Visual management method and system for drug IND stage research progress

Abstract

The invention relates to the technical field of data visualization, in particular to a drug IND stage research progress visual management method and system, and aims to extract drug effect test node sequence offset, identify delay sections and dense abnormal nodes, calculate task urgency and generate a drug IND stage task layer visual structure set. According to the method, through comparison of sequence numbers of task nodes on a plan level and an execution level, extraction of node offset directions and construction of a closed graph layer based on offset aggregation, clustering and boundary expression of time sequence abnormal nodes can be realized, and a structured region instead of isolated abnormal points can be formed; the calculation of urgent gradient values of plan time and residual periods is introduced into nodes in the areas, continuous numerical values form a mapping reference, and gray rendering is correspondingly established with parameters of a brightness interval, so that the time pressure is visually expressed in the visual dimension, and a traditional node color marking single expression mode is replaced.

Description

Technical Field

[0001] The present invention relates to the technical field of data visualization, and particularly to a method and system for visual management of the research progress in the IND stage of drugs. Background Art

[0002] The technical field of data visualization includes the processes of graphically transforming, displaying, and interacting with various types of data. The core content of this technical field is to convert raw data into visual graphs that are easy to understand and analyze through means such as image processing, graph generation, and data mapping, to assist users in identifying trends, comparing information, and making decisions.

[0003] Among them, the method for visual management of the research progress in the IND stage of drugs refers to a means of visually expressing and process managing the time nodes, task execution, data submission, and approval processes involved during the research preparation stage before a new drug clinical trial. It mainly covers the setting of data collection rules, the time-axis mapping of research nodes, the construction of status tracking logic for key tasks, and the graphical annotation of document progress, and is specifically completed through methods such as time series analysis, setting of task completion identification rules, and structured mapping of research process forms.

[0004] The prior art starts from the visual presentation of research tasks and realizes the display of project status through the annotation of processes such as time nodes, task execution, and data submission. However, it fails to quantitatively express the hidden problems such as structural delays, sequential offsets, execution intensification, and sample anomalies that occur in the process, and also lacks the ability to perform closed modeling on the task offset structure. In the case where the execution node order is misaligned, often only the changes in individual node information can be observed, and it is difficult to discover the phased time misorder of the node group, further resulting in the difficulty of overall identification of task slippage; due to the lack of an urgency mapping mechanism based on continuous values, it is impossible to distinguish high-urgency tasks from regular tasks in the graphical dimension, resulting in fuzzy scheduling judgment; the time overlap between nodes depends on manual identification, and it is easy to miss high-frequency parallel task groups. The sample number management is mainly based on static field records and cannot express its internal sequential logic or abnormal status in combination with the graphical interface. For example, sample skipping numbers are often missed or misclassified into the category of data problems, lacking the ability of structural positioning, making it difficult to trace and locate problems. Summary of the Invention

[0005] The purpose of the present invention is to solve the deficiencies existing in the prior art and to propose a method and system for visual management of the research progress in the IND stage of drugs.

[0006] To achieve the above purpose, the present invention adopts the following technical solutions: A method for visual management of the research progress in the IND stage of drugs, including the following steps:

[0007] S1: Obtain the sequential number of the pre-clinical efficacy test node in the original plan during the drug IND phase research task process, compare the numbers according to the planned and execution sequences, and generate a task node sequence offset set;

[0008] S2: Judge the offset direction of the pre-clinical efficacy test node in the task node sequence offset set, identify the offset continuous node group and screen the delayed nodes, construct a closed boundary layer, and generate a delay section layer structure set;

[0009] S3: Obtain the planned completion date, the current system time, and the remaining execution duration of the pre-clinical efficacy test node in the delay section layer structure set to calculate the task urgency gradient value, match the gray scale brightness interval to render the node tile, and generate a task urgency gray code layer;

[0010] S4: Obtain the start and end times of the pharmacokinetic detection node corresponding to the pre-clinical efficacy test node in the task urgency gray code layer, count the time overlap rate to judge the dense node pair, and obtain a node dense area structure set;

[0011] S5: Obtain the sample number sequence bound to each node pair in the node dense area structure set, compare the continuity of the sample numbers to identify the abnormal node pair, and generate a visualization structure set for the drug IND phase task layer.

[0012] The improvement of the present invention is that the task node sequence offset set includes a sequence number mapping relationship, a sequence misalignment direction identifier, and a node execution sequence index; the delay section layer structure set includes a delay aggregation section range identifier, a layer closed boundary box line parameter, and a task number offset screening record; the task urgency gray code layer includes a task urgency level interval mapping table, a gray scale brightness parameter binding configuration, and a node tile visual rendering parameter; the node dense area structure set includes a task overlap rate comparison result, a dense node pairing index, and a time section conflict marking label; the visualization structure set for the drug IND phase task layer is specifically an abnormal node identification graph, a tile boundary graph state style, and an abnormal annotation of sample number continuity.

[0013] The improvement of the present invention is that the specific steps for obtaining the sequential number of the pre-clinical efficacy test node in the original plan during the drug IND phase research task process and comparing the numbers according to the planned and execution sequences to generate a task node sequence offset set are as follows:

[0014] S101: Obtain the node list of the pre-clinical efficacy test node in the drug IND phase research task process, extract the sequential number of the pre-clinical efficacy test node in the original task breakdown table, and record the corresponding relationship between the sequential number and the node unique identifier to generate a node sequential number mapping result;

[0015] S102: Call the unique node identifier of the preclinical efficacy test node in the node sequence number mapping result, extract the registration execution time in the task progress record data, sort all preclinical efficacy test nodes in ascending order of the registration execution time, sequentially mark the sorting number as the current sequence number, and construct the correspondence between the sorting number and the unique node identifier to generate the actual execution sorting sequence of the nodes;

[0016] S103: According to the unique node identifiers in the node sequence number mapping result and the actual execution sorting sequence of the nodes, correspond the planned sequence number and the current sequence number of each preclinical efficacy test node, calculate the number difference value between the two numbers and record the direction attribute, and integrate the number offset value of each node in the sequential arrangement to generate the task node position offset set.

[0017] The improvement of the present invention is as follows.

[0018] The specific steps of judging the offset direction of the preclinical efficacy test nodes in the task node position offset set, identifying the offset continuous node group and screening the delayed nodes, constructing the closed boundary layer, and generating the delay section layer structure set are as follows:

[0019] S201: Based on the number offset values of the preclinical efficacy test nodes recorded in the task node position offset set, extract the offset value and the positive and negative direction marks of each node, identify the node offset direction as advance or delay according to the positive and negative status of the offset value, and correspondingly mark the node number after sorting the recognition results according to the current execution time of the nodes to generate the node offset direction identification sequence;

[0020] S202: Call the execution time data of adjacent nodes in the node offset direction identification sequence, compare the time intervals between node pairs in chronological order, classify the node pairs with time intervals not exceeding the node delay aggregation determination value into the same group, and extract the number offset values corresponding to the continuous nodes in the group to obtain the delay node group offset value sequence;

[0021] S203: According to the number offset values of the nodes in the delay node group offset value sequence, compare with the set delay judgment threshold, screen the continuous node segments with number offset values exceeding the delay judgment threshold, locate the first node and the last node in each continuous node segment to construct the closed boundary range, obtain the boundary coordinates and establish the regional layer border to generate the delay section layer structure set.

