Method for detecting and reporting operational errors in an in vitro diagnostic system, and in vitro diagnostic system

Abstract

The present disclosure relates to a method for detecting and reporting operational errors in an in vitro diagnostic system (1) for measuring a sample of a bodily fluid, the method including providing a plurality of sample containers (2), each containing a sample of the bodily fluid, a plurality of functional modules (3) comprising an analytical device (4) configured to measure the samples, a handling system (5) configured to handle the plurality of sample containers (2), and an automated track (6) provided by the handling system (5) and configured to transport the plurality of sample containers (2) to the analytical device (4).The method includes providing an operation control device (7) connected to at least one of the functional modules (3) and configured to control operation of the at least one functional module (3), the operation control device (7) comprising one or more data processors (8), application software running on the one or more data processors (8) for controlling operation of the at least one functional module (3); controlling operation of the at least one functional module (3) by the operation control device (7); detecting and reporting an operation error by an error detection and reporting device (9), wherein the error detection and reporting device (9) detects an operation error in operation of at least one of the plurality of functional modules (3) and the operation of the operation control device (7) and provides error data indicative of the operation error; receiving user input via a user interface (10) after detecting the operation error; and providing labeling data indicative of information regarding the operation error in addition to the error data in response to receiving the user input. and providing error report data including error data and labeling data, and transmitting the error report data to an error repository (11) located remotely with respect to both the plurality of functional modules (3) and the operation controller (7), receiving the error report data in a machine learning process running in a data processing device connected to the error repository (11), processing the error report data by the machine learning process in the data processing device, and providing an application software update for the application software in response to the processing of the error report data by the machine learning process in the data processing device, providing the application software update to the operation controller (7), and controlling operation of the at least one functional module (3) by the operation controller (7), including executing the application software including the application software update. Further provided is an in-vitro diagnostic system for measuring a sample of a body fluid.

Description

[Technical Field]

[0001] The present disclosure relates to a method for detecting and reporting operational errors in an in vitro diagnostic system configured to measure a sample of a bodily fluid, and to the in vitro diagnostic system. [Background technology]

[0002] In vitro diagnostic systems are applied to measure samples of body fluids. The samples are contained in sample containers or vessels that are processed in the in vitro diagnostic system. The sample containers can be handled by a sample container handling or distribution system, for example, to place the sample containers in a sample container carrier and to remove the sample containers from the sample container carrier. For example, a laboratory sample distribution system is disclosed in WO 2016 / 038014.

[0003] In an in vitro diagnostic system, sample containers are moved along a processing line for processing, and the sample containers and / or sample container carriers are moved by a transport device having one or more actuators and actuator drives or drive devices for driving the actuators. For example, a sample container can be moved or relocated from a first work station to a second work station provided in the processing line of the in vitro diagnostic system. A work station may also be referred to as a work position.

[0004] Diagnostic systems currently perform quality control in the field using pre-installed algorithms. In particular, current software solutions for sample quality checks in such diagnostic systems are typically camera-based and are not developed based on the behavior of the system installed in the field.

[0005] Systems for camera-based quality checks of samples are disclosed in U.S. Patent Application Publication No. 2019 / 0 033 209, European Patent Application Publication No. 3 408 640, and European Patent Application Publication No. 3 408 641. A model-based method for quantifying an analyte is disclosed. The method includes providing a sample, capturing images of the sample while irradiated with multiple spectra at different nominal wavelengths, exposing the sample, classifying the sample into various classes, including one or more of a serum or plasma portion, a sedimented blood portion, a gel separator (if used), air, a tube, a label, or a cap, and quantifying the sample. The quantification includes determining one or more of the location of the liquid-air interface, the location of the serum-blood interface, the location of the serum-gel interface, the location of the blood-gel interface, the volume and / or depth of the serum or plasma portion, or the volume and / or depth of the sedimented blood portion.

[0006] EP 3 171 302 A1 discloses a method for generating an entry in an electronic laboratory journal, comprising: selecting an experimental workflow specifying at least one action to be performed by an operator, the action including one or more laboratory devices; recording one or more images or videos of the actions by the operator by using a camera associated with an augmented reality device worn by the operator; and generating an entry in the electronic laboratory journal based on the recorded one or more images or videos.

[0007] The disclosure of China Patent Application Publication No. 108 984 613 refers to a method for cross-item classification of defect reports based on transfer learning, which includes: selecting data, determining source data and target data, pre-processing the data, calculating a vector representation of each defect report in steps by training a defect report semantic model, dividing the source data and the target data into training data and inspection data in steps, and adjusting the weights of the training data by transfer learning so that the error of the classification result is minimized, and classifying the target data across items via a machine learning classifier using the transferred data training classifier in steps.

[0008] WO 2019 / 089 578 discloses a system including a computing device including a memory configured to store instructions. The system also includes a processor for executing instructions to perform operations including receiving an image including text content in at least one font. The operations also include identifying at least one font represented in the received image using a machine learning system. The machine learning system is trained using images representing multiple training fonts. Some of the training images include text located in the foreground and positioned over a captured background image.

