A drug IND phase research progress visual management method and system

Abstract

The present application relates to the technical field of data visualization, in particular to a drug IND stage research progress visualization management method and system, extracts efficacy test node sequence offset, identifies delay section and dense abnormal node, calculates task urgency, and generates drug IND stage task graph layer visualization structure set. Through the comparison of the sequence numbers of the task nodes at the planning and execution levels, the present application extracts the node offset direction and constructs a closed layer based on the offset aggregation, can group and boundary express the time sequence abnormal node, forms a structured area instead of an isolated abnormal point, introduces the planned time and the remaining period of the nodes in these areas to calculate the urgency gradient value, forms a mapping reference with continuous numerical values, establishes a gray rendering by corresponding with the parameters of the brightness interval, and makes the time pressure obtain intuitive expression in the visual dimension, instead of the traditional single expression mode of node color marking.

Description

Technical Field

[0001] This invention relates to the field of data visualization technology, and in particular to a method and system for visual management of drug IND stage research progress. Background Technology

[0002] The field of data visualization technology encompasses the processing of various types of data through graphical transformation, display, and interaction. The core of this technology lies in using image processing, graphics generation, and data mapping to transform raw data into easily understandable and analyzable visual graphics, assisting users in identifying trends, comparing information, and making decisions.

[0003] The drug IND (Investigational New Drug) phase research progress visualization management method refers to the means of visually representing and managing the time nodes, task execution, data submission and approval processes involved in the research preparation stage before clinical trials of new drugs. It mainly covers the setting of data collection rules, timeline mapping of research nodes, construction of status tracking logic for key tasks, and graphical annotation of document progress, specifically achieved through time series analysis, setting of task completion identification rules, and structured mapping of research process forms.

[0004] Existing technologies focus on visualizing research tasks, displaying project status by annotating time nodes, task execution, and data submission. However, they fail to quantify and express implicit problems such as structural delays, sequence shifts, intensive execution, and sample anomalies that occur during the process. They also lack the ability to model the closed structure of task shifts. When the execution node sequence is misaligned, only changes in individual node information can often be observed, making it difficult to detect phased time misordering of node groups, further hindering the overall identification of task slippage. Due to the lack of a urgency mapping mechanism based on continuous values, it is impossible to distinguish between high-urgency tasks and regular tasks in a graphical dimension, resulting in ambiguous scheduling judgments. Time overlap between nodes relies on manual identification, easily overlooking high-frequency parallel task groups. Sample number management is mainly based on static field records, failing to combine with a graphical interface to express its internal sequence logic or abnormal states. For example, sample skipping numbers are often missed or incorrectly classified as data problems, lacking structural localization capabilities, making it difficult to trace and locate problems. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by proposing a method and system for visual management of drug IND research progress.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a method for visual management of drug IND stage research progress, comprising the following steps:

[0007] S1: Obtain the sequential number of the preclinical efficacy trial node in the original plan in the drug IND stage research task flow, and compare the number with the execution order to generate a task node order offset set.

[0008] S2: Determine the offset direction of the preclinical efficacy trial nodes in the task node order offset set, identify the offset continuous node group and filter the delayed nodes, construct a closed boundary layer, and generate a delay segment layer structure set;

[0009] S3: Obtain the planned completion date, current system time, and remaining execution time of the preclinical efficacy trial nodes in the delay segment layer structure set; calculate the task urgency gradient value; match the grayscale brightness interval rendering node blocks; and generate a task urgency grayscale encoding layer.

[0010] S4: Obtain the start and end times of the pharmacokinetic detection nodes to which the preclinical pharmacodynamic test nodes belong in the task urgency grayscale encoding layer, count the time overlap rate to determine dense node pairs, and obtain the node dense region structure set.

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

[0012] The present invention improves upon the following: 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 segment layer structure set includes a delay aggregation segment range identifier, layer closed boundary frame parameters, and task number offset filtering records; the task urgency grayscale encoding layer includes a task urgency level interval mapping table, grayscale brightness parameter binding configuration, and node tile visual rendering parameters; the node dense region structure set includes task overlap rate comparison results, dense node pairing index, and time segment conflict marker labels; and the drug IND stage task layer visualization structure set specifically includes abnormal node identifier graphics, tile boundary graphic status styles, and sample number continuity anomaly annotations.

[0013] The present invention improves upon this invention by obtaining the sequential numbers of preclinical efficacy trial nodes in the original plan within the drug IND stage research task flow, and comparing these numbers with the execution order to generate a task node sequence offset set. The specific steps are as follows:

[0014] S101: Obtain the node list of the preclinical pharmacodynamic trial nodes in the drug IND stage research task flow, extract the sequential number of the preclinical pharmacodynamic trial nodes in the original task decomposition table, record the correspondence between the sequential number and the node unique identifier, and generate the node sequential number mapping result.