[0022] The improvement of the present invention is as follows. The specific steps of obtaining the planned completion date, the current system time, and the remaining execution duration of the preclinical efficacy test nodes in the delay section layer structure set to calculate the task urgency gradient value, matching the gray brightness interval to render the node tiles, and generating the task urgency gray coding layer are as follows:

[0023] S301: Obtain the planned completion date, current time, and remaining execution duration of each pre-clinical pharmacodynamic test node in the delay area segment layer structure set, calculate the task urgency gradient value corresponding to each node, and generate a node urgency gradient numerical sequence;

[0024] S302: Based on each task urgency gradient value in the node urgency gradient numerical sequence, call the task urgency level division interval table, identify the section where each gradient value falls, match the corresponding level section with the preset gray scale brightness range table, extract the gray scale brightness parameter corresponding to the current gradient value, and obtain a task urgency brightness mapping parameter set;

[0025] S303: According to the brightness parameter corresponding to the pre-clinical pharmacodynamic test node in the task urgency brightness mapping parameter set, match the node graphic area index table, apply the gray scale brightness parameter to the brightness rendering channel of the node graphic area, establish the graphic brightness rendering configuration of each node, and generate a task urgency gray scale encoding layer.

[0026] The improvement of the present invention is that the specific steps for obtaining the start and end times of the pharmacokinetic detection nodes corresponding to the pre-clinical pharmacodynamic test nodes in the task urgency gray scale encoding layer, counting the node execution time overlap rate, and judging dense node pairs to obtain the node dense area structure set are as follows:

[0027] S401: Obtain the pharmacokinetic detection nodes corresponding to each pre-clinical pharmacodynamic test node in the task urgency gray scale encoding layer, extract the start time and end time of the two nodes, calculate the task time period overlap ratio of each node pair, and generate a node pair overlap rate numerical set;

[0028] S402: According to the task time period overlap ratio value of the node pairs in the node pair overlap rate numerical set, compare the overlap ratio of each node pair numerically according to the task density determination threshold, mark all node pairs with an overlap ratio greater than the task density determination threshold, and extract the node pair index number and record it as a task dense structure unit to generate a node dense area structure set.

[0029] The improvement of the present invention is that the specific steps for obtaining the sample number sequence bound to each node pair in the node dense area structure set, comparing the continuity of the sample numbers to identify abnormal node pairs and marking the graphic abnormal state, and generating a visualization structure set of the drug IND phase task layer are as follows:

[0030] S501: Obtain the sample number sequence bound to each node pair in the node dense area structure set, respectively extract the sample number lists of the pre-clinical pharmacodynamic test nodes and the pharmacokinetic detection nodes, and generate a node pair sample number sequence set;

[0031] S502: Calculate the consecutive number differences and identify their positions according to the arrangement order of each group of sample number data in the sample number sequence set of the nodes, detect the positions of duplicate numbers and reverse order numbers in the number list, determine whether there are cases of missing numbers, duplicates or out-of-order positions, and obtain the abnormal distribution positions of number continuity;

[0032] S503: Extract the corresponding node pair indexes and locate the graphic rendering areas based on the marked missing number, duplicate or misaligned situations in the abnormal distribution positions of number continuity, superimpose status abnormal symbols on each abnormal node pair in the graphic structure of the drug IND phase task, update the border identification information of the abnormal tiles, and generate a visualization structure set of the drug IND phase task layer.

[0033] A visualization management system for the research progress in the drug IND phase, the system comprising:

[0034] The sequence deviation recognition module obtains the sequence numbers of the preclinical pharmacodynamic test nodes in the original plan in the research task process of the drug IND phase, compares the numbers according to the planned and executed sequences, and generates a set of task node sequence deviations;

[0035] The delay layer construction module determines the deviation directions of the preclinical pharmacodynamic test nodes in the set of task node sequence deviations, identifies the continuous deviation node groups and filters the delayed nodes, constructs a closed boundary layer, and generates a set of delay section layer structures;

[0036] The task urgency mapping module obtains the planned completion date, the current system time and the remaining execution duration of the preclinical pharmacodynamic test nodes in the set of delay section layer structures to calculate the task urgency gradient value, matches the gray brightness interval to render the node tiles, and generates a task urgency gray coding layer;

[0037] The task density recognition module obtains the start and end times of the pharmacokinetic detection nodes to which the preclinical pharmacodynamic test nodes in the task urgency gray coding layer belong, counts the time overlap rate to judge the dense node pairs, and obtains a set of node dense area structures;

[0038] The abnormal node annotation module obtains the sample number sequences bound to each node pair in the set of node dense area structures, compares the sample number continuity to identify the abnormal node pairs, and generates a visualization structure set of the drug IND phase task layer.

[0039] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0040] In the present invention, by comparing the sequential numbers of task nodes at the planning and execution levels, extracting the node offset directions, and constructing a closed layer based on offset aggregation, it is possible to group and boundary-express the time-series abnormal nodes, forming a structured area instead of isolated abnormal points; introducing the calculation of the urgency gradient values of the planned time and the remaining cycle for the nodes within these areas, using continuous numerical values as the mapping basis, and establishing gray-scale rendering through parameter correspondence with the brightness interval, so that the time pressure can be visually expressed intuitively, replacing the single expression method of traditional node color marking; the introduction of the time overlap ratio between nodes quantifies the parallel conflict relationship, and the overlapping data drives the recognition of the dense structure, promoting the evolution of the relationship between nodes from single-point tracking to automatic recognition of the interaction density; the continuous analysis of the sample numbers not only expands the sampling dimension of the task execution results, but also maps the number anomalies to the layer area, forming an identification feedback in the visual layer, constituting a continuous processing flow starting from the execution offset, with the time pressure mapping as the central axis, and the task conflict and sample anomalies as the ends, forming a structure anomaly recognition system driven by layer data. Through the introduction of multi-dimensional correlation parameters and the structured analysis of the sample coding mechanism, the abstract abnormal states in the research management process are time-sequentially and graphically expressed, avoiding information omission and node escape. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 is the flowchart of the method of the present invention;

[0042] Figure 2 is the schematic diagram of the detailed process of step S1 of the present invention;

[0043] Figure 3 is the schematic diagram of the detailed process of step S2 of the present invention;

[0044] Figure 4 is the schematic diagram of the detailed process of step S3 of the present invention;

[0045] Figure 5 is the schematic diagram of the detailed process of step S4 of the present invention;

[0046] Figure 6 is the schematic diagram of the detailed process of step S5 of the present invention;

[0047] Figure 7 is the system module diagram of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0048] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0049] In the description of the present invention, it should be understood that the orientation or positional relationship indicated by the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as a limitation to the present invention. In addition, in the description of the present invention, the meaning of "a plurality of" is two or more, unless otherwise specifically defined.