[0009] U.S. Patent Application Publication No. 2018 / 232 657 discloses a method and system for generating cognitive insights, including receiving training data based on interactions between a user and a cognitive learning and reasoning system, performing a plurality of machine learning operations on the training data, generating a cognitive profile based on information generated by performing the plurality of machine learning operations, and generating cognitive insights based on the profile generated using the plurality of machine learning operations.

[0010] The disclosure of U.S. Patent Application Publication No. 2019 / 102 700 relates to an integrated machine learning platform. The machine learning platform can convert machine learning models with different schemas into machine learning models that share a common schema, organize the machine learning models into model groups based on specific criteria, and perform pre-deployment evaluation of the machine learning models. The machine learning models in the model groups can be evaluated or used individually or as a group. The machine learning platform can be used to deploy the model groups and a selector to a production environment, and the selector can be trained to dynamically select models from the model groups in the production environment in different contexts or with different input data based on scores determined using specific scoring metrics, such as specific business objectives.

[0011] A fault repair report management method based on customer service is disclosed in China Patent Application Publication No. 109 635 968. The method includes the steps of obtaining voice information of a repair report user, analyzing the voice information to obtain text information corresponding to the voice information, generating a repair work order through a keyword capturing algorithm according to the text information and a preset keyword library, and delivering the repair work order to a target terminal and receiving feedback information sent by the target terminal.

[0012] Chinese Patent No. 1 043 011 368 discloses a method and apparatus for reporting and processing fault information. The method includes analyzing a log of a user equipment according to preset related parameters of the fault information, where the fault information can be obtained from the log, and the preset related parameters of the fault information include fault keywords. According to the obtained fault information, a fault file is generated and reported to a server, where the fault file includes the fault keywords.

[0013] European Patent Application Publication No. 3 517 970 discloses a system for a service department to contact a customer to improve customer service when an alarm occurs due to an abnormality in an automated analyzer and the automated analyzer stops, and it is determined that the customer has difficulty handling the alarm themselves. The automated analyzer includes a display that displays an alarm screen, a computer that generates alarm occurrence information indicating that an alarm has occurred due to an abnormality and alarm deletion information indicating that the customer has deleted the alarm displayed on the alarm screen based on the alarm occurrence information, and an information sharing unit that shares the alarm generation information and alarm deletion information generated by the computer with an external alarm maintenance system.

[0014] US Patent Application Publication No. 2009 / 199 052 refers to a management system including a plurality of analytical instruments, and a computer system connected to the analytical instruments via a network, each of the analytical instruments having a data transmitter for transmitting data generated by the analytical instruments to the computer system via the network, the computer system including a memory under the control of a processor, the memory storing instructions that enable the processor to perform operations, the operations including: (a) receiving a plurality of data transmitted from the data transmitters of the plurality of analytical instruments; (b) generating an aggregated result based on the plurality of received data, the aggregated result being used to determine a decision condition for determining whether a notification to a user of the analytical instrument is necessary; and (c) outputting the aggregated result. A computer system and a method for providing information are also disclosed.

[0015] U.S. Patent Application Publication No. 2007 / 174653 discloses an automated laboratory apparatus comprising a mechanism for performing operations on laboratory samples, a scheduler for causing the mechanism to process the laboratory samples according to a programmed process, logic for detecting errors that occur in the process controlled by the scheduler, logic for accepting user-defined error handling routines for the errors, and logic for executing the error handling routine when an error is encountered. Laboratory automation systems, methods of laboratory automation, computer-implemented software program products, methods of doing business, and laboratory automation networks are also described.

[0016] GB 2 573 336 A1 discloses a modular medical device, such as an in vitro diagnostic (IVD) device, that is self-tested for faults using sensors and actuators already present in the device's modules. These are used to perform test runs to check for abnormal operation within the device. The actuators can be pump or stepper motors, Peltier elements, RFID sensors, acoustic or optical sensors, or sensors for capacitance, pressure, flow rate, or fill level. The sensors can also be rotary or linear encoders or RFID receivers. The modular device includes a pipettor, a pump, a module for cleaning reaction vessels, a thermostat module, or a module for measuring reaction strength. The test sequence is repeated over time, and data is compared to indicate impending faults.

[0017] The specification of U.S. Patent No. 9,430,312 refers to a system including an analytical instrument that analyzes samples or a processing instrument that pre-processes samples, and a management device that manages at least one of the analytical instrument and the processing instrument, wherein the management device includes an error detection means that detects errors in the analytical instrument or the processing instrument, a storage means that stores an operator notification management table in which at least one operator is registered for each type of error, an error notification means that notifies an operator who should individually deal with the error detected by the error detection means of the error detected by the error detection means based on the operator notification management table and according to the type of error detected by the error detection means, and a registration means that registers an operator who handled the error as a troubleshooter, among the operators notified of the error by the error notification means. Summary of the Invention

[0018] It is an object of the present invention to provide a method for detecting and reporting operational errors in an in vitro diagnostic system configured to measure a sample of a body fluid using improved techniques for avoiding operational errors, and to provide an in vitro diagnostic system that allows for safe operation.