[0015] S102: Call the unique identifier of the preclinical pharmacodynamic trial node in the node sequence number mapping result, extract the registration execution time in the task progress record data, sort all preclinical pharmacodynamic trial nodes from earliest to latest according to the registration execution time, mark the sort number in sequence as the current sequence number, and construct the correspondence between the sort number and the node unique identifier to generate the actual execution sorting sequence of the node.

[0016] S103: Based on the node sequence number mapping result and the unique identifier of the node in the actual execution sorting sequence, the planned sequence number of each preclinical efficacy trial node is matched with the current sequence number, the number difference value between the two numbers is calculated and the direction attribute is recorded, the number offset value of each node in the sequence arrangement is integrated, and a task node order offset set is generated.

[0017] The present invention has the following improvements:

[0018] The specific steps for determining the offset direction of preclinical efficacy trial nodes in the task node order offset set, identifying continuous offset node groups and filtering delayed nodes, constructing a closed boundary layer, and generating a delay segment layer structure set are as follows:

[0019] S201: Based on the number offset value of the preclinical pharmacodynamic test node recorded in the task node order offset set, extract the offset value and positive and negative direction mark of each node, identify whether the node offset direction is ahead or behind according to the positive and negative state of the offset value, and sort the identification results according to the current execution time of the node and mark the corresponding node number to generate a node offset direction identification sequence.

[0020] S202: Call the execution time data of adjacent nodes in the node offset direction identifier sequence, compare the time interval between node pairs according to the time order, group node pairs whose time interval does not exceed the node delay aggregation judgment value into the same group, and extract the number offset value corresponding to the consecutive nodes in the group to obtain the delay node group offset value sequence.

[0021] S203: Based on the number offset value of the node in the offset numerical sequence of the delayed node group, compare the numerical values ​​with the set delay judgment threshold, filter out continuous node segments whose number offset values ​​exceed the delay judgment threshold, locate the first and last nodes in each continuous node segment to construct a closed boundary range, obtain the boundary coordinates and establish the region layer border, and generate a delay segment layer structure set.

[0022] The present invention improves upon this invention by obtaining the planned completion date, current system time, and remaining execution time of the preclinical efficacy trial nodes in the delay segment layer structure set, calculating the task urgency gradient value, matching grayscale brightness intervals to render node blocks, and generating a task urgency grayscale encoding layer. The specific steps are as follows:

[0023] S301: Obtain the planned completion date, current time and remaining execution time of each preclinical pharmacodynamic trial node in the delay segment layer structure set, calculate the task urgency gradient value corresponding to each node, and generate a node urgency gradient value 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 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.

[0025] S303: Based on the brightness parameters corresponding to the preclinical efficacy trial nodes in the task urgency brightness mapping parameter set, match the node graphic region index table, apply the grayscale brightness parameters to the brightness rendering channel of the node graphic region, establish the graphic brightness rendering configuration for each node, and generate a task urgency grayscale encoding layer.

[0026] The present invention improves upon this invention by obtaining the start and end times of the pharmacokinetic detection nodes to which the preclinical pharmacodynamic trial nodes in the task-urgent grayscale encoding layer belong, calculating the node execution time overlap rate, and identifying dense node pairs to obtain the node-dense region structure set. The specific steps are as follows:

[0027] S401: Obtain the pharmacokinetic detection node corresponding to each preclinical pharmacodynamic test node in the task urgency grayscale 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 set of node pair overlap ratio values.

[0028] S402: Based on the task time period overlap ratio of the node pairs in the node pair overlap ratio value set, and according to the task density judgment threshold, the overlap ratio of each node pair is numerically compared, all node pairs with overlap ratios greater than the task density judgment threshold are marked, and the node pair index numbers are extracted and recorded as task-dense structural units to generate a node-dense region structure set.

[0029] The present invention improves upon this invention by obtaining the sample number sequence bound to each node pair in the dense node region structure set, comparing the continuity of sample numbers to identify abnormal node pairs and marking abnormal graphic states, and generating a visualization structure set for the drug IND stage task layer. The specific steps are as follows:

[0030] S501: Obtain the sample number sequence corresponding to each node pair in the node-dense region structure set, extract the sample number lists of preclinical pharmacodynamic test nodes and pharmacokinetic detection nodes respectively, and generate a set of node pair sample number sequences.

[0031] S502: Based on the arrangement order of each group of sample number data in the sample number sequence set of the node, calculate and mark the position of the consecutive number difference, and detect the position of duplicate and reversed numbers in the number list to determine whether there are skipped numbers, duplicates or misaligned sequences, and obtain the position of abnormal distribution of number continuity.

[0032] S503: Based on the skipped, repeated, or misaligned numbers marked in the serial number continuity anomaly distribution position, extract the corresponding node pair index and locate the graphic rendering area. In the graphic structure of the drug IND stage task, update the abnormal block border identification information for each abnormal node pair superimposed state abnormal symbol, and generate the drug IND stage task layer visualization structure set.