[0050] Please refer to Figure 1 , the present invention provides a technical solution: a visualization management method for the research progress in the IND stage of a drug, including the following steps:

[0051] Divide the research task process in the IND stage of the drug into multiple nodes, obtain the sequence number of the preclinical efficacy test node in the original plan in the research task process in the IND stage of the drug, call the registration execution time data of the preclinical efficacy test node, sort all the preclinical efficacy test nodes in the order of execution time and mark the actual sequence number, then correspond the planned sequence number of the preclinical efficacy test node with the actual sequence number obtained by sorting in terms of the number position, calculate and record the number offset value of the preclinical efficacy test node in the sequential arrangement, and generate a task node sequence offset set;

[0052] Judge whether the offset direction of the preclinical efficacy test node in the task node sequence offset set is advanced or delayed, identify the node group with a continuous time interval not exceeding the node delay aggregation determination value in the judgment result, compare the number offset value in each node group with the preset delay judgment threshold, screen the task node aggregation area exceeding the delay judgment threshold, and mark the first node and the last node of the aggregation area as the closed boundary points, establish the layer closed frame line of the aggregation area, and generate a set of layer structures of the delay section;

[0053] Obtain the planned completion date, the current system time and the remaining execution duration of the preclinical efficacy test node in the set of layer structures of the delay section to calculate the corresponding task urgency gradient value, divide the task urgency into multiple grade sections, assign a set of task tile gray scale brightness ranges to each section, judge the gray scale brightness range corresponding to the current task urgency gradient value, and render the graphic area of the preclinical efficacy test node with the gray scale brightness range as the tile brightness parameter to generate a task urgency gray code layer;

[0054] Obtain the actual start time and actual end time of the pharmacokinetic detection nodes corresponding to the preclinical pharmacodynamic test nodes in the task urgent gray-coded layer, calculate the task overlap rate according to the time period range, compare the task overlap rate with the task density determination threshold, and mark the node pairs of the preclinical pharmacodynamic test nodes and pharmacokinetic detection nodes that exceed the task density determination threshold to obtain the node dense area structure set;

[0055] Obtain the sample number sequence bound to each node pair in the node dense area structure set, check the skip positions, duplicate positions, and order misalignment positions in the number sequence, and determine whether there is any non-consecutive numbering situation in the two-node sample number sequences. If there is discontinuous numbering, embed abnormal graphic symbols for the corresponding node pairs in the drug IND stage task graphic structure respectively to generate the drug IND stage task layer visualization structure set.

[0056] The task node sequence offset set includes the sequence number mapping relationship, the sequence misalignment direction identifier, and the node execution sequence index. The delay section layer structure set includes the delay aggregation section range identifier, the layer closed boundary box line parameters, and the task number offset screening record. The task urgent gray-coded layer includes the task urgent level interval mapping table, the gray scale brightness parameter binding configuration, and the node block visual rendering parameters. The node dense area structure set includes the task overlap rate comparison result, the dense node pairing index, and the time section conflict marking label. The drug IND stage task layer visualization structure set is specifically the abnormal node identification graphic, the graphic state style of the block boundary, and the abnormal annotation of the sample number continuity.

[0057] Please refer to Figure 2 , to obtain the sequence numbers of the preclinical pharmacodynamic test nodes in the original plan in the drug IND stage research task process, and compare the numbers according to the planned and execution sequences. The specific steps to generate the task node sequence offset set are as follows:

[0058] S101: Obtain the node list of the preclinical pharmacodynamic test nodes in the drug IND stage research task process, extract the sequence numbers of the preclinical pharmacodynamic test nodes in the original task breakdown table, and record the corresponding relationship between the sequence numbers and the node unique identifiers to generate the node sequence number mapping result;

[0059] Obtain the node list of the preclinical efficacy test node in the research task process of the drug IND stage, that is, all nodes marked as "preclinical efficacy test" need to be extracted from the work breakdown structure. This process usually screens based on the task structured data in the project management system. First, extract all records classified as "preclinical efficacy test" among all task nodes through the node category field set in the system, and use task management tools such as Project, JIRA or the structure fields in the self-built database for query. For example, if the task table contains the field "node type", the nodes with its value being "efficacy test" are selected; then it is necessary to extract the sequence number of this node in the original task structure. The sequence number is generally generated through the arrangement order in the task tree structure and usually exists in the form of an integer number, such as 001, 002, 003, etc. To ensure the accuracy of the number sequence, the "sorting field" or "priority field" in the task structure can be called. For example, if the node ID is "ND001" and its sequence number in the structure is 3; after that, a mapping relationship between the sequence number and the unique identifier of the node needs to be established. A key-value pair structure is established through programming for mapping. If implemented using Python code, a dictionary structure such as {3: "ND001", 4: "ND002"} is the required mapping result, where the sequence number is the key and the unique identifier of the node is the value. This mapping structure can be used for verification and comparison operations in the subsequent execution process. The above process needs to consider the uniqueness and integrity of the node information in the task breakdown table. If there are multiple nodes with the same classification but different IDs, it should be ensured that the valid nodes in the task version are extracted; for example, in the IND stage of drug A, there are 5 preclinical efficacy test nodes, numbered from "ND010" to "ND014" in sequence, and the corresponding sequence numbers in the task structure are 1 to 5, then the mapping result is {1: "ND010", 2: "ND011", 3: "ND012", 4: "ND013", 5: "ND014"}, and this data constitutes the input basic data for the subsequent steps.

[0060] S102: Call the unique identifier of the preclinical efficacy test node in the mapping result of the node sequence number, extract the registration execution time in the task progress record data, sort all preclinical efficacy test nodes in ascending order of the registration execution time, sequentially mark the sorting number as the current sequence number, and construct the corresponding relationship between the sorting number and the unique identifier of the node to generate the actual execution sorting sequence of the node;

[0061] Call the node unique identifier of the preclinical efficacy test node in the node sequence number mapping result, that is, extract all preclinical efficacy test node IDs from the mapping structure obtained in the previous paragraph. For example, the "ND010" to "ND014" in {1: "ND010", 2: "ND011", 3: "ND012", 4: "ND013", 5: "ND014"}. Search for the task progress registration records corresponding to these IDs in the project task database, and specifically extract the content of the "registration execution time" field. This field is in timestamp format, such as "2024-03-01 10:00:00". Then sort all nodes in ascending order according to the registration execution time. For example, if the registration time of "ND012" is "2024-03-01", "ND010" is "2024-03-05", and "ND011" is "2024-03-07", then the sorted order is "ND012", "ND010", "ND011", etc. The system can use sorting algorithms such as quicksort, bubble sort, or the ORDER BY statement in the database to achieve sorting. After sorting, assign a current sequence number to each node, that is, increment from 1 according to its position in the sorting. For example, if the sorted list is ["ND012", "ND010", "ND011", "ND013", "ND014"], then the current sequence number is {1: "ND012", 2: "ND010", 3: "ND011", 4: "ND013", 5: "ND014"}. This structure is the actual execution sorting sequence of the nodes. For example, the original planned sequence of the preclinical efficacy test tasks of drug B was from "ND020" to "ND024", but the actual execution time records show that the sequence is "ND022", "ND020", "ND021", "ND023", "ND024". Therefore, the mapping of the sorting number to the node unique identifier is {1: "ND022", 2: "ND020", 3: "ND021", 4: "ND023", 5: "ND024"}, and this structure is used for subsequent offset calculations. Nodes with missing registration time should be excluded or processed using the most recently updated time in the sorting logic.

