A method and apparatus for visual interactive method of a LIMS laboratory system

Abstract

The present application relates to the technical field of data processing, in particular to a kind of LIMS laboratory system visual interactive method and equipment, comprising: obtaining the original detection signal of same batch same type sample in LIMS laboratory system, and is divided into several detection signal sections, all detection signal sections are clustered, reference normal cluster and abnormal cluster are filtered from all cluster, to determine the suspected interference factor and the severity of the original detection signal itself problem, then determine the step factor required by original detection signal, to adaptively filter original detection signal, obtain filtered detection signal, and carry out the visual interaction of LIMS laboratory system.The present application can make detection signal more smooth, stable, characteristic clear, significantly improve the visual display effect.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and specifically to a visualization and interaction method and device for a LIMS laboratory system. Background Technology

[0002] Laboratory Information Management System (LIMS) is a core digital tool supporting laboratory sample flow, data management, process control, and compliance auditing. It is widely used in various laboratory scenarios such as food testing, environmental testing, biological testing, and pharmaceutical testing. LIMS systems typically use charts, dashboards, flowcharts, and other methods to visualize sample testing, equipment status, task progress, and quality data. Data querying and process monitoring are achieved through standard interactive operations such as filtering, drill-down, drag-and-drop adjustments, anomaly alerts, and viewing / exporting.

[0003] Existing problems: Due to the susceptibility of laboratory instrument outputs and sensor signals to electromagnetic interference, voltage fluctuations, environmental noise, poor contact, and other factors, the raw real-time data collected often contains random noise, pulse interference, outliers, jitter, and other issues. If these data are directly fed into the visualization interface or used for control logic, it will cause irregular jitter and distortion in the raw signal curves, reduce the clarity and readability of the visualization, obscure the true detection characteristics, and also cause problems such as abnormal fluctuations in the interactive interface, false triggering, and decreased reliability of automatic judgment. Summary of the Invention

[0004] This invention provides a visualization and interaction method and device for a LIMS laboratory system to solve existing problems.

[0005] The present invention provides a visualization and interaction method and device for a LIMS laboratory system, which adopts the following technical solution:

[0006] One embodiment of the present invention provides a visualization and interaction method for a LIMS laboratory system, the method comprising the following steps:

[0007] Acquire the raw detection signals of the same batch and type of samples within the LIMS laboratory system; divide the raw detection signals into several detection signal segments;

[0008] Based on the differences between the detected signal segments, a robust abnormal score is determined for each detected signal segment; clustering is performed on all detected signal segments to obtain several clusters; based on the magnitude of the robust abnormal score of the detected signal segments in each cluster, a baseline normal cluster and an abnormal cluster are selected from all clusters; based on the differences between the detected signal segments in the baseline normal cluster and the abnormal cluster, combined with the number of abnormal clusters and the number of detected signal segments in the abnormal clusters, the suspected interference factor of the original detected signal is determined;

[0009] Based on the similarity between the detected signal segments in the abnormal clusters and the matching between them and the detected signal segments in the benchmark normal clusters, the severity of the problem in the original detected signal is determined; based on the magnitude of the suspected interference factor and the severity of the problem in the original detected signal, the required step size factor for the original detected signal is determined.

[0010] Based on the step size factor required for the original detection signal, the original detection signal is adaptively filtered to obtain the filtered detection signal, and then visualized and interacted with the LIMS laboratory system.

[0011] Furthermore, the specific steps for determining the robust abnormal score for each detected signal segment are as follows:

[0012] The median of the DTW distance values ​​between each detection signal segment and all other detection signal segments is denoted as the robust abnormal score for each detection signal segment.

[0013] Furthermore, the specific steps for selecting the benchmark normal clusters and abnormal clusters from all clusters are as follows:

[0014] The mean of the robust abnormal scores of all detected signal segments in each cluster is recorded as the abnormal value of each cluster, and the cluster corresponding to the minimum abnormal value is recorded as the baseline normal cluster.

[0015] Clusters other than the baseline normal clusters are denoted as abnormal clusters.

[0016] Furthermore, the specific steps for determining the suspected interference factors of the original detection signal are as follows:

[0017] The total number of all detected signal segments in the original detected signal is denoted as the total number of signal segments;

[0018] The number of all detected signal segments in all abnormal clusters is denoted as the number of abnormal signal segments;

[0019] The ratio of the number of abnormal signal segments to the total number of signal segments is denoted as the percentage of abnormal segments.

[0020] The ratio of the number of anomalous clusters to the total number of clusters is denoted as the percentage of anomalous types.