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

[0034] The sequence offset identification module obtains the sequential number of the preclinical efficacy trial node in the drug IND stage research task flow in the original plan, and compares the numbering according to the plan and execution order to generate a task node sequence offset set.

[0035] The delay layer construction module determines the offset direction of the preclinical efficacy trial nodes in the task node order offset set, identifies the offset continuous node group and filters the delayed nodes, constructs a closed boundary layer, and generates a delay segment layer structure set.

[0036] The task urgency mapping module obtains the planned completion date, current system time, and remaining execution time of the preclinical efficacy trial nodes in the delay segment layer structure set, calculates the task urgency gradient value, matches the grayscale brightness interval rendering node blocks, and generates a task urgency grayscale encoding layer.

[0037] 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 grayscale encoding layer, calculates the time overlap rate to determine dense node pairs, and obtains the node dense region structure set.

[0038] The abnormal node annotation module obtains the sample number sequence bound to each node pair in the node-dense region structure set, compares the continuity of sample numbers to identify abnormal node pairs, and generates a visual structure set of the task layer for the drug IND stage.

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

[0040] In this invention, by comparing the sequential numbering of task nodes at the planning and execution levels, the node offset direction is extracted, and a closed layer is constructed based on offset aggregation. This enables the grouping and boundary-based expression of time-series anomalous nodes, forming structured regions rather than isolated anomalous points. For nodes within these regions, urgent gradient values ​​of planned time and remaining cycles are calculated, using continuous values ​​to form a mapping benchmark. Grayscale rendering is established by corresponding parameters with brightness ranges, allowing time pressure to be visually expressed intuitively, replacing the traditional single-expression method of node color marking. The introduction of the time overlap ratio between nodes quantifies parallel conflict relationships, with overlapping data driving dense... Structural identification drives the evolution of relationships between nodes from single-point tracking to automatic identification of interaction density. The continuity analysis of sample numbers not only expands the dimensions of sampling of task execution results, but also maps number anomalies to layer areas, forming a visual feedback label. This constitutes a continuous processing flow with execution offset as the starting point, time pressure mapping as the central axis, and task conflict and sample anomaly as the end points. This forms a layer data-driven structural anomaly identification system. Through the introduction of multi-dimensional correlation parameters and the structured analysis of sample coding mechanisms, abstract abnormal states in the research management process are expressed temporally and graphically, avoiding information omissions and node jumps. Attached Figure Description

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

[0042] Figure 2 This is a detailed flowchart of step S1 of the present invention;

[0043] Figure 3 This is a detailed flowchart of step S2 of the present invention;

[0044] Figure 4 This is a detailed flowchart of step S3 of the present invention;

[0045] Figure 5 This is a detailed flowchart of step S4 of the present invention;

[0046] Figure 6 This is a detailed flowchart of step S5 of the present invention;

[0047] Figure 7 This is a system module diagram of the present invention. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of this invention clearer, the 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 merely illustrative and not intended to limit the invention.

[0049] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0050] Please see Figure 1 This invention provides a technical solution: a method for visual management of drug IND stage research progress, comprising the following steps:

[0051] The drug IND phase research task process is divided into multiple nodes. The sequential number of the preclinical pharmacodynamic test node in the original plan is obtained. The registration execution time data of the preclinical pharmacodynamic test node is called. All preclinical pharmacodynamic test nodes are sorted according to the execution time and marked with actual sequential number. Then, the planned sequential number of the preclinical pharmacodynamic test node is matched with the actual sequential number obtained by sorting. The number offset value of the preclinical pharmacodynamic test node in the sequential arrangement is calculated and recorded to generate a task node order offset set.

[0052] The system determines whether the offset direction of the preclinical efficacy trial nodes in the task node order offset set is forward or backward. It identifies the node groups whose continuous time interval does not exceed the node delay aggregation judgment value in the judgment results. It compares the number offset value in each node group with the preset delay judgment threshold, filters the task node aggregation area that exceeds the delay judgment threshold, and marks the first and last nodes of the aggregation area as closed boundary points. It establishes the layer closed frame of the aggregation area and generates the delay segment layer structure set.

[0053] The system obtains the planned completion date, current system time, and remaining execution time of the preclinical pharmacodynamics trial nodes in the delayed segment layer structure to calculate the corresponding task urgency gradient value. The task urgency is divided into multiple level segments, and each segment is assigned a set of task tile grayscale brightness ranges. The grayscale brightness range corresponding to the current task urgency gradient value is determined, and the grayscale brightness range is used as the tile brightness parameter to render the graphic area of ​​the preclinical pharmacodynamics trial node, generating a task urgency grayscale encoding layer.

[0054] The actual start and end times of the pharmacokinetic detection nodes corresponding to the preclinical pharmacodynamic test nodes in the task urgency grayscale encoding layer are obtained. The task overlap rate is calculated according to the time period range. The task overlap rate is compared with the task density judgment threshold. Node pairs of preclinical pharmacodynamic test nodes and pharmacokinetic detection nodes that exceed the task density judgment threshold are marked to obtain the node density region structure set.