[0062] S103: According to the node unique identifier in the node sequence number mapping result and the actual execution sorting sequence of the nodes, correspond the planned sequence number and the current sequence number of each preclinical efficacy test node, calculate the number difference value between the two numbers and record the direction attribute, and integrate the number offset value of each node in the sequential arrangement to generate a task node position offset set;

[0063] According to the node unique identifier in the node sequence number mapping result and the actual execution sorting sequence of the nodes, correspond the planned sequence number and the actual execution number of each node, compare the number differences one by one and record their offset directions. The difference is defined as Δ i= i actual -i planned , where Δ i represents the sequential offset value of the i-th node, and i actual represents the sequential number of the node in the actual execution of the sorting, and i planned represents the planned sequential number of the node in the original task decomposition. If Δ i > 0, it means that the node sequence lags behind, and the direction flag is "+"; if Δ i < 0, it means that the node sequence is advanced, and the direction flag is "-"; if Δ i = 0, it means that the node sequence has not changed, and the direction flag is "0"; this process is achieved by traversing the node list and calculating one by one. For example, the planned sequence of "ND010" is 2 and the actual execution is 3, then Δ = 3 - 2 = 1, and the direction is "+", and the sequential offset data structure of each node can be constructed, such as {"ND010": {"planned": 2, "actual": 3, "delta": 1, "direction": "+"}}; example: In the IND stage of drug C, the planned sequence and actual sorting of its test nodes are as follows - planned: ["ND030", "ND031", "ND032", "ND033", "ND034"], actual: ["ND031", "ND030", "ND032", "ND034", "ND033"], then the planned sequence of the node "ND031" is 2, the actual sequence is 1, Δ = 1 - 2 = -1, and the direction is "-"; based on this, a complete set of sequential offsets of task nodes is constructed, such as {"ND030": {" ": 1, "direction": "+"}, "ND031": {" ": -1, "direction": "-"}...}; in the actual execution process, the absolute value |Δ| of the offset value Δ can be used to judge the sorting stability. If |Δ| > 2, it is regarded as a high-offset node, and the setting logic of this "high-offset" threshold is where T is the offset threshold, N is the total number of task nodes, and the symbol represents rounding up. If N = 9, then at this time, |Δ| ≥ 3 is judged as a significantly offset node. This judgment criterion can dynamically adapt to research tasks with different numbers of nodes, thereby forming a stable basis for the analysis of sequential offsets of node executions.

[0064] Please refer to Figure 3 , the specific steps to judge the offset direction of preclinical pharmacodynamic test nodes in the set of sequential offsets of task nodes, identify continuous offset node groups and screen for delayed nodes, construct a closed boundary layer, and generate a set of delay zone layer structures are as follows:

[0065] S201: Based on the number offset values of preclinical pharmacodynamic test nodes recorded in the task node sequence offset set, extract the offset numerical values and positive / negative direction markers of each node. Identify whether the node offset direction is advanced or delayed according to the positive / negative status of the offset value. After sorting the recognition results according to the current execution time of the nodes, label the node numbers correspondingly to generate a node offset direction identification sequence;

[0066] First, read the sequence offset value data of all preclinical pharmacodynamic test nodes from the established data structure. For each node, the system needs to traverse its key-value pair structure and read the "delta" (number offset value) and "direction" (offset direction) fields. For example, if the offset value of node "ND101" is -2, its direction is advanced, and the offset direction is "-". Another example is that the offset value of "ND102" is 3, then the offset direction is "+", indicating delay. The identification of the offset direction is completed by judging the positive / negative status of the offset value, that is, if the value is negative, the offset direction is advanced, if it is positive, it is delayed, and if it is 0, there is no offset. This judgment can be implemented through conditional statements. Subsequently, call the actual execution time field of each node, which is a specific timestamp, such as "2024-05-01 09:00:00". Sort all nodes in ascending order according to the time sequence. The sorting algorithm can call the built-in sorted function in Python or the database ORDER BY statement. The sorting field is time. After sorting, re-label the node numbers and their offset directions at the corresponding sorting positions. For example, the original structure is {"ND101": {"delta": -2, "direction": "-", "time": "2024-04-02"}, "ND102": {"delta": 3, "direction": "+", "time": "2024-04-03"}}, and the sorted structure becomes: ["ND101 - advanced", "ND102 - delayed"], where the offset direction fields "advanced" and "delayed" are obtained by converting the direction markers "-" and "+", respectively. This labeled result generates a node offset direction identification sequence. For example, the preclinical nodes of drug D are "ND201" (2024-03-01), "ND202" (2024-03-03), "ND203" (2024-03-06) according to the execution time, and their offset values are -1, 0, 2 respectively, corresponding to the identifications "advanced", "no offset", "delayed", then the node offset direction identification sequence is ["ND201 - advanced", "ND202 - no offset", "ND203 - delayed"].

[0067] S202: Call the execution time data of adjacent nodes in the node offset direction identification sequence, compare the time intervals between node pairs in chronological order, group the node pairs with time intervals not exceeding the node delay aggregation determination value into the same group, and extract the number offset values corresponding to consecutive nodes in the group to obtain the delay node group offset numerical sequence;

[0068] Call the execution time data of adjacent nodes in the node offset direction identification sequence. It is necessary to extract the interval between each pair of adjacent nodes in the time dimension. Suppose the time of node "ND301" is April 1, 2024, and "ND302" is April 3, 2024, and the interval between them is 2 days. If the delay aggregation determination value is set to 3 days, then this node pair meets the grouping condition; continue to judge "ND303", the execution time is April 4, 2024, and the interval from "ND302" is 1 day, still less than 3 days, and the three can be combined into a delay node group; if the time of "ND304" is April 10, 2024, and the interval from the previous node "ND303" is 6 days, exceeding the determination value, it will not be included in the previous group and will be used as the first node of the next group. The above process can be achieved by traversing the node time sequence in a sliding window manner; after each group is formed, it is necessary to extract the number offset values of each node in it. For example, group 1 is "ND301", "ND302", "ND303", and the offset values are -2, 0, 3 respectively, then the offset numerical sequence is [-2, 0, 3]; group 2 is "ND304", "ND305", and the offset values are 1, 2, and the sequence is [1, 2]; this delay node group offset numerical sequence will be used as the input basis for subsequent judgment of delay segments. The entire grouping operation needs to ensure that the time differences between all adjacent nodes do not exceed the set determination value, and the number of nodes in each group is at least 2, meeting the basic requirement of continuity.