[0021] In the baseline normal cluster, the detection signal segment corresponding to the minimum robust abnormal score is denoted as the baseline normal detection signal segment;

[0022] The mean DTW distance between the baseline normal detection signal segment and all detection signal segments in all abnormal clusters is denoted as the degree of difference.

[0023] Based on the proportion of abnormal numbers, the proportion of abnormal types, and the degree of difference, the suspected interference factors of the original detection signal are determined.

[0024] Furthermore, the specific steps for determining the suspected interference factors of the original detection signal based on the proportion of abnormal quantity, the proportion of abnormal type, and the degree of difference are as follows:

[0025] The product of the mean of the percentage of abnormal numbers and the percentage of abnormal types and the degree of difference is recorded as the suspected interference factor of the original detection signal.

[0026] Furthermore, the specific steps involved in determining the severity of the inherent problems in the original detection signal are as follows:

[0027] In all abnormal clusters and all detected signal segments, the inverse proportional normalized value of the minimum DTW distance value between each detected signal segment and all other detected signal segments is denoted as the probability of the existence of quality abnormality for each detected signal segment.

[0028] Any detection signal segment in all abnormal clusters is designated as the reference detection signal segment;

[0029] Based on the matching between the reference detection signal segment and the baseline normal detection signal segment, determine the abnormal continuous occurrence performance value of the reference detection signal segment;

[0030] The maximum value between the continuous occurrence of anomalies in the reference detection signal segment and the probability of its own quality anomalies is recorded as the self-problem performance value of the reference detection signal segment.

[0031] The mean of the self-problem performance values ​​of all detection signal segments in all abnormal clusters is denoted as the self-problem severity of the original detection signal.

[0032] Furthermore, the specific steps for determining the abnormal continuous occurrence performance value of the reference detection signal segment are as follows:

[0033] The reference detection signal segment is matched with the baseline normal detection signal segment to obtain an optimal matching path, and each node on the optimal matching path corresponds to a set of matching distances;

[0034] On the optimal matching path, the mean and standard deviation of the matching distance of all nodes are obtained. The standard deviation of the matching distance is multiplied by a preset coefficient, and the product is added to the mean of the matching distance. The sum is defined as the segmentation threshold. Nodes whose matching distance is greater than the segmentation threshold are recorded as abnormal nodes, and abnormal path segments are formed by adjacent abnormal nodes. The length of each abnormal path segment is calculated, and the mean of the lengths of all abnormal path segments is recorded as the length threshold. The sum of the lengths of all abnormal path segments is recorded as the total length. The sum of the lengths of all abnormal path segments whose lengths are greater than the length threshold is recorded as the continuous length. The ratio of the continuous length to the total length is recorded as the abnormal continuous occurrence performance value of the reference detection signal segment.

[0035] Furthermore, the specific steps involved in determining the step size factor required for the original detection signal are as follows:

[0036] Based on the severity of the inherent problems and suspected interference factors of the original detection signal, determine the required step size factor adjustment coefficient for the original detection signal;

[0037] The difference between the upper limit and the lower limit of the preset step size factor is recorded as the step size factor range value. The sum of the product of the step size factor adjustment coefficient required for the original detection signal and the step size factor range value and the preset lower limit of the step size factor is recorded as the step size factor required for the original detection signal.

[0038] Furthermore, the specific steps involved in determining the step size factor adjustment coefficient required for the original detection signal are as follows:

[0039] For the original detection signal, the product of the preset constant and the sum of its own problem severity and the suspected interference factor is recorded as the filtering effect requirement coefficient. The inverse proportional normalized value of the filtering effect requirement coefficient is recorded as the step size factor adjustment coefficient required for the original detection signal.

[0040] The present invention also proposes a visualization and interactive device for a LIMS laboratory system, including a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the computer program stored in the memory to implement the steps of the aforementioned visualization and interactive method for a LIMS laboratory system.

[0041] The beneficial effects of the technical solution of the present invention are:

[0042] In this embodiment of the invention, the original detection signals of samples of the same batch and type within the LIMS laboratory system are acquired and divided into several detection signal segments. Based on the differences between the detection signal segments, a robust abnormality score is determined for each segment. Clustering is performed on all detection signal segments to obtain several clusters. Based on the robust abnormality scores of the detection signal segments within each cluster, a baseline normal cluster and anomaly clusters are selected from all clusters. Based on the differences between the detection signal segments in the baseline normal clusters and the anomaly clusters, combined with the number of anomaly clusters and the number of detection signal segments within each anomaly cluster, the suspected interference factor of the original detection signal is determined. Thus, based on the high similarity of detection signals from the same type and batch of samples, the severity of suspected noise interference is analyzed to ensure the denoising effect of the filtering. Based on the similarity between detection signal segments in the anomaly clusters and their matching with those in the baseline normal clusters, the severity of the inherent problems in the original detection signal is determined. This analyzes the problems inherent in the detection sample itself, ensuring that the sample's inherent quality abnormalities are preserved during filtering. Based on the suspected interference factor and the severity of the problem in the original detection signal, the required step size factor is determined for adaptive filtering of the original detection signal. The filtered detection signal is then obtained and visualized interactively within the LIMS laboratory system. Thus, this invention, by improving the filtering effect of the original detection signal, effectively removes environmental interference, circuit noise, and random fluctuations, making the detection signal smoother, more stable, and with clearer characteristics. On the one hand, it significantly improves the visualization effect, enhancing the readability, standardization, and traceability of the signal curve. On the other hand, it greatly improves the reliability, smoothness, and automation of system interaction, reducing the probability of human intervention and misjudgment, thereby improving the accuracy, stability, and efficiency of the entire rapid detection process. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 This is a flowchart illustrating the steps of a visualization and interaction method for a LIMS laboratory system according to the present invention.

[0045] Figure 2 This is a schematic diagram of the structure of a visualization and interactive device for a LIMS laboratory system. Detailed Implementation

[0046] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a visualization and interactive method and device for a LIMS laboratory system proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0047] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0048] The following description, in conjunction with the accompanying drawings, details a specific scheme for a visualization and interactive method and equipment for a LIMS laboratory system provided by the present invention.

[0049] Please see Figure 1 The diagram illustrates a flowchart of a visualization and interaction method for a LIMS laboratory system according to an embodiment of the present invention. The method includes the following steps:

[0050] Step S001: Obtain the original detection signals of the same batch and type of samples within the LIMS laboratory system; divide the original detection signals into several detection signal segments.

[0051] It should be noted that most current Laboratory Information Management Systems (LIMS) are designed for specific niche areas, exhibiting strong targeting but weak general applicability. For example, LIMS for food testing mainly focus on food safety indicators, testing standards, sample flow, and compliance reporting; while LIMS for biological testing and medical laboratory testing emphasize biological sample management, experimental process traceability, quality control systems, and biosafety regulations. These systems are deeply industry-dependent and highly customized, making them difficult to reuse across different fields.

[0052] In this embodiment, within the LIMS system for rapid testing of agricultural products, sample information, origin traceability, production line processes, raw data from rapid testing instruments, reagent kits and quality control information, test results, personnel operation and qualified disposal data are collected. This enables automatic data collection, real-time judgment and full traceability throughout the entire process from sampling to certification. This is the standard basic operation for applying existing LIMS systems to the field of rapid testing of agricultural products.

[0053] The time-series data types that need to be filtered in the collected data include the raw detection signals of rapid testing instruments in the rapid testing line for agricultural products, the operating status parameters of the line equipment, and the time-series data of ambient temperature and humidity.

[0054] For non-core time-series data such as operating parameters of production line equipment and ambient temperature and humidity, moving average filtering can be directly used to meet the operational monitoring requirements. Moving average filtering is a well-known technique, and the specific method will not be described here. However, for the raw detection signals output by rapid testing instruments, since they directly determine the accuracy and reliability of agricultural product testing results, even small fluctuations in the signal can easily lead to deviations in the test results. Therefore, the following more precise adaptive filtering process is required.

[0055] It should be noted that rapid testing lines for agricultural products typically employ batch and continuous testing, performing uninterrupted rapid testing on a large number of similar agricultural products. In this embodiment, for a large number of samples of the same batch and type, the original detection signals and full data from the rapid testing instrument are first collected completely. After data verification and format parsing, the data is then uniformly visualized. This method ensures the integrity, validity, and reliability of the test data, avoids incomplete display issues caused by segmented data transmission or real-time rendering, improves system stability and result credibility, and meets the compliance and traceability requirements of rapid testing for agricultural products. The full data refers to all unedited original detection signals and related auxiliary information (including but not limited to: sample information, origin traceability, production line procedures, filtered production line equipment operating status parameters, and environmental temperature and humidity time-series data) output by the rapid testing instrument during a complete testing process for samples of the same batch and type.

[0056] It should be further explained that in this embodiment, apples are used as samples, and the detection signal is an absorbance signal (horizontal axis is time, vertical axis is absorbance), with a sampling frequency of 100 Hz. For each apple, it passes uniformly through the detection spot from left to right. The instrument collects 100 absorbance points per second, forming a continuous time-absorbance curve. The first part of the curve corresponds to the absorbance of the apple's head (stem area), the middle part to the absorbance of the central flesh area, and the last part to the absorbance of the apple's tail (calyx area). Fluctuations on the curve generally correspond to surface defects, bruises, rot, color differences, and internal lesions. In this embodiment, the LIMS system sends data acquisition commands to the edge computing node, which is responsible for processing the 100Hz raw detection signal from the rapid testing instrument. The purpose of this processing is to filter out environmental noise while retaining the subtle signal characteristics of true quality anomalies in agricultural products, ensuring the accuracy of subsequent system archiving, judgment, and visual traceability data.