[0055] The sample number sequence bound to each node pair in the dense node region structure set is obtained. The skip positions, repeated positions and misaligned positions in the number sequence are checked to determine whether there is any non-continuous numbering in the sample number sequence of two nodes. If there is non-continuous numbering, the corresponding node pair is embedded with the abnormal state graphic symbol in the task graphic structure of the drug IND stage, and the task layer visualization structure set of the drug IND stage is generated.

[0056] The task node sequence offset set includes sequence number mapping relationship, sequence misalignment direction identifier, and node execution sequence index. The delay segment layer structure set includes delay aggregation segment range identifier, layer closed boundary frame parameters, and task number offset filtering records. The task urgency grayscale encoding layer includes task urgency level interval mapping table, grayscale brightness parameter binding configuration, and node tile visual rendering parameters. The node dense area structure set includes task overlap rate comparison results, dense node pairing index, and time segment conflict marker labels. The drug IND stage task layer visualization structure set specifically includes abnormal node identifier graphics, tile boundary graphic status style, and sample number continuity anomaly annotation.

[0057] Please see Figure 2 The specific steps for obtaining the sequential numbers of preclinical efficacy trial nodes in the drug IND stage research task flow within the original plan, and comparing these numbers with the execution order to generate a task node sequence offset set are as follows:

[0058] S101: Obtain the node list of the preclinical pharmacodynamic trial nodes in the drug IND stage research task flow, extract the sequential number of the preclinical pharmacodynamic trial nodes in the original task decomposition table, record the correspondence between the sequential number and the node unique identifier, and generate the node sequential number mapping result.

[0059] To obtain the node list for preclinical efficacy trials in the drug IND phase research task flow, it is necessary to extract all nodes marked "preclinical efficacy trials" from the task decomposition structure. This process is typically based on screening the structured task data in the project management system. First, records categorized as "preclinical efficacy trials" are extracted from all task nodes using the system-defined node category field. This is done using task management tools such as Project, JIRA, or a self-built database with structure fields. For example, if the task table contains a "node type" field, nodes with a value of "efficacy trial" are selected. Next, the sequential number of this node in the original task structure needs to be extracted. The sequential number is generally generated by the arrangement order in the task tree structure and is usually in integer form, such as 001, 002, 003, etc. To ensure accurate numbering, the "sort field" or "priority field" in the task structure can be used. For example, if the node ID is "ND001", its sequential number in the structure is 3. Next, a mapping relationship between the sequential number and the unique identifier of the node needs to be established. A key-value pair structure is created programmatically for mapping. If implemented using Python code, a dictionary structure such as {3: "ND001", 4: "ND002"} is the required mapping result, where the sequential 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 during the execution process. The above process needs to consider the uniqueness and completeness of the node information in the task decomposition table. If there are multiple nodes with the same category but different IDs, it should be ensured that the extracted nodes are valid nodes in the task version. For example, in the IND stage, drug A has 5 nodes in its preclinical efficacy trial, numbered "ND010" to "ND014" in sequence. The corresponding sequential numbers in the task structure are 1 to 5. The mapping result is {1: "ND010", 2: "ND011", 3: "ND012", 4: "ND013", 5: "ND014"}. This data constitutes the basic input data for subsequent steps.

[0060] S102: Call the unique identifier of the preclinical pharmacodynamic trial node in the node sequence number mapping result, extract the registration execution time in the task progress record data, sort all preclinical pharmacodynamic trial nodes from earliest to latest according to the registration execution time, mark the sort number in sequence as the current sequence number, construct the correspondence between the sort number and the node unique identifier, and generate the actual execution sorting sequence of the node.

[0061] The unique identifier of the preclinical pharmacodynamic trial node in the node sequence number mapping result is retrieved. Specifically, all preclinical pharmacodynamic trial node IDs are extracted from the mapping structure obtained in the previous section, such as "ND010" to "ND014" in {1: "ND010", 2: "ND011", 3: "ND012", 4: "ND013", 5: "ND014"}. The task progress registration records corresponding to these IDs are then searched in the project task database, and the "Registration Execution Time" field is extracted. This field is in timestamp format. For example, "2024-03-01 10:00:00", then sort all nodes in ascending order according to their registration execution time. For example, if "ND012" was registered on "2024-03-01", "ND010" on "2024-03-05", and "ND011" on "2024-03-07", then the sorted nodes would be "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, each node is assigned a current sequence number, starting from 1 and incrementing according to its position in the sort. 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 represents the actual sorting sequence of the nodes. For example, the original plan for the preclinical efficacy trial of drug B... The order is from "ND020" to "ND024", but the actual execution time record shows that the order is "ND022", "ND020", "ND021", "ND023", "ND024". Therefore, the mapping between the sorting number and the unique identifier of the node is {1: "ND022", 2: "ND020", 3: "ND021", 4: "ND023", 5: "ND024"}. This structure is used for subsequent offset calculations. In the sorting logic, nodes with missing registration time should be excluded or the most recent update time should be used as a substitute.