[0069] S203: According to the number offset values of the nodes in the delay node group offset numerical sequence, compare them with the set delay judgment threshold, screen out the consecutive node segments with number offset values exceeding the delay judgment threshold, and locate the first node and the last node in each segment of consecutive nodes to construct a closed boundary range, obtain the boundary coordinates and establish the border of the regional layer, and generate the delay segment layer structure set;

[0070] According to the node number offset values in the delay node group offset value sequence, the system needs to compare each offset value in each group with the set delay judgment threshold. The threshold can be set by dividing the total number of task nodes by 5. For example, if the task contains 20 nodes, the delay judgment threshold is 4, which means that if the absolute value of the number offset value exceeds 4, it is a delayed node. The system checks the size of the offset value for each group sequence in turn and determines whether it meets the delay standard. For example, for group 1 which is [-1, 5, 6], where 5 and 6 are both greater than 4, they are determined to be delayed nodes. Two consecutive delayed nodes "ND402" and "ND403" form a delay segment; another example is group 2 which is [1, 3], the maximum offset is 3, which does not exceed the threshold and does not form a delay segment; the system identifies the start and end nodes of the segment, obtains "ND402" as the start node and "ND403" as the end node, and determines the closed boundary range as ["ND402" - "ND403"]. Then, it obtains the coordinate information of these two nodes in the layer. The horizontal axis is the timestamp and the vertical axis is the sequence number. For example, the time of "ND402" is May 1, 2024, and the number is 7, and "ND403" is May 4, 2024, and the number is 8. Then the upper left corner coordinates of the area are (May 1, 2024, 7), and the lower right corner coordinates are (May 4, 2024, 8), and a closed rectangular border is constructed, and this border is recorded as a layer element; after all delay segments are identified, the boundary set such as: [Boundary 1: ("ND402" - "ND403"), Boundary 2: ("ND407" - "ND409")], all form the border coordinate range with the time and number of the nodes and form a layer structure set, which is used to construct the graphical description of the delay segment.

[0071] Please refer to Figure 4 , to obtain the planned completion date, the current time of the system, and the remaining execution duration of the preclinical pharmacodynamic test nodes in the layer structure set of the delay segment, calculate the task urgency gradient value, match the gray brightness interval to render the node tiles, and the specific steps to generate the task urgency gray code layer are as follows:

[0072] S301: Obtain the planned completion date, the current time, and the remaining execution duration of each preclinical pharmacodynamic test node in the layer structure set of the delay segment, calculate the task urgency gradient value corresponding to each node, and generate a node urgency gradient numerical sequence;

[0073] After obtaining the planned completion date, current time, and remaining execution duration of each preclinical pharmacodynamic test node in the delay area layer structure set, the units of the three need to be unified to the same dimension to meet the logical relationship in the formula. The project system will automatically traverse all node IDs from the node layer structure index, call the node attribute table fields one by one, and extract the fields including: planned completion date (such as 2025-05-10), current system time (such as 2025-03-30), remaining execution duration (such as 36 hours), and convert the dates to "relative days" uniformly with reference to the starting baseline date set by the project (for example, 2025-01-01, set as day 0). The system converts the planned completion time (denoted as Date plan ) and the current time (denoted as Date now ) into the D plan -th and D now -th days respectively from the starting date. For example, if the planned completion time is May 10, 2025, it is converted to the 130th day, and the current time of March 30, 2025, is converted to the 89th day. The remaining execution duration is in hours (such as 36 hours) and needs to be uniformly converted to the unit of "days", that is is the original remaining execution duration, and h is hours. Using the formula:

[0074]

[0075] calculate the task urgency gradient value U;

[0076] where D plan is the "relative days" corresponding to the planned completion date, starting from the project start date as Day0 (unit: days), D now is the relative days corresponding to the current system time (unit: days), is the remaining execution duration, is the ideal execution cycle, d is in days, obtained from historical data statistics, δ is a very small constant to prevent the denominator in the square root term from approaching 0 (the value is such as 0.01), ε is a stable correction term to prevent oscillation when dividing by zero or in the extreme value of the inverse trigonometric function (the value is such as 0.01), α is an error term regulation coefficient, adjusting the sensitivity of the planned time difference and the remaining cycle offset (unitless, usually taking 1.0), arcsin() is the arcsine function, calculating the radian value of the angle corresponding to a given proportional offset, used to compress the change rate of the deviation term.

[0077] Assume the parameters of node A are as follows: D plan = 130: The planned completion is on May 10, D now = 89: The current time is on March 30, After conversion The ideal completion period of the node under standard resource configuration, δ=0.01, ε=0.01: positive constant to prevent division by zero, α=1.0: adjustment coefficient to control the strength of the error term.

[0078] Denominator square root calculation:

[0079] Inverse sine term calculation:

[0080] Gradient value calculation:

[0081] This result indicates that the current node A task urgency gradient is 4.74. By performing the same process on all nodes, a task urgency gradient numerical sequence such as {4.74, 3.01, 6.82, 1.45, …} is formed for subsequent mapping and scheduling module calls.

[0082] S302: Based on each task urgency gradient value in the node urgency gradient value sequence, call the task urgency level division interval table, identify the segment where each gradient value falls, match the corresponding level segment with the preset grayscale brightness range table, extract the grayscale brightness parameter corresponding to the current gradient value, and obtain the task urgency brightness mapping parameter set;

[0083] Each task urgency gradient value in the node urgency gradient value sequence needs to call the task urgency level division interval table to determine its numerical segment. The system divides the interval according to the preset, such as: [0, 1) = "very low", [1, 2) = "low", [2, 4) = "medium", [4, 6) = "high", [6, 8) = "very high", [8, +∞) = "extremely high", and classifies each node urgency value into the corresponding segment and then searches for the grayscale brightness parameter range mapped to the segment. For example, the grayscale brightness value corresponding to the "high" segment is [160, 192]. The system performs linear interpolation on the current node gradient value within the level segment to which it belongs and converts it into a grayscale brightness value. During the interpolation process, the linear interpolation formula used is as follows:

[0084]

[0085] Among them, L is the final grayscale brightness value of the target node, L min , L max are the lower and upper limits of the brightness interval corresponding to the gradient level, U is the urgent gradient value of the current node, and U min , U max It is the upper and lower limits of the level range where the current gradient value is located.

[0086] Taking this node as an example, its gradient value is 4.74, which falls into the range of [4, 6), that is, the level "high", and the corresponding brightness interval is [160, 192]. After substituting into the interpolation calculation, the grayscale brightness value is 171.84. The system performs the same operation on all nodes, and finally generates a task urgency brightness mapping parameter set, such as {171.84, 148.12, 200.00,...}, which provides input parameters for the brightness rendering of graphic nodes.