[0057] For the original detection signals of samples of the same type in the same batch, the original detection signals are divided into several detection signal segments according to the timestamp after the detection of each sample is completed. Each detection signal segment corresponds to one sample.

[0058] It should be noted that in the rapid testing line for agricultural products, the rapid testing instruments can collect the timestamp of each sample after the test is completed.

[0059] Step S002: Based on the differences between the detected signal segments, determine the robust abnormal score for each detected signal segment; perform clustering operations on all detected signal segments to obtain several clusters; based on the magnitude of the robust abnormal score of the detected signal segments in each cluster, select the baseline normal cluster and the abnormal cluster from all clusters; based on the differences between the detected signal segments in the baseline normal cluster and the abnormal cluster, combined with the number of abnormal clusters and the number of detected signal segments in the abnormal clusters, determine the suspected interference factor of the original detected signal.

[0060] It should be noted that because the tested samples are of the same type, from the same batch, and have similar morphology and composition, the original detection signals obtained by the rapid testing instrument for each sample are highly similar. Correspondingly, during continuous testing, the original detection signals output by the rapid testing instrument generally exhibit a temporal characteristic of approximately periodic variation. Therefore, the greater the difference between the detection signals of different samples, the greater the possibility of noise interference, and thus the higher the requirements for filtering effectiveness.

[0061] Therefore, the robustness score of each detection signal segment is first determined based on the differences between the detection signal segments. That is, the more similar a detection signal segment is to all other detection signal segments, the more likely it is to be a detection signal segment of a normal sample. Conversely, the greater the difference between a detection signal segment and all other detection signal segments, the more likely it is to be a detection signal segment of an abnormal sample, and the higher the robustness score will be.

[0062] Then, the DTW algorithm is used to obtain the DTW distance value between any two detected signal segments.

[0063] It should be noted that the DTW (Dynamic Time Warping) algorithm is a well-known technique, and its specific method will not be described here. The smaller the DTW distance value, the more similar the two detected signal segments are.

[0064] Using the DTW distance between any two detected signal segments as the clustering distance, the K-centroid clustering algorithm is used to cluster all detected signal segments to obtain several clusters.

[0065] The K-centroid clustering algorithm is a well-known technique, and its specific method will not be described here.

[0066] Since the smaller the robust abnormal score of the detection signal segment, the more likely it is to correspond to a normal sample, the benchmark normal cluster and abnormal cluster can be screened from all clusters based on the size of the robust abnormal score of the detection signal segment in each cluster.

[0067] Each anomalous cluster represents a type of anomalous event. If no anomalous cluster exists, the step size factor adjustment coefficient for the subsequent raw detection signal is set to 1.

[0068] Finally, during the detection process, the more numerous and diverse the abnormal detection signal segments, and the more significant their differences from normal detection signal segments, the higher the probability of interference. Therefore, the suspected interference factors of the original detection signal can be determined based on the differences between the detection signal segments in the baseline normal clusters and the abnormal clusters, combined with the number of abnormal clusters and the number of detection signal segments within the abnormal clusters.

[0069] Step S003: Determine the severity of the problem in the original detection signal based on the similarity between the detection signal segments in the abnormal clusters and the matching between the detection signal segments and the benchmark normal clusters; determine the required step size factor for the original detection signal based on the magnitude of the suspected interference factor and the severity of the problem.

[0070] It should be noted that significant differences from the baseline normal detection signal segment can be caused by environmental noise, instrument interference, or even by inherent quality abnormalities in the sample itself (such as surface blemishes, dents, or rot on an apple). Therefore, a larger suspected interference factor indicates a greater likelihood of noise interference, requiring better filtering. If the larger suspected interference factor is caused by inherent quality abnormalities in the sample, the requirements for filtering are even higher; that is, the filtering must both ensure noise reduction and preserve the inherent quality abnormalities of the sample.

[0071] It is further important to clarify that the signal difference between normal and problematic samples is a real detection difference that is regular, consistent, and reproducible. In contrast, the signal difference between normal samples and normal samples containing noise is a random, irregular, and unreproducible interference difference. The two are fundamentally different in their source, regularity, reproducibility, and practical significance. Based on the reproducibility of similar problems in samples, when there are highly similar detection signal segments in all abnormal clusters, it indicates that the problem is more likely to be with the original detection signal itself. However, if a certain quality problem occurs only once in the same batch of samples of the same type, it may lead to analytical errors, mistaking it for noise interference, when in fact it is a quality anomaly within the sample itself. Signal anomalies caused by inherent sample quality problems often appear continuously within a signal segment, while signal fluctuations caused by noise interference appear only randomly, intermittently, and discretely.