[0062] S103: Based on the node sequence number mapping result and the unique identifier of the node in the actual execution sorting sequence, the planned sequence number of each preclinical pharmacodynamics trial node is matched with the current sequence number, the number difference value between the two numbers is calculated and the directional attribute is recorded, the number offset value of each node in the sequence arrangement is integrated, and a task node order offset set is generated.

[0063] Based on the mapping result of node sequence numbers and the unique identifiers of nodes in the actual execution sequence, the planned sequence number of each node is mapped to its actual execution number. The difference between the numbers is compared one by one, and the offset direction is recorded. The difference is defined as Δ. i=i actual -i planned , where Δ i This represents the sequential offset value of the i-th node, i actual Indicates the sequential number of the node in the actual sorting process, i planned This indicates the planned sequence number of the node in the original task decomposition. If Δ i If Δ > 0, it indicates that the node sequence is lagging, and the direction is marked as "+"; if Δ i If Δ < 0, it indicates that the node order is advanced, and the direction is marked as "-"; if Δ i =0 indicates that the node order has not changed, and the direction is marked as "0". This process is implemented by traversing the node list and calculating one by one. For example, if the planned order of "ND010" is 2, and the actual execution order is 3, then Δ = 3 - 2 = 1, and the direction is "+". The order offset data structure of each node can be constructed, such as {"ND010": {"planned": 2, "actual": 3, "delta": 1, "direction": "+"}}. For example, in the IND stage of drug C, the planned order and actual order of its test nodes are as follows: Planned: ["ND030", "ND031", "ND032", "ND033"]. [ND034], actual: [ND031", "ND030", "ND032", "ND034", "ND033"], then the planned order of node "ND031" is 2, the actual order is 1, Δ = 1 - 2 = -1, and the direction is "-"; based on this, construct a complete set of task node order offsets, such as {"ND030": {"": 1, "direction": "+"}, "ND031": {"": -1, "direction": "-"}...}; during actual execution, the absolute value of the offset Δ, |Δ|, can be used to judge the sorting stability. If |Δ| > 2, it is considered a high offset node. The logic for setting this "high offset" threshold is as follows: Where T is the offset threshold, N is the total number of task nodes, and the symbol is... This indicates rounding up. If N = 9, then... At this point, |Δ|≥3 is considered a significant offset node. This criterion can dynamically adapt to research tasks with different numbers of nodes, thereby forming a stable basis for node execution sequence offset analysis.

[0064] Please see Figure 3 The specific steps for determining the offset direction of preclinical efficacy trial nodes in the task node order offset set, identifying continuous offset node groups and filtering delayed nodes, constructing a closed boundary layer, and generating a delay segment layer structure set are as follows:

[0065] S201: Based on the number offset value of the preclinical pharmacodynamic trial node recorded in the task node order offset set, extract the offset value and positive / negative direction mark of each node, identify whether the node offset direction is ahead or behind according to the positive / negative state of the offset value, and sort the identification results according to the current execution time of the node and mark the corresponding node number to generate a node offset direction mark sequence.

[0066] First, the system reads the order offset values ​​of all preclinical pharmacodynamic trial nodes from the established data structure. For each node, the system traverses its key-value pair structure, reading the "delta" (offset value) and "direction" (offset direction) fields. For example, if the offset value of node "ND101" is -2, its direction is forward, and the offset direction is "-". Similarly, if the offset value of "ND102" is 3, the offset direction is "+", indicating a backward shift. The offset direction is identified by judging the positive or negative state of the offset value: a negative value indicates forward, a positive value indicates a backward shift, and a value of 0 indicates no offset. This judgment can be achieved using conditional statements. Then, the system calls the actual execution time field of each node, which is a specific timestamp, such as "2024-05-01 09:00:00". All nodes are then sorted in ascending order according to time. The sorting algorithm can call Python's built-in `sorted` function or the database `ORDERBY` statement, with the sorting field being time. After sorting, the node number and its offset direction are re-sorted according to the corresponding sorting position. The original structure, {"ND101":{"delta":-2","direction":"-","time":"2024-04-02"}, "ND102":{"delta":3","direction":"+","time":2024-04-03"}}, becomes ["ND101-early", "ND102-late"] after sorting. The offset direction fields "early" and "late" are determined by the direction identifier "-". The "+" conversion generates a node offset direction identifier sequence. For example, the preclinical nodes of drug D are "ND201" (2024-03-01), "ND202" (2024-03-03), and "ND203" (2024-03-06) according to the execution time. Their offset values ​​are -1, 0, and 2, respectively, and the corresponding identifiers are "early", "no offset", and "delayed". Then the offset direction identifier sequence is ["ND201-early", "ND202-no offset", "ND203-delayed"].