[0019] To solve this problem, there is provided a method for detecting and reporting an operating error in an in vitro diagnostic system configured to measure a sample of body fluid, and an in vitro diagnostic system, respectively, according to independent claims 1 and 12. Further embodiments are set out in the dependent claims.

[0020] According to one aspect, a method is provided for detecting and reporting operational errors in an in vitro diagnostic system configured to measure a sample of a bodily fluid, the method including: providing a plurality of sample containers in the in vitro diagnostic system, each containing a sample of the bodily fluid; and providing a plurality of functional modules, the plurality of functional modules comprising: an analytical device configured to measure the samples; a handling system configured to handle the plurality of sample containers; and an automated track provided by the handling system and configured to transport the plurality of sample containers to the analytical device. The method further includes providing an operation control device connected to at least one of the functional modules and configured to control operation of the at least one functional module, the operation control device comprising one or more data processors, application software running on the one or more data processors to control operation of the at least one functional module; controlling the operation of the at least one functional module by the operation control device; and detecting and reporting an operation error by an error detection and reporting device, the detecting and reporting operation error including detecting an operation error in operation of at least one functional module of the plurality of functional modules and the operation control device, providing error data indicative of the operation error, receiving user input via a user interface after detecting the operation error, and providing labeling data indicative of information regarding the operation error in addition to the error data in response to receiving the user input, providing error report data including the error data and the labeling data, and transmitting the error report data to an error repository located remotely with respect to both the plurality of functional modules and the operation control device.The method further includes receiving error report data at a machine learning process running in a data processing device connected to the error repository, processing the error report data by the machine learning process in the data processing device, and in response to the processing of the error report data by the machine learning process in the data processing device, providing an application software update for the application software, providing the application software update to an operation control device, and controlling operation of at least one functional module by the operation control device, including executing the application software including the application software update.

[0021] According to another aspect, an in-vitro diagnostic system for measuring a sample of a bodily fluid is provided, the in-vitro diagnostic system comprising: a plurality of functional modules, each of which includes a plurality of sample containers containing a sample of the bodily fluid; an analytical device configured to measure the samples; a handling system configured to handle the plurality of sample containers; and an automated track provided by the handling system and configured to transport the plurality of sample containers to the analytical device. An operation controller is connected to at least one of the functional modules and configured to control the operation of the at least one functional module, and comprises one or more data processors. Application software executes on the one or more data processors to control the operation of the at least one functional module. The error detection and reporting device is configured to detect and report an operation error in the operation of at least one of the plurality of functional modules and the operation controller, provide error data indicative of the operation error, receive user input via a user interface after detecting the operation error, provide labeling data indicative of information regarding the operation error in addition to the error data in response to receiving the user input, provide error report data including the error data and the labeling data, and transmit the error report data to an error repository located remotely from both the plurality of functional modules and the operation controller. The in vitro diagnostic system further comprises a data processing device coupled to the error repository, the data processing device configured to receive error report data at a machine learning process executing on the data processing device, process the error report data by the machine learning process, and provide an application software update for the application software in response to the processing of the error report data by the machine learning process, wherein the application software update is provided to the operation controller for controlling operation of the at least one functional module by the operation controller, including executing the application software that includes the application software update.

[0022] Not only is data indicative of such operational errors, i.e., error data, provided, but additional information collected from users implementing operational measures to operate the in vitro diagnostic system is also provided. Such additional information is provided by the labeling data. Thus, the error report data, including both the error data and the labeling data received (and stored) in the error repository, provides an improved method for error or failure analysis of the in vitro diagnostic system not only on-site, i.e., at the location where the in vitro diagnostic system is used, but also remotely. For example, a development team can access the error report data.

[0023] The machine learning process may be configured to provide application software updates.

[0024] In an embodiment, during operation of the in-vitro diagnostic system in the field, the operation controller may initially be provided with pre-installed application software. However, the pre-installed software may cause operation errors. In particular, the operation errors may be systematic errors. Without a learning process, errors, such as systematic errors, typically remain. By reporting errors, for example, via error report data, a machine learning process can learn from these errors, and specifically, systematic errors can be identified. In other words, errors (especially error report data) can be provided as training data and / or validation data as input for the machine learning process. The machine learning process may be the same as that used for software pre-installed on the in-vitro diagnostic system, particularly the operation controller. The machine learning process is provided to calculate or determine updated application software (e.g., an operation model) for the operation controller that has fewer errors than the pre-installed application software.

[0025] The error report data received by the machine learning process may indicate an error in the in-situ in-vitro diagnostic system and / or an error in a further in-vitro diagnostic system.