[0072] Therefore, based on the analysis of similarity between signal segments detected in abnormal clusters, and combined with the matching between signal segments detected in abnormal clusters and signal segments detected in benchmark normal clusters, the severity of the problem in the original detection signal can be determined.

[0073] It should be noted that in this embodiment, LMS adaptive filtering (Least Mean Square) is used to filter the original detection signal. This is a well-known technique, and the specific method will not be described here. In LMS adaptive filtering, a larger step size factor allows for faster adaptation to the dynamic changes of the signal in the pipeline, preventing filtering lag. However, the filtered signal will exhibit significant "jitter," with unnecessary fluctuations superimposed on the originally smooth real signal. It may even filter out effective signals as noise, leading to deviations in the detection results. Conversely, a smaller step size factor results in a very smooth filtered signal with well-suppressed noise, good repeatability of the rapid detection results, and a more gradual and conservative filtering process. It is less likely to filter out sudden changes, anomalies, and differences as noise, preserving the true characteristics of problematic samples. However, convergence is slow, requiring more iterations to reach a steady state, resulting in a slower response.

[0074] The greater the suspected interference factor and the severity of the problem in the original detection signal, the higher the requirements for denoising effect and preservation of the abnormal information features of the problem itself, i.e., a smaller step size factor is needed. Therefore, the required step size factor for the original detection signal can be determined based on the magnitude of the suspected interference factor and the severity of the problem.

[0075] Step S004: Based on the step size factor required for the original detection signal, perform adaptive filtering on the original detection signal to obtain the filtered detection signal, and perform visualization interaction on the LIMS laboratory system.

[0076] Based on the step size factor required for the original detection signal, LMS adaptive filtering is performed on the original detection signal to obtain the filtered detection signal.

[0077] Based on the filtered detection signals of samples of the same type and batch, combined with the corresponding full data, a visual interaction of the LIMS laboratory system is performed.

[0078] It should be noted that the visual interaction in LIMS (Laboratory Information Management System) is a well-known technique in the field. Existing LIMS systems typically use charts, dashboards, flowcharts, and other methods to visualize sample testing, equipment status, task progress, and quality data. Data querying and process monitoring are achieved through conventional interactive operations such as filtering, drill-down, drag-and-drop adjustment, anomaly alerts, and viewing / exporting. The aforementioned visualization and interaction methods are all conventional techniques well-known to those skilled in the art.

[0079] Preferably, in some possible implementations of the embodiments of the present invention, the median of the DTW distance values ​​between each detection signal segment and all other detection signal segments is recorded as the robust abnormal score of each detection signal segment.

[0080] The median is chosen to avoid interference from a few abnormal detection signal segments or bad sample detection signal segments.

[0081] Preferably, in some possible implementations of the embodiments of the present invention, the mean of the robust abnormal scores of all detection signal segments in each cluster is recorded as the abnormal value of each cluster, and the cluster corresponding to the smallest abnormal value is recorded as the benchmark normal cluster.

[0082] If there are multiple minimum abnormal values, then all clusters corresponding to the minimum abnormal values ​​are denoted as the baseline normal clusters.

[0083] Clusters other than the baseline normal clusters are denoted as abnormal clusters.

[0084] Preferably, in some possible implementations of the embodiments of the present invention, the number of all detected signal segments in the original detected signal is denoted as the total number of signal segments. The number of all detected signal segments in all abnormal clusters is denoted as the number of abnormal signal segments, and the ratio of the number of abnormal signal segments to the total number of signal segments is denoted as the percentage of abnormal numbers. The ratio of the number of anomalous clusters to the total number of clusters is denoted as the anomaly rate. .

[0085] Among them, the more abnormal clusters there are, the higher the proportion of abnormal types. The larger the value, the higher the proportion of outlier clusters. This is because the baseline normal cluster size is typically 1, so as the number of outlier clusters increases, the proportion of outlier types also increases. In order: , , , ..., these are sequentially increasing.

[0086] In the baseline normal cluster, the detection signal segment corresponding to the smallest robust abnormal score is denoted as the baseline normal detection signal segment.

[0087] If there are multiple detection signal segments corresponding to the minimum robust abnormal scores, then any one of them shall be selected as the benchmark normal detection signal segment.

[0088] The mean DTW distance between the baseline normal detection signal segment and all detection signal segments in all abnormal clusters is denoted as the degree of difference. .

