5) is being handled in the automated laboratory system (1), the device receives second update information indicating that the identification tag (9) is readable by the tag reader device (10) of the second laboratory device, and updates the classification model in response to the second update information. The method according to claim 6, further comprising at least one of the following.

8. The classification module (8) is equipped with a machine learning classifier, and the purpose of providing the classification module (8) is to The present invention provides training data for at least a subset of the laboratory devices (4) from the plurality of laboratory devices, wherein the training data is provided for one or more tag reader devices (10) assigned to the subset of laboratory devices. The first tag characteristics of the identification tag on the first sample tube, as recognized from reading the identification tag (9) on the first sample tube, and To provide training data showing the second tag characteristics of the identification tag (9) on the second sample tube that was not recognized from reading the identification tag (9) on the second sample tube, Using the aforementioned training data, a training procedure is performed for the classification module (8), thereby generating the machine-learned classifier. The method according to any one of claims 1 to 7, further comprising:

9. The method according to any one of claims 1 to 8, further comprising providing the classification device (2) in at least one of a handheld device, a mobile communication device, and an input device of the automated laboratory system (1).

10. The method according to any one of claims 1 to 9, wherein providing the sample tube (5) includes at least one of providing a sample tube to be processed in an automated sample pre-processing laboratory system and providing a sample tube to be processed in an automated analytical laboratory system.

11. An automated laboratory system (1) for processing sample tubes containing a sample for at least one of sample pretreatment and sample analysis, A sample tube (5) having an identification tag (9) and containing a sample (6) to be processed for at least one of sample pretreatment and sample analysis, Multiple laboratory devices (4), each assigned a tag reader device (10) configured to read the aforementioned identification tag (9), Classification device (2), For each of the plurality of laboratory devices (4), a classification module (8) is provided which is configured to predict whether the identification tag (9) can be recognized from measurement tag data that is detected by the tag reader device (10) assigned to the laboratory device and that indicates the characteristics of the identification tag (9), The identification tag (9) on the sample tube (5) is read by the classification reader device (7), thereby providing the measurement tag data of the identification tag (9) that indicates the tag characteristics of the identification tag (9) on the sample tube (5). The tag characteristics of the identification tag (9) are determined from the measurement tag data. In the classification module (8), the tag characteristics are received, and it is predicted whether at least one of the plurality of laboratory devices (4) can read the identification tag (9). To provide classification data, the classification data is First classification data indicating that the classification module (8) predicts that the identification tag (9) is readable by at least one of the multiple laboratory devices (4) and the tag reader device (10), and The classification data provided is a second classification data, different from the first classification data, indicating that the classification module (8) predicts that the identification tag (9) is unreadable by at least one of the tag reader devices (10) among the plurality of laboratory devices (4). A classification device and Equipped with, The automated inspection room system (1) further comprises a control device (3), and the automated inspection room system (1) The control device (3) is provided with the classification data, The handling of the sample tube (5) is controlled according to the classification data. It is configured in such a way, An automated laboratory system (1) wherein the workflow of the sample tube (5) in the automated laboratory system (1) is determined not to include some or all of the laboratory devices (4) in which the tag reader device (10) of the laboratory device (4) is expected not to be able to read the identification tag (9).

12. An automated laboratory system (1) according to claim 11, provided as one of an automated sample preparation laboratory system and an automated analysis laboratory system.

13. The automated laboratory system (1) according to claim 11 or 12, provided as a connected automated laboratory system (1).

14. The automated laboratory system (1) according to any one of claims 11 to 13, wherein the classification device (2) is at least one of a handle-held device, a mobile communication device, and an input device of the automated laboratory system (1).

Images