[0026] The machine learning process may learn without developer supervision, thereby allowing the machine learning to occur automatically in response to receiving error report data. Such an automated machine learning process may be initiated automatically after receiving error report data indicative of an operational error or in response to receiving error report data for a predetermined number of operational errors in an error repository.

[0027] Alternatively, the machine learning process may be overseen by a developer. The error report data in the error repository may be provided to the developer. The developer may provide a second machine learning process that is different from the first machine learning process. The developer may select one or more error report data from the plurality of error report data, and the selected error report data may or may not be included in the training data. In an embodiment, only a selection of the error report data may be received by the machine learning process.

[0028] Typical types of errors that may be reported via error report data (error report data) may include at least one of the following: errors caused by the system related to the amount or type of substance / body fluid detected in the sample vessel; errors caused related to the identification of the sample vessel; errors caused related to the determination of the quality of the substance in the sample vessel;

[0029] In response to processing the error report data by the machine learning process, new application software for the motion controller may be provided. The new application software for the motion controller may be different from the application software for the motion controller currently provided in the in-situ in-vitro diagnostic system. The application software for the motion controller currently provided in the in-situ in-vitro diagnostic system (the old application software) may be overwritten by the new application software. Alternatively, only portions representative of the differences between the old and new application software may be incorporated into the old application software (i.e., providing an application software update).

[0030] In an embodiment, the method further includes preprocessing the error report data in a data processing device, where the preprocessing includes formatting the error report data into formatted error report data having a data format processable by the machine learning process. The data processing device may be configured with or operatively connected to an error repository for accessing the error report data. The preprocessing may include, for example, at least one of unpacking and / or decrypting data packages and reconstructing data if the data is stored in a hierarchical format.

[0031] Preprocessing of the error report data in the data processing device may include converting analog data to digital data, particularly converting analog error data or analog labeling data to digital error data or digital labeling data. Furthermore, for example, one or more features may be selected from the error data. The features may indicate one or more operating parameters of the in-vitro diagnostic system. The processing device may be provided for such feature selection of the error data. Alternatively, an error repository may be provided for such feature selection from the error data. The error data may be formatted to provide a feature vector that can include these selected features. The feature vector may be an n-dimensional vector indicating n features. Alternatively, the feature vector may be a scalar. In a preferred embodiment, the feature vector is a numeric vector.

[0032] The labeling data may be formatted such that a target is provided that is linked to the error data, in particular the feature vector. In other words, the target labels the error data, in particular the feature vector. The target may indicate the labeling data. The target may indicate the type of motion error. The target may be a scalar or a vector. The target may be a numerical scalar or vector.

[0033] The target and feature vector may be provided as input data for a machine learning process or algorithm. The target and feature vector may be incorporated into a training dataset. The training dataset includes training entries. The feature vector, together with the target, may define one training entry. The feature vector may be incorporated into a feature matrix, and the target may be incorporated into a target vector / matrix, respectively. In particular, the feature vector and target may be provided so that they can be used to train algorithms such as nearest neighbor classification, neural networks, and Bayesian approaches, as statistical techniques. Alternatively, the feature vector and target may be provided to train other machine learning algorithms / processes. The scalars, vectors, and matrices disclosed herein may be provided as bit vectors. The newly formatted error report data can be easily fed into existing machine learning algorithms. In particular, the newly formatted error report data can complement previously formatted error report data. As each error is detected, new error report data can be formatted and fed into the machine learning algorithm, generating a successively increasing set of error report data. The successively increasing set of error report data can accompany a successively increasing set of training data. A larger set of training data can be associated with more accurate results for machine learning algorithms.

[0034] The error data may indicate an operational error of the analyzer. In another embodiment, the error data may indicate an operational error of the handling system. Additionally or alternatively, the error data may indicate an operational error of the automated truck. The error data may indicate at least one operational parameter of the analyzer, the handling system, or the automated truck. In particular, the error data may indicate at least one operational parameter of the analyzer, the handling system, or the automated truck when the error occurred, i.e., was detected or reported. Alternatively, the error data may indicate at least one operational parameter of the analyzer, the handling system, or the automated truck at a time before the error occurred. The error data may indicate a function value determined by a function of the operational parameters of the analyzer, the handling system, and / or the automated truck. For example, the error data may indicate an average value (or a maximum value, minimum value, median value, etc.) of the operational parameter of the analyzer, the handling system, and / or the automated truck at a time before the error occurred.

[0035] Error data can exhibit one of two types of errors: a false positive error and a false negative error. In medical testing, and more generally in binary classification, a false positive is a data reporting error in which a test result inappropriately indicates the presence of a condition, such as a disease, when in fact it is not present (a positive result), and a false negative is an error in which a test result inappropriately does not indicate the presence of a condition, when in fact it is present (a negative result). These are two types of errors in binary testing (contrasted with a correct result, either a true positive or true negative). They are also known in medicine as false positive (respectively negative) diagnoses and in statistical classification as false positive (respectively negative) errors. False positives are distinct from overdiagnosis and from overtesting. For binary classification of errors, users only need two options for labeling error data, e.g., "0" or "1," i.e., false negative or false positive. For binary classification of errors, the target may only be a scalar with two possible entries, e.g., the entries "0" or "1." Alternatively, for binary classification of errors, the target may be a scalar with three possible entries, e.g., false negative, false positive, or correct. The target may also have four possible entries, e.g., false negative, false positive, correct negative, or correct positive. For example, a false negative can be detected when a tube or sample container is incorrectly classified into an error rack. The error rack can include a sample container that is classified as negative.