[0089] Preferably, in some possible implementations of the embodiments of the present invention, the product of the average of the proportion of abnormal quantity and the proportion of abnormal type and the degree of difference is recorded as the suspected interference factor of the original detection signal. The specific calculation formula is: Suspected interference factor of the original detection signal ,in, and All are dimensionless data values ​​between 0 and 1, therefore Dimensionless data values ​​between 0 and 1 can be used as The adjustment value.

[0090] Preferably, in some possible implementations of the embodiments of the present invention, any one of the detection signal segments in all abnormal clusters is designated as the reference detection signal segment.

[0091] The DTW algorithm is used to match the reference detection signal segment with the baseline normal detection signal segment to obtain an optimal matching path. Each node on the optimal matching path corresponds to a set of matching distances.

[0092] The default coefficient is 3, and we will use this as an example for explanation.

[0093] The average of the matching distances of all nodes along the optimal matching path. with standard deviation ,Will Let be the segmentation threshold. Nodes whose matching distance is greater than the segmentation threshold are designated as abnormal nodes, and adjacent abnormal nodes form an abnormal path segment. An isolated single abnormal node also constitutes an abnormal path segment.

[0094] On the optimal matching path, the length of each abnormal path segment is calculated. The average length of all abnormal path segments is recorded as the length threshold. The sum of the lengths of all abnormal path segments is recorded as the total length. The sum of the lengths of all abnormal path segments whose lengths are greater than the length threshold is recorded as the continuous length. The ratio of the continuous length to the total length is recorded as the abnormal persistence performance value of the reference detection signal segment. The abnormal persistence performance value is a dimensionless data value between 0 and 1.

[0095] It should be noted that: if there are no abnormal path segments, the abnormal continuous occurrence performance value of the reference detection signal segment is directly set to 0. If there are abnormal path segments, the total length will definitely not be 0, so the ratio of continuous length to total length is valid.

[0096] Preferably, in some possible implementations of the embodiments of the present invention, the minimum DTW distance value between each detection signal segment and all other detection signal segments is obtained among all detection signal segments in all abnormal clusters. The inversely proportional normalized value is used as the probability of inherent quality anomalies for each detected signal segment.

[0097] It should be noted that in this embodiment, Let be the most similar value of each detection signal segment in all anomaly clusters. Using the min-max normalization method, normalize the most similar values ​​of all detection signal segments in all anomaly clusters to a range between 0 and 1. Subtract 1 from 1. The difference between the normalized values ​​is used as The inversely proportional normalized value. The minimum-maximum normalization method is a well-known technique, and its specific method will not be described here. The smaller the value, the more likely the same quality problem will occur repeatedly in the same batch of samples of the same type, meaning that the probability of it having quality abnormalities is greater.

[0098] The maximum value between the persistent abnormality performance value of the reference detection signal segment and the probability of its own quality abnormality is denoted as the self-problem performance value of the reference detection signal segment. The self-problem performance value is a dimensionless data value between 0 and 1.

[0099] In all abnormal clusters and all detected signal segments, the mean of the self-problem performance values ​​of all detected signal segments is obtained, and denoted as the self-problem severity of the original detected signal. The severity of the problem itself is a dimensionless data value between 0 and 1.

[0100] Preferably, in some possible implementations of the embodiments of the present invention, for the original detection signal, the product of the sum of the preset constant and the severity of its own problem and the suspected interference factor is recorded as the filtering effect requirement coefficient, and the inverse proportional normalized value of the filtering effect requirement coefficient is recorded as the step size factor adjustment coefficient required for the original detection signal.

[0101] In this embodiment, the preset constant is 1. Taking this as an example, the required coefficient for the filtering effect is... Among them, the severity of its own problems For dimensionless data values ​​between 0 and 1, therefore, can be used right Adjustments are made. The required step size factor adjustment coefficient for the original detection signal. This embodiment uses To present The inverse proportional relationship and normalization processing are determined by the implementer, who can set the inverse proportional function and normalization function according to the actual situation. The standard deviation of the DTW distance values ​​between the baseline normal detection signal segment and all detection signal segments in all abnormal clusters is denoted as the exponential decay scaling factor. , used to eliminate dimensions.

[0102] Preferably, in some possible implementations of the embodiments of the present invention, the difference between the upper limit of the preset step size factor and the lower limit of the preset step size factor is recorded as the step size factor range value, and the sum of the product of the step size factor adjustment coefficient required for the original detection signal and the step size factor range value and the preset step size factor lower limit is recorded as the step size factor required for the original detection signal.