[0067] S202: Call the execution time data of adjacent nodes in the node offset direction identifier sequence, compare the time interval between node pairs according to the time order, group node pairs whose time interval does not exceed the node delay aggregation judgment value into the same group, and extract the number offset value corresponding to the consecutive nodes in the group to obtain the delay node group offset value sequence.

[0068] To retrieve the execution time data of adjacent nodes in the node offset direction identifier sequence, it is necessary to extract the time interval of each pair of adjacent nodes. Let the time of node "ND301" be 2024-04-01 and "ND302" be 2024-04-03, with an interval of 2 days. If the delay aggregation judgment value is set to 3 days, then this node pair meets the grouping condition. Continuing to judge "ND303", its execution time is 2024-04-04, with an interval of 1 day from "ND302", still less than 3 days, so the three can be merged into one delayed node group. If the time of "ND304" is 2024-04-10, with an interval of 6 days from the previous node "ND303", exceeding the judgment value, then it is not included in the group. Each group serves as the first node of the next group. The above process can be achieved by traversing the node time series using a sliding window. After each group is formed, the offset values ​​of each node number need to be extracted. For example, group 1 consists of “ND301”, “ND302”, and “ND303”, with offset values ​​of -2, 0, and 3 respectively, so the offset value sequence is [-2, 0, 3]. Group 2 consists of “ND304” and “ND305”, with offset values ​​of 1 and 2, so the sequence is [1, 2]. The offset value sequence of this delayed node group will serve as the input basis for subsequent judgment of delayed segments. The entire grouping operation must satisfy that the time difference of all adjacent nodes does not exceed the set judgment value, and the number of nodes in each group is at least 2, satisfying the basic requirement of continuity.

[0069] S203: Based on the offset value of the node number in the offset numerical sequence of the delayed node group, compare the numerical values ​​with the set delay judgment threshold, filter out continuous node segments whose offset values ​​exceed the delay judgment threshold, locate the first and last nodes in each continuous node segment to construct a closed boundary range, obtain the boundary coordinates and establish the region layer border, and generate a delay segment layer structure set.

[0070] Based on the offset values ​​of nodes in the delay node group offset sequence, the system compares the offset values ​​in each group with a set delay judgment threshold. This threshold can be calculated by dividing the total number of nodes in the task by 5. For example, if the task contains 20 nodes and the delay judgment threshold is 4, it means that if the absolute value of the offset value exceeds 4, it is a delayed node. The system checks the offset value of each group sequence sequentially and determines whether it meets the delay standard. For example, group 1 is [-1, 5, 6], where 5 and 6 are both greater than 4, and are judged as delayed nodes. Two consecutive delayed nodes "ND402" and "ND403" constitute a delayed segment. Another example is group 2, [1, 3], where the maximum offset is 3, which does not exceed the threshold and does not constitute a delayed segment. The system identifies the first and last nodes of the segment, takes "ND402" as the start node and "ND403" as the end node, and determines the closed boundary range. ["ND402" - "ND403"], then obtain the coordinate information of these two nodes in the layer. The horizontal axis is the timestamp, and the vertical axis is the sequential number. For example, the time of "ND402" is 2024-05-01 and the number is 7, and the time of "ND403" is 2024-05-04 and the number is 8. Then the coordinates of the upper left corner of the area are (2024-05-01, 7) and the coordinates of the lower right corner are (2024-05-04, 8). Construct a closed rectangular border, which is recorded as a layer element. After all delay segments are identified, the boundary set is as follows: [Boundary 1: ("ND402" - "ND403"), Boundary 2: ("ND407" - "ND409")], which are all based on the time and number of the nodes to form the coordinate range of the border, and form a layer structure set, which is used to construct a graphical description of the delay segment.

[0071] Please see Figure 4 The specific steps for obtaining the planned completion date, current system time, and remaining execution time of preclinical efficacy trial nodes in the delayed segment layer structure, calculating the task urgency gradient value, matching grayscale brightness intervals to render node tiles, and generating a task urgency grayscale encoding layer are as follows:

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

[0073] After obtaining the planned completion date, current time, and remaining execution time for each preclinical pharmacodynamic trial node in the delay segment layer structure set, the units of these three data points need to be unified to a consistent dimension to satisfy the logical relationships in the formula. The project system will automatically traverse all node IDs from the node layer structure index, calling the node attribute table fields one by one to extract fields including: planned completion date (e.g., 2025-05-10), current system time (e.g., 2025-03-30), and remaining execution time (e.g., 36 hours). The system will then uniformly convert the date to a "relative number of days" by referring to the project's set start date (e.g., 2025-01-01, set as day 0). The system will then use a date conversion function to convert the planned completion time (denoted as Date) to a "relative number of days." plan ) and the current time (denoted as Date) now ) respectively convert to the Dth number starting from the start date plan D now Days, for example, a planned completion date of May 10, 2025, is converted to day 130; the current date, March 30, 2025, is converted to day 89. Remaining execution time, expressed in hours (e.g., 36 hours), must be uniformly converted to days. This is the original remaining execution time, where h represents the number of hours. The formula used is:

[0074]

[0075] Calculate the task urgency gradient value U;

[0076] Among them, D plan It is the "relative number of days" corresponding to the planned completion date, counting from the project start date as Day 0 (unit: days). now It is the relative number of days corresponding to the current system time (unit: days). It is the remaining execution time. δ is the ideal execution cycle, d is the number of days, derived from historical data statistics, δ is a minimal constant (e.g., 0.01) to prevent the denominator in the square root term from approaching 0, ε is a stabilizing correction term (e.g., 0.01) to prevent oscillations when dividing by zero or when the inverse trigonometric function reaches an extreme value, α is the error term control coefficient, which adjusts the sensitivity of the plan time difference and the remaining cycle offset (unitless, usually taken as 1.0), and arcsin() is the arcsine function, which calculates the radian value of the angle corresponding to a given proportional offset, used to compress the rate of change of the deviation term.

[0077] Suppose that node A has the following parameters: D plan =130: The plan is to be completed on May 10th, D now =89: The current time is March 30th. After conversion The ideal completion cycle of this node under standard resource configuration, δ=0.01, ε=0.01: positive constants to prevent division by zero, α=1.0: adjustment coefficient to control the strength of the error term.

[0078] Calculation of the square root term in the denominator:

[0079] Arcsine calculation:

[0080] Gradient value calculation:

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

[0082] 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 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] For each task urgency gradient value in the node urgency gradient value sequence, the corresponding task urgency level division interval table needs to be called to determine its numerical segment. The system, based on preset interval divisions such as [0, 1) = "Very Low", [1, 2) = "Relatively Low", [2, 4) = "Medium", [4, 6) = "High", [6, 8) = "Very High", and [8, +∞) = "Extremely High", categorizes each node urgency value into its corresponding segment and then looks up the grayscale brightness parameter range mapped to that segment. For example, the "High" segment corresponds to grayscale brightness values ​​of [160, 192]. The system performs linear interpolation on the current node gradient value within its respective level segment to convert it into a grayscale brightness value. The linear interpolation formula used in the interpolation process is as follows:

[0084]

[0085] Where L is the final grayscale value of the target node, L min L max These are the lower and upper limits of the brightness range corresponding to this gradient level, where U is the urgent gradient value of the current node. min U max It represents the upper and lower limits of the grade range in which the current gradient value is located.

[0086] Taking this node as an example, its gradient value is 4.74, falling into the [4, 6) segment, which is the "high" level, and the corresponding brightness range 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 set of task-urgent brightness mapping parameters, such as {171.84, 148.12, 200.00, ...}, which provides input parameters for the brightness rendering of graphics nodes.

[0087] S303: Based on the brightness parameters corresponding to the preclinical efficacy trial nodes in the task urgency brightness mapping parameter set, match the node graphic region index table, apply the grayscale brightness parameters to the brightness rendering channel of the node graphic region, establish the graphic brightness rendering configuration for each node, and generate the task urgency grayscale encoding layer.

[0088] Based on the brightness values ​​corresponding to the preclinical efficacy trial nodes in the task-urgent brightness mapping parameter set, the system loads a layer region index table in the graphics module. This index table contains the pixel region coordinate range corresponding to each node in the rendering layer. For example, the graphics region of node A is a rectangular area with a horizontal axis of 120 to 140 and a vertical axis of 200 to 220. The system matches the node number with the graphics region index and writes the grayscale brightness value of the node into the grayscale rendering channel of the corresponding region. The grayscale brightness value of each node is applied to all pixels in its corresponding region. For example, if the brightness value corresponding to node A is 171.84, the system will uniformly set all pixel values ​​in its pixel region to 171 to ensure that the node forms a clear visual intensity in the layer. The system performs the above brightness value injection operation node by node until all trial nodes are traversed, completing the graphics brightness rendering configuration. All rendering actions are executed in the pixel rendering engine. The rendered brightness encoding data is synthesized into the final task-urgent grayscale encoding layer through the layer update mechanism, which serves as one of the input sources for the layer visualization module.

[0089] Please see Figure 5 The specific steps for obtaining the start and end times of the pharmacokinetic detection nodes corresponding to the preclinical pharmacodynamic trial nodes in the task-urgent grayscale encoding layer, calculating the node execution time overlap rate, identifying dense node pairs, and obtaining the node dense region structure set are as follows:

[0090] S401: Obtain the pharmacokinetic detection node corresponding to each preclinical pharmacodynamic test node in the task urgency grayscale 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 set of node pair overlap ratio values.