[0036] Receiving user input may further include providing user information data indicating a plurality of types of operation errors; presenting a menu of the plurality of types of operation errors on the display device by outputting the user information data via the display device; receiving user input via the user interface, the user input indicating a user selection of at least one of the plurality of types of operation errors; and providing labeling data assigned to at least one of the plurality of types of operation errors selected by the user.

[0037] A memory containing multiple types of operation errors may be provided, for example, in a list. In response to receiving a first user input, the list may be shown to the user on the display device. Alternatively, multiple types of operation errors may be stored in the memory in different categories (optionally, different subcategories or sub-subcategories). In response to receiving a first user input, the categories may be shown to the user on the display device. In response to a second user input indicating a category selected by the user, operation errors of the corresponding types in the selected category may be shown to the user on the display device. If several subcategories (sub-subcategories) are provided, further user input (multiple user inputs) may be received until the type of operation error corresponding to the finally selected subcategory (sub-subcategory) is presented / displayed to the user on the display device.

[0038] For example, in the memory, the multiple types of operational errors may be categorized according to operational errors of different functional modules or motion control devices. In particular, the multiple types of operational errors may be categorized into operational errors of analytical devices, handling devices, and automated trucks. The multiple types of operational errors may be stored in the memory such that, for example, for each operational error of an analytical device, handling device, automated truck, and / or motion control device, a corresponding (standard) type of operational error is linked.

[0039] The user interface for receiving user input may include buttons, a joystick, a touch screen, a microphone, a scanning device, and / or a video camera.

[0040] Presenting the menu of multiple types of operational errors on the display device can include presenting a handful of possible standard errors for the user to select from. Additionally, an option can be provided that allows the user to write something else indicating a non-standard error. In response to receiving user input indicating a non-standard error, the non-standard error can be stored in memory. The non-standard error can be linked to at least one of an analytical device, a handling device, an automated truck, and / or an operational control device.

[0041] The user input can indicate a new type of operation error. The new type of operation error need not be stored in memory prior to the user input. In response to the user input indicating the new type of operation error, the new type of operation error can be stored in memory. The new type of operation error may be incorporated into multiple operation errors. The user can input the new type of operation error via, for example, a keypad, a gesture recognition system, or a voice recognition system.

[0042] The user input may also indicate a new type of operation error for one of the categories. The new type of operation error for one of the categories may not be stored in the memory in the corresponding category before the user input. In response to the user input indicating a new type of operation error for one of the categories, the new type of operation error for one of the categories may be stored in the memory. The new type of operation error for one of the categories may be incorporated into the corresponding category in the memory. In other words, the new type of operation error for one of the categories may be linked to this category in the memory. The user may input a new type of operation error for one of the categories (e.g., via a keypad, a gesture recognition system, or a voice recognition system).

[0043] Providing the user information data may include receiving read data from a data carrier reader, the read data indicating sample vessel data stored on a data carrier provided on the sample vessel, and providing the user information data in response to receiving the read data.

[0044] The data carrier reader may include a scanning device, such as a barcode scanner or a QR code scanner. Alternatively, the data carrier reader may include a camera device or an RFID reader. The readout data indicative of the sample vessel data stored in the data carrier provided on the sample vessel may include typical user information data associated with the sample vessel, in particular the type of operational error. These typical types of operational error may be provided, for example, in response to historically detected and reported operational errors associated with the sample vessel. Alternatively, the readout data indicative of the sample vessel data stored in the data carrier provided on the sample vessel may include a unique identification number. For example, in the memory device, one or more unique identification numbers may be associated with each of a plurality of types of operational error. In response to receiving the unique identification number, the operational error associated with the unique identification number may be provided. These operational errors associated with the unique identification number may then be provided / displayed to the user on the display device. Which of the plurality of types of operational error the unique identification number is associated with may depend on the historically detected and reported operational errors associated with the corresponding sample vessel and / or may be preset.

[0045] Each sample vessel can be provided with a corresponding category containing one or more types of operational errors. A category containing one or more types of operational errors corresponding to one of the sample vessels can contain the type of operational error associated with this sample vessel. In the memory, one or more sample vessels can be linked to each of the stored operational errors. In response to receiving the read data, one of the categories corresponding to one of the sample vessels can be selected. The operational errors stored in the selected category can then be provided / displayed to the user on the display device.

[0046] Providing the user information data may include receiving user message data from an input device, the user message data indicating at least one of a user video message input and a user audio message input; and providing the user information data in response to receiving the user message data.