[0087] S303: According to the brightness parameter corresponding to the preclinical pharmacodynamic test node in the task urgency brightness mapping parameter set, match the node graphic area index table, apply the grayscale brightness parameter to the brightness rendering channel of the node graphic area, establish the graphic brightness rendering configuration for each node, and generate a task urgency grayscale coding layer;

[0088] According to the brightness value corresponding to the preclinical pharmacodynamic test node in the task urgency brightness mapping parameter set, the system loads the layer area index table in the graphic module. This index table contains the pixel area coordinate range corresponding to each node in the rendering layer. For example, the graphic area of node A is a rectangular area with the horizontal axis from 120 to 140 and the vertical axis from 200 to 220. The system matches the node number with this graphic area index and writes the grayscale brightness value of the node into the grayscale rendering channel of the corresponding area. The grayscale brightness value of each node is applied to all pixel points within its corresponding area as a whole. For example, the brightness value corresponding to node A is 171.84, and the system uniformly sets all pixel values in its pixel area to 171 to ensure that a clear visual intensity is formed for this node in the layer. The system performs the above brightness value injection operation for each node one by one until all test nodes are traversed to complete the graphic brightness rendering configuration. All rendering actions are executed in the pixel rendering engine, and the rendered brightness encoded data is synthesized into the final task urgency grayscale coding layer through the layer update mechanism, which serves as one of the input sources for the layer visualization module.

[0089] Please refer to Figure 5 , to obtain the start and end times of the pharmacokinetic detection nodes corresponding to the preclinical pharmacodynamic test nodes in the task urgency grayscale coding layer, calculate the node execution time overlap rate and judge the dense node pairs. The specific steps to obtain the node dense area structure set are as follows:

[0090] S401: Obtain the pharmacokinetic detection nodes corresponding to each preclinical pharmacodynamic test node in the task urgency grayscale coding layer, extract the start time and end time of the two nodes, calculate the task time period overlap ratio of each node pair, and generate a set of node pair overlap rate values;

[0091] To obtain the pharmacokinetic detection nodes corresponding to each preclinical pharmacodynamic test node in the task-urgent gray-coded layer, it is necessary to first establish an index table of the attribution relationship between the two types of nodes. This table is constructed through the hierarchical dependency field in the project process definition. Each pharmacodynamic test node establishes a logical mapping relationship with a pharmacokinetic detection node through the field "attributed PK node ID". For example, if the pharmacodynamic node E104 belongs to the pharmacokinetic node PK21, then establish the node pair relationship of E104→PK21. The system reads the list of all rendered pharmacodynamic test nodes from the task-urgent gray-coded layer. For each node, it calls the layer index interface and the task schedule to extract its task time period, that is, the "start time" and "end time" fields. For example, for a node E104, the start time is 2025-03-28 and the end time is 2025-04-03, which is converted to the 87th to 93rd day of the project; the corresponding pharmacokinetic detection node PK21 has a start time of 2025-04-01 and an end time of 2025-04-07, which is converted to the 91st to 97th day of the project. Then the system calculates the interval overlap of the two nodes' task time periods. The method is to divide the number of intersection days of the two intervals by the number of union days. That is, the overlap ratio calculation formula is:

[0092]

[0093] Among them, represents the start and end time of the pharmacodynamic test node, is the time period of the start and end time belonging to the pharmacokinetic node. +1 means that both the start and end days are included. In actual execution, for example, the interval of E104 is [87, 93], the interval of PK21 is [91, 97], the intersection is [91, 93] for a total of 3 days, and the union is [87, 97] for a total of 11 days. Substituting into the calculation gives: The system records this ratio as the overlap ratio of the node pair (E104, PK21). Calculate the time period overlap ratios of all node pairs in turn to generate a set of node pair overlap rate values such as {(E104, PK21, 0.273), (E105, PK22, 0.615), (E106, PK22, 0),...}. Each item in it includes the pharmacodynamic node number, the number of its attributed pharmacokinetic node, and the value of its task time period overlap ratio.

[0094] S402: According to the task time period overlap ratio values of the node pairs in the set of node pair overlap rate values, and according to the task density determination threshold, compare the overlap ratios of each node pair numerically, mark all node pairs with overlap ratios greater than the task density determination threshold, and extract the node pair index numbers and record them as task density structural units to generate a set of node dense area structures;

[0095] According to the task time period overlap ratio values of node pairs in the node pair overlap rate numerical set, numerical comparison and conditional screening need to be combined with the task density determination threshold. This threshold is set to determine whether there is an actual execution overlap of tasks. Generally, it is set to 0.6, that is, only when 60% of the time periods overlap is it considered a task-intensive structure. This value can be set by setting the mean and standard deviation of the average overlap ratio of past projects. The reasonable range is between [0.5, 0.7]. If the set value is taken as 0.6, the system reads each node pair in the overlap rate set and compares its ratio value with the threshold one by one. The comparison method is direct numerical judgment. When the overlap ratio is greater than 0.6, the node pair is marked as a task-intensive node pair. For example, the overlap ratio of the node pair (E105, PK22) is 0.615, which is greater than 0.6, so it meets the condition and is marked as a task-intensive node pair. For all node pairs that meet the conditions, their node pair index numbers are extracted. For example, "E105→PK22", and the system records it in the task-intensive structure unit set to construct a dense area structure set, such as {(E105, PK22), (E110, PK24), (E111, PK24)}. Finally, a node dense area structure set is generated as an input parameter for the subsequent time scheduling and layer conflict detection module.

[0096] Please refer to Figure 6 , to obtain the sample number sequence bound to each node pair in the node dense area structure set, compare the continuity of the sample numbers to identify abnormal node pairs and mark the graphic abnormal state. The specific steps to generate the visualization structure set of the drug IND phase task layer are as follows:

[0097] S501: Obtain the sample number sequence bound to each node pair in the node dense area structure set, respectively extract the sample number lists of the preclinical pharmacodynamic test nodes and the pharmacokinetic detection nodes, and generate a node pair sample number sequence set;

[0098] In the process of obtaining the sequence of sample numbers corresponding to each node pair in the concentrated structure of node-dense regions, the system first needs to read the node pair indexes one by one from the generated set of node-dense region structures. Each element in this structure set contains a preclinical pharmacodynamic test node number and the pharmacokinetic detection node number to which it belongs. Subsequently, the system accesses the preclinical experiment sample information database and the drug metabolism detection sample registration form, and retrieves the sample information records bound to the two node ID fields respectively. The data fields usually include sample number, sample ID, sampling time, the node ID to which it belongs, etc. The system creates a sample number list for each node. For example, the preclinical node E301 is associated with sample numbers [S1001, S1002, S1003], and the corresponding pharmacokinetic node PK72 is associated with sample numbers [S1003, S1004, S1005]. The system records the two lists respectively and binds them to the node pair to form a structure item. The acquisition process supports a many-to-many relationship, and some samples may appear repeatedly in the two nodes. In this process, the system needs to use a filtering function to exclude non-target node samples and ensure the accuracy of each pair of sample number lists through intersection verification logic. Finally, the sample number data structure is organized in the following way: {(E301, PK72): [[S1001, S1002, S1003], [S1003, S1004, S1005]]}. After extracting the sample numbers for each node pair in this way, the system stores all the node pairs and their corresponding sample number combinations in the sample number sequence set.