[0103] In this embodiment, based on the empirical calibration range of the sensor of this type of rapid detection instrument at the current production line operating speed, the preset lower limit of the step size factor is 0.001, and the preset upper limit of the step size factor is 0.01. Taking this as an example, the required step size factor for the original detection signal is... .

[0104] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0105] The present invention also provides a visualization and interactive device for a LIMS laboratory system, including a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the computer program stored in the memory to implement the steps of the aforementioned visualization and interactive method for a LIMS laboratory system.

[0106] In this embodiment, a schematic diagram of the structure of a visualization and interactive device for a LIMS laboratory system is shown, as follows: Figure 2 As shown. Figure 2 The system comprises: a physical interaction layer, a data transmission layer, a LIMS interface layer, and a visualization layer. The physical interaction layer consists of hardware entities, including: a large touchscreen or multi-screen display, an industrial tablet, a barcode scanner, an NFC reader, a physical button panel, a biometric module (fingerprint or face recognition), and an indicator light array. The data transmission layer includes: an edge computing unit, a network security module, and a protocol conversion gateway, responsible for encrypting and transmitting interaction commands, device data, and sample IDs to the LIMS server. The LIMS interface layer (software interface) includes: a RESTful API interface, a WebSocket real-time channel, a LIMS database driver, and an authentication module, enabling task synchronization with the LIMS system, sample status query, data writing, and report retrieval. The visualization layer includes: an overview dashboard (equipment utilization, task progress, and anomaly alarms), a sample tracking interface (process nodes, responsible persons, and timestamps), an equipment control panel (start or stop, parameter settings, and calibration records), report or chart modules (bar charts, line charts, and heatmaps), and multi-platform adaptation (responsive layout for large screens, tablets, and mobile devices).

[0107] This invention is now complete.

[0108] In summary, in this embodiment of the invention, the original detection signals of samples of the same batch and type within the LIMS laboratory system are acquired and divided into several detection signal segments. Based on the differences between the detection signal segments, a robust abnormal score is determined for each detection signal segment. Clustering is performed on all detection signal segments to obtain several clusters. Based on the magnitude of the robust abnormal score of the detection signal segments in each cluster, a baseline normal cluster and anomaly clusters are selected from all clusters. Based on the differences between the detection signal segments in the baseline normal clusters and the anomaly clusters, combined with the number of anomaly clusters and the number of detection signal segments in the anomaly clusters, a suspected interference factor of the original detection signal is determined. Based on the similarity between the detection signal segments in the anomaly clusters and their matching with the detection signal segments in the baseline normal clusters, the severity of the problem in the original detection signal is determined. Based on the magnitude of the suspected interference factor and the severity of the problem in the original detection signal, the required step size factor for the original detection signal is determined. Based on the required step size factor, adaptive filtering is performed on the original detection signal to obtain the filtered detection signal, and then the visualization interaction of the LIMS laboratory system is performed. This invention enables the detection signal to be smoother, more stable, and more clearly characterized, significantly improving the visualization effect.

[0109] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A visualization and interaction method for a LIMS laboratory system, characterized in that, The method includes the following steps: Acquire the raw detection signals of the same batch and type of samples within the LIMS laboratory system; divide the raw detection signals into several detection signal segments; Based on the differences between the detected signal segments, a robust abnormal score is determined for each detected signal segment; clustering is performed on all detected signal segments to obtain several clusters; based on the magnitude of the robust abnormal score of the detected signal segments in each cluster, a baseline normal cluster and an abnormal cluster are selected from all clusters; based on the differences between the detected signal segments in the baseline normal cluster and the abnormal cluster, combined with the number of abnormal clusters and the number of detected signal segments in the abnormal clusters, the suspected interference factor of the original detected signal is determined; Based on the similarity between the detected signal segments in the abnormal clusters and the matching between them and the detected signal segments in the benchmark normal clusters, the severity of the problem in the original detected signal is determined; based on the magnitude of the suspected interference factor and the severity of the problem in the original detected signal, the required step size factor for the original detected signal is determined. Based on the step size factor required for the original detection signal, the original detection signal is adaptively filtered to obtain the filtered detection signal, and then visualized and interacted with the LIMS laboratory system.

2. The visualization and interaction method for a LIMS laboratory system according to claim 1, characterized in that, The specific steps for determining the robust abnormal score for each detected signal segment are as follows: The median of the DTW distance values ​​between each detection signal segment and all other detection signal segments is denoted as the robust abnormal score for each detection signal segment.

3. The visualization and interaction method for a LIMS laboratory system according to claim 1, characterized in that, The specific steps involved in selecting the baseline normal clusters and abnormal clusters from all clusters are as follows: The mean of the robust abnormal scores of all detected signal segments in each cluster is recorded as the abnormal value of each cluster, and the cluster corresponding to the minimum abnormal value is recorded as the baseline normal cluster. Clusters other than the baseline normal clusters are denoted as abnormal clusters.