[0047] For example, the memory may be provided with multiple types of operational errors. It may therefore be contemplated that a user may make a pre-selection via a user input device or input devices, for example, before user information data indicative of the (selected) multiple operational errors is output to the display device. The user input device may comprise an input device. A portion of the (overall) multiple operational errors may be provided in response to receiving user message data. In particular, a particular category (or subcategory or sub-subcategory) of the (overall) multiple operational errors (e.g., related to an analytical device, a handling device, an automated truck, or a sample container) may be provided in response to receiving a user message. The user information data may be indicative of a portion (category, subcategory, or sub-subcategory) of the (overall) multiple operational errors.

[0048] The video message input can be provided by a camera. A gesture recognizer may be provided. The input device can include a camera and / or a gesture recognizer. The camera can constantly monitor the user's attention. If the user is not looking in the direction of the camera, the attention can be classified as negative, and the camera is in standby mode. If the attention is classified as positive, i.e., if the user is looking in the direction of the camera, the camera can be turned on. Alternatively, the camera can be always on. In another exemplary embodiment, the camera can be switched between on mode and standby mode by other input devices such as a button, a touch screen, a joystick, or a microphone. When the camera is turned on, a user's photo (video) can be provided and received by the gesture recognizer. The gesture recognizer can recognize the user's gestures (captured by the camera). In response to recognizing the user's gestures, corresponding user (video) message data can be provided by the gesture recognizer. When the camera is in standby mode, the user's photo (video) may not be received by the gesture recognizer. If the user provides gesture language characters, the gesture recognizer can convert the gesture language characters into vtext. The user (video) message data can include vtext.

[0049] The audio message input can be provided by a microphone. A voice recognizer may be provided. The input device can include a microphone and / or a voice recognizer. The microphone can include a standby mode and an on mode. When the microphone is turned on, user audio data can be provided and received by the voice recognizer. In response to receiving the user audio data, corresponding user (audio) message data can be provided by the voice recognizer. When the microphone is in standby mode, user audio data may not be received by the voice recognizer. The voice recognizer can convert the audio data into text. The user (video) message data can include text. The microphone can be always on. Alternatively, the microphone is generally in standby mode and is turned on in response to a specific user input. The specific user input can be a specific spoken phrase. Examples of such spoken phrases can be: "Hello <name of in vitro diagnostic system>," "Bug report," "Error report," etc. Alternatively, the microphone can be switched between the on mode and the standby mode by other input devices such as a button, a touchscreen, a joystick, and / or a camera.

[0050] In the method, providing user information data may include outputting a visual representation of one of the functional modules from the plurality of functional modules and operation control device via the display device, and receiving a user selection input indicating a user selection of the visual representation.

[0051] All of the multiple types of operational errors may be stored in the memory. Furthermore, in the memory, each operational error may or may not be linked to one or more functional modules or operational controllers. In response to a user selection of a visual representation of one of the functional modules or operational controllers, a portion of the overall multiple types of operational errors may be provided. The portion of the (overall) multiple types of operational errors may be selected to include the type of operational error associated with the selected functional module or operational controller. In other words, the multiple types of operational errors are classified into different categories in the memory. In response to a selection of a visual representation of one of the multiple functional modules and an operational controller, one of the categories may be selected. A display device may be provided for user selection. Alternatively, a button, a joystick, a touchscreen, a microphone, a scanning device, and / or a video camera may be provided for user selection. Furthermore, in response to a user selection of a visual representation of one of the functional modules or operational controllers, the user may input a new type of operational error (e.g., via a keyboard, a gesture recognizer, or a voice recognizer).

[0052] In response to a user selecting the visual representation of one of the functional modules or operational controller, the display device may present a menu of a subset of the plurality of types of operational errors, which may indicate a type of operational error linked to the selected functional module or operational controller.

[0053] The method may further include receiving a user error report message within receiving the user input, generating user error report message data indicative of the user error report message, and providing labeling data including the user error report message data. In addition to selecting a type of operation error, the user error report message may be provided. The user error report message may include further information regarding the selected type of operation error. In other words, the user error report message indicates details regarding the selected type of operation error, particularly details not provided by the error data or inherently included in the type of operation error. The user error report message may be a text message, an audio message, an image message, and / or a video message. The user report message data may include text, an audio file, an image, and / or a video. In a preferred embodiment, the user report message data may include only text data. The user input device (providing the user input) may include a voice recognition device with a microphone. The voice recognition device may convert the voice message into a voice-to-text message (indicating a text message). The user input device (providing the user input) may include a gesture recognition device with a camera. The gesture recognizer can convert a (user's) video message into a video-text message (representing a text message). The gesture recognizer can convert a (user's) image message into an image-text message (representing a text message).

[0054] With respect to an in-vitro diagnostic system configured to measure a sample of a body fluid, the embodiments described for a method for detecting and reporting operational errors in an in-vitro diagnostic system may be applied mutatis mutandis.