[0099] S502: According to the arrangement order of each group of sample number data in the sample number sequence set of node pairs, calculate the difference between consecutive numbers and identify their positions, and detect the positions of duplicate numbers and reverse-order numbers in the number list, and determine whether there are cases of missing numbers, duplicates, or sequential misalignment, so as to obtain the abnormal distribution positions of number continuity;

[0100] According to the arrangement order of each group of sample number data in the sample number sequence set by the node, calculate the difference between consecutive numbers and identify their positions, and detect the positions of duplicate numbers and reverse-order numbers in the number list to determine whether there are cases of missing numbers, duplicates, or out-of-order sequences. The system needs to first parse the number values of each group of sample number lists, that is, extract the pure numerical part in the sample number and convert it into a comparable numerical list. For example, [S1001, S1002, S1005] is converted to [1001, 1002, 1005]. Then the system performs continuity detection on the sample numbers, uses the difference analysis method to process the number intervals, and calculates the difference between adjacent numbers. If the difference is not 1, it is marked as a missing number. For example, in the numbers [1001, 1002, 1005], the difference between 1002 and 1005 is 3, indicating that there are two missing numbers, and the missing number positions are recorded as 1002→1005. Duplicate number detection is done by comparing the length of the original list with the length of the set after removing duplicates. For example, after removing duplicates from [1001, 1002, 1002, 1003], it becomes [1001, 1002, 1003]. If the length decreases, it is determined that there are duplicates, and the duplicate positions are recorded as the 2nd to the 3rd. Reverse-order detection uses the current number list and its ascending-order result for bit-by-bit comparison. For example, if the current is [1003, 1001, 1002] and the ascending order is [1001, 1002, 1003], the system compares the index differences and determines that there is a reverse order at the 1st and 2nd positions. The system records the abnormal number continuity situations in a structured form, including missing number marks, duplicate position indexes, and reverse-order position marks, and at the same time gives abnormal type labels. The system uses the following classification criteria: if the difference > 1, it is a missing number; if the difference = 0, it is a duplicate; if the current number value is less than the previous number, it is a reverse order. The system performs a complete inspection process on the sample number list according to the above criteria and maps all abnormal numbers and their positions back to the node pair indexes to form a set of records of the abnormal distribution positions of number continuity.

[0101] S503: According to the marked missing numbers, duplicates, or misalignments in the abnormal distribution positions of number continuity, extract the corresponding node pair indexes and locate the graphic rendering areas, and superimpose status abnormal symbols on each abnormal node pair in the graphic structure of the drug IND stage task respectively, update the border identification information of the abnormal tiles, and generate a set of visual structures of the drug IND stage task layer.

[0102] Based on the skipped numbers, repetitions, or misalignments marked in the abnormal distribution positions of the numbering continuity, the system extracts the corresponding node pair indexes and accesses the layer structure. First, it matches the node pair with the graphic index table to locate the graphic area of the node pair in the task graphic structure. The graphic structure takes the node pair as the basic unit, and each node pair occupies a specific rectangular area in the graphic rendering structure. The system adds identification information to the graphic areas of all abnormal node pairs. The abnormal types are overlaid by the symbol layer. For example, skipped numbers are identified with the "!" icon, repetitions are identified with "△", and sequential misalignments are identified with the icon, and the identification is located in the upper right corner of the tile area, and the border color is also adjusted synchronously. The border color of the skipped number is set to red (#FF0000), the repeated one is orange (#FFA500), and the misaligned one is purple (#800080). The rendering process of the graphic area is handled by the abnormal identification rendering module. The system combines the abnormal symbol primitives into texture materials according to the abnormal types and performs a local layer update operation on the target area. After the update, the layer content contains the visible identification of the abnormal state, and finally forms a visualization structure set of the drug IND stage task layers, including the state rendering content and layer coordinate information of all abnormal node pairs.

[0103] Please refer to Figure 7 , a visualization management system for the research progress in the drug IND stage, the system includes:

[0104] The sequence offset recognition module obtains the sequence numbers of the preclinical pharmacodynamic test nodes in the original plan in the research task process of the drug IND stage, compares the numbers according to the planned and executed sequences, and generates a set of task node sequence offsets;

[0105] The delay layer construction module judges the offset direction of the preclinical pharmacodynamic test nodes in the set of task node sequence offsets, identifies the continuous offset node groups and filters the delayed nodes, constructs a closed boundary layer, and generates a set of delay section layer structures;

[0106] The task urgency mapping module obtains the planned completion date, the current system time, and the remaining execution duration of the preclinical pharmacodynamic test nodes in the set of delay section layer structures to calculate the task urgency gradient value, matches the gray scale brightness interval to render the node tiles, and generates a task urgency gray scale coding layer;

[0107] The task density recognition module obtains the start and end times of the pharmacokinetic detection nodes corresponding to the preclinical pharmacodynamic test nodes in the task urgency gray scale coding layer, statistically calculates the time overlap rate to judge the dense node pairs, and obtains a set of node dense area structures;

[0108] The abnormal node annotation module obtains the sample number sequences bound to each node pair in the set of node dense area structures, compares the continuity of the sample numbers to identify the abnormal node pairs, and generates a visualization structure set of the drug IND stage task layers.

[0109] The above are only the preferred embodiments of the present invention, and are not intended to limit the present invention in other forms. Any person skilled in the art may use the technical content disclosed above to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, as long as it does not depart from the technical solution content of the present invention, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. A visual management method for the research progress in the IND stage of a drug, characterized in that, It includes the following steps: S1: Obtain the sequence number of the preclinical pharmacodynamic test node in the original plan of the research task process in the IND stage of the drug, compare the numbers according to the planned and execution order, and generate a task node sequence offset set; S2: Judge the offset direction of the preclinical pharmacodynamic test node in the task node sequence offset set, identify the offset continuous node group and screen the delayed nodes, construct a closed boundary layer, and generate a delay section layer structure set; S3: Obtain the planned completion date, the current system time and the remaining execution duration of the preclinical pharmacodynamic test node in the delay section layer structure set to calculate the task urgency gradient value, match the gray brightness interval to render the node tile, and generate a task urgency gray coding layer; S4: Obtain the start and end times of the pharmacokinetic detection node corresponding to the preclinical pharmacodynamic test node in the task urgency gray coding layer, count the time overlap rate to judge the dense node pair, and obtain a node dense area structure set; S5: Obtain the sample number sequence bound to each node pair in the node dense area structure set, compare the continuity of the sample numbers to identify the abnormal node pair, and generate a visualization structure set of the drug IND stage task layer.

2. The visualization management method for the research progress of the drug in the IND stage according to claim 1, wherein: The task node sequence offset set includes the sequence number mapping relationship, the sequence misalignment direction identifier, and the node execution sequence index. The delay section layer structure set includes the delay aggregation section range identifier, the layer closed boundary box line parameters, and the task number offset screening record. The task urgency gray coding layer includes the task urgency level interval mapping table, the gray brightness parameter binding configuration, and the node tile visual rendering parameters. The node dense area structure set includes the task overlap rate comparison result, the dense node pairing index, and the time section conflict marker label. The visualization structure set of the drug IND stage task layer is specifically the abnormal node identifier graph, the tile boundary graph state style, and the sample number continuity abnormality annotation.