4. The visualization and interaction method for a LIMS laboratory system according to claim 1, characterized in that, The specific steps for determining the suspected interference factors of the original detection signal are as follows: The total number of all detected signal segments in the original detected signal is denoted as the total number of signal segments; The number of all detected signal segments in all abnormal clusters is denoted as the number of abnormal signal segments; The ratio of the number of abnormal signal segments to the total number of signal segments is denoted as the percentage of abnormal segments. The ratio of the number of anomalous clusters to the total number of clusters is denoted as the percentage of anomalous types. In the baseline normal cluster, the detection signal segment corresponding to the minimum robust abnormal score is denoted as the baseline normal detection signal segment; The mean DTW distance between the baseline normal detection signal segment and all detection signal segments in all abnormal clusters is denoted as the degree of difference. Based on the proportion of abnormal numbers, the proportion of abnormal types, and the degree of difference, the suspected interference factors of the original detection signal are determined.

5. The visualization and interaction method for a LIMS laboratory system according to claim 4, characterized in that, The specific steps for determining the suspected interference factors of the original detection signal based on the proportion of abnormal quantity, the proportion of abnormal type, and the degree of difference are as follows: The product of the mean of the percentage of abnormal numbers and the percentage of abnormal types and the degree of difference is recorded as the suspected interference factor of the original detection signal.

6. The visualization and interaction method for a LIMS laboratory system according to claim 4, characterized in that, The specific steps involved in determining the severity of the inherent problems in the original detection signal are as follows: In all abnormal clusters and all detected signal segments, the inverse proportional normalized value of the minimum DTW distance value between each detected signal segment and all other detected signal segments is denoted as the probability of the existence of quality abnormality for each detected signal segment. Any detection signal segment in all abnormal clusters is designated as the reference detection signal segment; Based on the matching between the reference detection signal segment and the baseline normal detection signal segment, determine the abnormal continuous occurrence performance value of the reference detection signal segment; The maximum value between the continuous occurrence of anomalies in the reference detection signal segment and the probability of its own quality anomalies is recorded as the self-problem performance value of the reference detection signal segment. The mean of the self-problem performance values ​​of all detection signal segments in all abnormal clusters is denoted as the self-problem severity of the original detection signal.

7. The visualization and interaction method for a LIMS laboratory system according to claim 6, characterized in that, The specific steps involved in determining the abnormal persistent occurrence value of the reference detection signal segment are as follows: The reference detection signal segment is matched with the baseline normal detection signal segment to obtain an optimal matching path, and each node on the optimal matching path corresponds to a set of matching distances; On the optimal matching path, the mean matching distance and standard deviation of the matching distance of all nodes are obtained. The standard deviation of the matching distance is multiplied by a preset coefficient, and the product is added to the mean matching distance. The sum is defined as the segmentation threshold. Nodes whose matching distance is greater than the segmentation threshold are recorded as abnormal nodes, and adjacent abnormal nodes form an abnormal path segment. The length of each abnormal path segment is calculated. The average length of all abnormal path segments is recorded as the length threshold. The sum of the lengths of all abnormal path segments is recorded as the total length. The sum of the lengths of all abnormal path segments whose lengths are greater than the length threshold is recorded as the continuous length. The ratio of the continuous length to the total length is recorded as the abnormal continuous occurrence performance value of the reference detection signal segment.

8. The visualization and interaction method for a LIMS laboratory system according to claim 1, characterized in that, The specific steps involved in determining the step size factor required for the original detection signal are as follows: Based on the severity of the inherent problems and suspected interference factors of the original detection signal, determine the required step size factor adjustment coefficient for the original detection signal; The difference between the upper limit and the lower limit of the preset step size factor is recorded as the step size factor range value. The sum of the product of the step size factor adjustment coefficient required for the original detection signal and the step size factor range value and the preset lower limit of the step size factor is recorded as the step size factor required for the original detection signal.

9. The visualization and interaction method for a LIMS laboratory system according to claim 8, characterized in that, The specific steps involved in determining the step size factor adjustment coefficient required for the original detection signal are as follows: For the original detection signal, the product of the preset constant and the sum of its own problem severity and the suspected interference factor is recorded as the filtering effect requirement coefficient. The inverse proportional normalized value of the filtering effect requirement coefficient is recorded as the step size factor adjustment coefficient required for the original detection signal.

10. A visualization and interactive device for a LIMS laboratory system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the computer program is executed by the processor, it implements the steps of a visualization and interaction method for a LIMS laboratory system as described in any one of claims 1-9.

Images