[0055] Further embodiments will now be described with reference to the figures. [Brief explanation of the drawings]

[0056] [Figure 1] 1 is a schematic diagram of functional modules or components of an in vitro diagnostic system configured to measure a sample of bodily fluid. [Figure 2] FIG. 1 is a schematic block diagram of a method for detecting and reporting operational errors in an in vitro diagnostic system. DETAILED DESCRIPTION OF THE INVENTION

[0057] FIG. 1 shows a schematic diagram of the functional modules or components of an in vitro diagnostic system 1 configured to measure a sample of a body fluid.

[0058] The body fluid is contained in a sample container 2. In particular, for handling and analyzing the sample, a number of functional modules 3 are provided. The number of functional modules 3 may comprise an analyzer 4, a handling system 5, and an automated track 6. The analyzer 4 is configured to measure the sample. For example, a blood glucose level may be determined for the sample. Alternatively or additionally, some other physical or chemical parameter may be determined for the sample received in the sample container.

[0059] A handling system 5 is configured to handle a plurality of sample containers 2, and an automated track 6 is provided by the handling system 5 and configured to transport a plurality of sample containers 2 from a receiving location where the sample containers 2 can be received into a (transport) rack to an analytical device 4, and optionally follow to some output device or location.

[0060] Further, an operation controller 7 is provided. The operation controller 7 is connected to at least one of the functional modules 3 and configured to control the operation of at least one of the functional modules 3. In the example shown in FIG. 1 , the operation controller 7 is functionally connected (for operation control) to the analyzer 4, the handling system 5, and the automated truck 6. For the conduction operation control of the functional modules 3, the operation controller 7 comprises one or more data processors 8 and application software running on the one or more data processors 8.

[0061] The in-vitro diagnostic system 1 further includes an error detection and reporting device 9 for detecting and reporting operational errors. In particular, the error detection and reporting device 9 is provided for detecting operational errors in the operation of at least one of the plurality of functional modules 3 and the operational controller 7. The error detection and reporting device 9 may be implemented, for example, in common with the operational controller 7. The error detection and reporting device 9 may also be implemented, at least in part, by application software running on one or more data processors 8. Alternatively, the error detection and reporting device 9 may be provided separately from the operational controller 7.

[0062] The provided user interface 10 is in communication with the error detection and reporting device 9. Furthermore, an error repository 11 is provided. The error repository 11 is located remotely with respect to both the plurality of functional modules 3 and the operation controller 7. The error repository 11 may be provided at a remote target or server terminal. The error repository 11 may be connectable to multiple in vitro diagnostic systems located at different locations to receive or collect error report data from the multiple in vitro diagnostic systems.

[0063] FIG. 2 shows a schematic block diagram of a method for detecting and reporting an operational error 100 in an in-vitro diagnostic system 1 .

[0064] In step 101, an error in the operation of at least one of the plurality of functional modules 3 and the operation controller 7 is detected. In response thereto, in step 102, error data indicative of the operation error is provided to the error detection and reporting device 9. The error data may include information about which functional module 3 the operation error was detected for. The operation error may be of a type selected from the following: unidentifiable sample container: the sample identification unit was unable to determine the type of sample container or container 2; incorrectly prepared sample container 2: for example, an aliquot was requested but too little material was determined to be available in the sample container 2; the barcode could not be read; and the sample container 2 is not allowed in the in vitro diagnostic system 1.

[0065] In step 103, in response to detecting the operational error, user input is received via the user interface 10. In response to receiving the user input, labeling data is provided in step 104. The labeling data indicates information about the operational error in addition to the error data. The labeling data can indicate at least one of the following: Is the error correct? disturbances in environmental conditions in the laboratory that are not recorded by sensors in the device; The intended use and destination of the material in sample container 2 after the error occurred; The intended workflow of the entire solution when a module throws an error, and More contextual information.

[0066] In steps 105 and 106, error report data including error data and labeling data is provided and transmitted to an error repository 11, which is remotely located relative to both the plurality of functional modules 3 and the operation controller 7. The error report data can be used for error and fault analysis of the plurality of in-vitro diagnostic systems, and some or all of such in-vitro diagnostic systems provide the error report data to the error repository 11. For example, the error report data can provide input data, such as at least one of training data and validation data, for a machine learning process. As a result of such processing by the machine learning process, updated or new application software can be generated and provided to some or all of the plurality of in-vitro diagnostic systems in the field. Thus, centralized error and fault analysis can be performed to improve the operation of the in-vitro diagnostic systems in the field. For such analysis and processing, not only is the error data itself collected and transmitted to the error repository 11, but labeling data generated in response to user input received at the in-vitro diagnostic system implementation site is also collected and transmitted.