3. The visualization management method for the research progress of the drug in the IND stage according to claim 1, wherein: The specific steps to obtain the sequence number of the preclinical pharmacodynamic test node in the original plan of the research task process in the IND stage of the drug, compare the numbers according to the planned and execution order, and generate a task node sequence offset set are as follows: S101: Obtain the node list of the preclinical pharmacodynamic test node in the research task process in the IND stage of the drug, extract the sequence number of the preclinical pharmacodynamic test node in the original task breakdown table, and record the corresponding relationship between the sequence number and the node unique identifier to generate a node sequence number mapping result; S102: Call the node unique identifier of the preclinical pharmacodynamic test node in the node sequence number mapping result, extract the registered execution time in the task progress record data, sort all the preclinical pharmacodynamic test nodes from the earliest to the latest according to the registered execution time, sequentially mark the sorting number as the current sequence number, and construct the corresponding relationship between the sorting number and the node unique identifier to generate a node actual execution sorting sequence; S103: According to the node sequence number mapping result and the unique identifier of the nodes in the actual execution sorting sequence of the nodes, the planned sequence number of each pre-clinical pharmacodynamic test node is corresponding to the current sequence number in terms of position, calculate the number difference value between the two numbers and record the direction attribute, and integrate the number offset values of each node in the sequential arrangement to generate a task node sequence offset set.

4. The visualization management method for the research progress of the drug in the IND stage according to claim 3, wherein: The specific steps for judging the offset direction of the pre-clinical pharmacodynamic test nodes in the task node sequence offset set, identifying the offset continuous node groups and screening the delayed nodes, and constructing a closed boundary layer to generate a set of delay section layer structures are as follows: S201: Based on the number offset values of the pre-clinical pharmacodynamic test nodes recorded in the task node sequence offset set, extract the offset numerical value and positive / negative direction mark of each node, identify whether the node offset direction is advanced or delayed according to the positive / negative state of the offset value, and after sorting the recognition results according to the current execution time of the nodes, mark the node numbers correspondingly to generate a node offset direction identification sequence; S202: Call the execution time data of adjacent nodes in the node offset direction identification sequence, compare the time intervals between node pairs in chronological order, group the node pairs with time intervals not exceeding the node delay aggregation determination value into the same group, and extract the number offset values corresponding to the continuous nodes in the group to obtain a sequence of offset numerical values of the delayed node group; S203: According to the number offset values of the nodes in the sequence of offset numerical values of the delayed node group, compare the numerical values with the set delay judgment threshold, screen the continuous node segments with number offset values exceeding the delay judgment threshold, and locate the first node and the last node in each continuous node segment to construct a closed boundary range, obtain the boundary coordinates and establish a regional layer border to generate a set of delay section layer structures.

5. The visualization management method for the research progress of the drug in the IND stage according to claim 4, wherein: The specific steps for obtaining the planned completion date, the current system time and the remaining execution duration of the pre-clinical pharmacodynamic test nodes in the set of delay section layer structures to calculate the task urgency gradient value, matching the gray scale brightness interval to render the node tiles, and generating a task urgency gray code layer are as follows: S301: Obtain the planned completion date, the current time and the remaining execution duration of each pre-clinical pharmacodynamic test node in the set of delay section layer structures, calculate the task urgency gradient value corresponding to each node, and generate a sequence of node urgency gradient numerical values; S302: Based on each task urgency gradient value in the sequence of node urgency gradient numerical values, call the task urgency level division interval table, identify the section where each gradient value falls, match the corresponding level section with the preset gray scale brightness range table, and extract the gray scale brightness parameter corresponding to the current gradient value to obtain a set of task urgency brightness mapping parameters; S303: According to the brightness parameters corresponding to the pre-clinical pharmacodynamic test nodes in the set of task urgency brightness mapping parameters, match the node graphic area index table, apply the gray scale brightness parameter to the brightness rendering channel of the node graphic area, establish the graphic brightness rendering configuration of each node, and generate a task urgency gray code layer.

6. The visualization management method for the research progress of the drug in the IND stage according to claim 5, wherein: Obtain the start and end times of the pharmacokinetic detection nodes corresponding to the preclinical pharmacodynamic test nodes in the task urgency gray coding layer, count the node execution time overlap rate and judge the dense node pairs, and the specific steps to obtain the node dense area structure set are as follows: S401: Obtain the pharmacokinetic detection nodes corresponding to each preclinical pharmacodynamic test node in the task urgency gray coding layer, extract the start time and end time of the two nodes, calculate the task time period overlap ratio of each node pair, and generate a node pair overlap rate value set; S402: According to the task time period overlap ratio value of the node pairs in the node pair overlap rate value set, and according to the task density determination threshold, compare the overlap ratios of each node pair numerically, mark all node pairs with an overlap ratio greater than the task density determination threshold, and extract the node pair index numbers and record them as task dense structure units to generate a node dense area structure set.

7. The visual management method for the research progress of the drug in the IND stage according to claim 6, wherein: Obtain the sample number sequence bound to each node pair in the node dense area structure set, compare the continuity of the sample numbers to identify abnormal node pairs and mark the graphic abnormal status, and the specific steps to generate the visualization structure set of the drug IND stage task layer are as follows: S501: Obtain the sample number sequence bound to each node pair in the node dense area structure set, respectively extract the sample number lists of the preclinical pharmacodynamic test nodes and the pharmacokinetic detection nodes, and generate a node pair sample number sequence set; S502: According to the arrangement order of each group of sample number data in the node pair sample number sequence set, calculate and position the continuous number difference, and detect the positions of duplicate numbers and reverse order numbers in the number list, and judge whether there are cases of missing numbers, duplicates or order misalignments, and obtain the abnormal distribution positions of number continuity; S503: According to the missing number, duplicate or misalignment situations marked in the abnormal distribution positions of number continuity, extract the corresponding node pair index and locate the graphic rendering area, and superimpose status abnormal symbols on each abnormal node pair in the drug IND stage task graphic structure respectively, and update the abnormal tile border identification information to generate a visualization structure set of the drug IND stage task layer.

8. A visual management system for the research progress in the IND stage of a drug, characterized in that, Executed according to the drug IND stage research progress visualization management method according to any one of claims 1-7, the system includes: The sequence offset recognition module obtains the sequence numbers of the preclinical pharmacodynamic test nodes in the original plan in the drug IND stage research task process, and compares the numbers according to the planned and executed sequences to generate a task node sequence offset set; The delay layer construction module judges the offset direction of the preclinical pharmacodynamic test nodes in the task node sequence offset set, identifies the offset continuous node groups and screens the delayed nodes, constructs a closed boundary layer, and generates a delay section layer structure set; The task urgency mapping module obtains the planned completion date of the preclinical pharmacodynamic test nodes, the current system time and the remaining execution duration in the delay section layer structure set to calculate the task urgency gradient value, matches the gray brightness interval to render the node tiles, and generates a task urgency gray coding layer; The task density recognition module obtains the start and end times of the pharmacokinetic detection nodes corresponding to the preclinical pharmacodynamic test nodes in the task urgency gray coding layer, statistically calculates the time overlap rate to determine dense node pairs, and obtains a node dense area structure set; The abnormal node annotation module obtains the sample number sequence bound to each node pair in the node dense area structure set, compares the continuity of the sample numbers to identify abnormal node pairs, and generates a visualization structure set of the drug IND stage task layer.

Images