Claims

1. 1. A method for detecting and reporting operational errors in an in-vitro diagnostic system (1) configured to measure a sample of a body fluid, the method comprising: providing a plurality of sample vessels (2) each containing a sample of a bodily fluid; A plurality of functional modules (3) is provided, the plurality of functional modules (3) comprising: an analytical device (4) configured to measure the sample; a handling system (5) configured to handle the plurality of sample vessels (2); an automated truck (6) provided by the handling system (5) and configured to transport the plurality of sample containers (2) to the analytical device (4); providing a plurality of functional modules (3) comprising: providing an operation control device (7) connected to at least one of the functional modules (3) and configured to control the operation of the at least one functional module (3), the operation control device (7) comprising one or more data processors (8), wherein application software is running on the one or more data processors (8) to control the operation of the at least one functional module (3); controlling the operation of the at least one functional module (3) by the operation control device (7); Detecting and reporting operational errors by an error detection and reporting device (9), Detecting the operational error in the operation of at least one of the plurality of functional modules (3) and the operation control device (7); providing error data indicative of the operational error; receiving a user input via a user interface (10) after detecting the operational error; in response to receiving the user input, providing labeling data indicative of information about the operational error in addition to the error data; providing error report data including the error data and the labeling data; transmitting the error report data to an error repository (11) located remotely to both the plurality of functional modules (3) and the operation control device (7); Detecting and reporting operational errors, including receiving the error report data in a machine learning process running on a data processing device connected to the error repository (11); processing the error report data by the machine learning process in the data processing device; providing an application software update for the application software in response to the processing of the error report data by the machine learning process in the data processing device; providing the application software update to the operation controller (7); controlling operation of the at least one functional module (3) by the operation controller (7), including executing the application software including the application software update; A method comprising:

2. 2. The method of claim 1, further comprising preprocessing the error report data in the data processing device, wherein the preprocessing comprises formatting the error report data into formatted error report data having a data format that can be processed by the machine learning process.

3. 3. The method of claim 1, wherein the error data indicates an operational error of the analytical device (4).

4. 4. The method according to any one of claims 1 to 3, wherein the error data indicates an operation error of the handling system (5).

5. 5. The method according to claim 1, wherein the error data indicates an operation error of the automated truck (6).

6. The method of claim 1 , wherein the error data indicates one of a false positive error and a false negative error.

7. receiving the user input; providing user information data indicative of a plurality of types of operational errors; outputting the user information data via a display device to present a menu of the plurality of types of operational errors on the display device; receiving the user input via the user interface (10), the user input indicating a user selection of at least one of the plurality of types of operational errors; providing labeling data assigned to the at least one of the plurality of types of motion error selected by the user; 7. The method of claim 1, further comprising:

8. providing the user information data, receiving read data from a data carrier reader, the read data being indicative of sample vessel data stored on a data carrier provided in the sample vessel (2); providing said user information data in response to receiving said read data; The method of claim 7, comprising:

9. providing the user information data, receiving user message data from an input device, the user message data indicating at least one of a user video message input and a user audio message input; providing said user information data in response to receiving said user message data; 9. The method of claim 7 or 8, comprising:

10. providing the user information data, outputting, via the display device, a visual representation of one of the functional modules from the plurality of functional modules and the motion control device; receiving a user selection input indicating that the user has selected the visual representation; 10. The method of any one of claims 7 to 9, comprising:

11. receiving a user error report message in said receiving user input; generating user error report message data indicative of the user error report message; providing the labeling data including the user error report message data; 11. The method of claim 7, further comprising:

12. An in vitro diagnostic system (1) for measuring a sample of a body fluid, comprising: a plurality of sample containers (2) each containing a sample of a body fluid; A plurality of functional modules (3), an analytical device (4) configured to measure the sample; a handling system (5) configured to handle the plurality of sample vessels (2); an automated truck (6) provided by the handling system (5) and configured to transport the plurality of sample containers (2) to the analytical device (4); a plurality of functional modules (3) comprising: A motion control device (7), a control unit connected to at least one of the functional modules (3) and configured to control the operation of the at least one functional module (3); one or more data processors (8), application software running on said one or more data processors (8) for controlling the operation of said at least one functional module (3); A motion control device (7); An error detection and reporting device (9) configured to detect and report operational errors, comprising: Detecting the operational error in the operation of at least one of the plurality of functional modules (3) and the operation control device (7); providing error data indicative of the operational error; receiving a user input via a user interface (10) after detecting the operational error; in response to receiving the user input, providing labeling data indicative of information about the operational error in addition to the error data; providing error report data including the error data and the labeling data; transmitting the error report data to an error repository (11) located remotely to both the plurality of functional modules (3) and the operation control device (7); an error detection and reporting device (9) including: a data processing device connected to the error repository (11), receiving the error report data in a machine learning process running on the data processing device; processing the error report data with the machine learning process; a data processing device configured to provide an application software update for the application software in response to the processing of the error report data by the machine learning process, the application software update being provided to the operation controller (7) for controlling operation of the at least one functional module (3) by the operation controller (7), including executing the application software that includes the application software update; and An in vitro diagnostic system comprising:

Images