[0091] To obtain the corresponding pharmacokinetic testing node for each preclinical pharmacodynamic trial node in the task-urgent grayscale encoding layer, it is necessary to first establish an index table of the attribution relationship between the two types of nodes. This table is constructed using the hierarchical dependency field defined in the project workflow. Each pharmacodynamic trial node establishes a logical mapping relationship with a pharmacokinetic testing node through the field "Attribution PK Node ID". For example, if the pharmacodynamic node E104 belongs to the pharmacokinetic node PK21, then a node pair relationship of E104→PK21 is established. The system reads the list of all rendered pharmacodynamic trial nodes from the task-urgent grayscale encoding layer, and for each node... The system calls the layer index interface and the task schedule table to extract the task time periods, i.e., the "start time" and "end time" fields. For example, node E104 has a start time of 2025-03-28 and an end time of 2025-04-03, which corresponds to days 87 to 93 of the project. The corresponding pharmacokinetics testing node PK21 has a start time of 2025-04-01 and an end time of 2025-04-07, which corresponds to days 91 to 97 of the project. The system then calculates the overlap between the task time periods of the two nodes by dividing the number of days of intersection by the number of days of union. The overlap ratio is calculated using the following formula:

[0092]

[0093] in, Indicates the start and end times of the efficacy test milestones. This represents the time period attributable to the pharmacist node, with +1 indicating that the start and end days are included. In actual execution, if the E104 interval is [87, 93] and the PK21 interval is [91, 97], the intersection is [91, 93] (3 days), and the union is [87, 97] (11 days). Substituting these values ​​into the calculation yields: The system records this ratio as the overlap ratio of node pair (E104, PK21), calculates the time period overlap ratio of all node pairs in turn, and generates a set of node pair overlap ratio values ​​such as {(E104, PK21, 0.273), (E105, PK22, 0.615), (E106, PK22, 0), ...}, where each item contains the pharmacodynamic node number, its associated pharmacokinetic node number, and its task time period overlap ratio value.

[0094] S402: Based on the task time period overlap ratio of the node pairs in the node pair overlap ratio value set, and according to the task density judgment threshold, the overlap ratio of each node pair is numerically compared, all node pairs with overlap ratios greater than the task density judgment threshold are marked, and the node pair index number is extracted and recorded as a task-dense structural unit to generate a node-dense region structure set.

[0095] Based on the task time period overlap ratio of node pairs in the node pair overlap rate value set, it is necessary to perform numerical comparison and conditional filtering in conjunction with the task density judgment threshold. This threshold is set to determine whether there is actual task execution overlap, and is generally set to 0.6, that is, 60% of the time periods overlap to be considered a task-intensive structure. This value can be set by setting the mean and standard deviation of the average overlap ratio of previous projects, and the reasonable range is between [0.5, 0.7]. If the value is set to 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 the threshold, the system will determine whether the task is task-intensive. When the overlap ratio is 0.6, the node pair is marked as a task-dense node pair. For example, if the overlap ratio of the node pair (E105, PK22) is 0.615, which is greater than 0.6, then the condition is met and it is marked as a task-dense node pair. The node pair index number is extracted for all node pairs that meet the condition, such as "E105→PK22". The system records it in the task-dense structural unit set and constructs a dense region structure set, such as {(E105, PK22), (E110, PK24), (E111, PK24)}. Finally, a node-dense region structure set is generated, which serves as the input parameter for the subsequent time scheduling and layer conflict detection module.

[0096] Please see Figure 6 The specific steps for generating a visualization structure set for the drug IND stage task layer are as follows: Obtain the sample number sequence bound to each node pair in the dense node region structure set, compare the continuity of sample numbers to identify abnormal node pairs and mark abnormal graphical states.

[0097] S501: Obtain the sample number sequence corresponding to each node pair in the node-dense region structure set, extract the sample number lists of preclinical pharmacodynamic test nodes and pharmacokinetic test nodes respectively, and generate a set of node pair sample number sequences.

[0098] In the process of obtaining the sample number sequence associated with each node pair in the node-dense region structure set, the system first needs to read the node pair index one by one from the generated node-dense region structure set. Each element in this structure set contains a preclinical pharmacodynamic test node number and its associated pharmacokinetic detection node number. Subsequently, the system accesses the preclinical experimental sample information database and the pharmacokinetic detection sample registration table, respectively retrieving sample information records bound to the two node ID fields. The data fields typically include sample number, sample ID, sampling time, and associated node ID. The system establishes a sample number list for each node. For example, the preclinical node E301 is associated with sample numbers [S1001, S1002, S1003], corresponding to the pharmacokinetic node PK. 72 associated sample numbers are [S1003, S1004, S1005]. The system records the two lists and binds them to the node pairs to form structure items. The acquisition process supports many-to-many relationships. Some samples may appear repeatedly in the two nodes. During this process, the system needs to use a filtering function to exclude samples from non-target nodes and use intersection verification logic to ensure the accuracy of each pair of sample number lists. The final sample number data structure is organized as follows: {(E301, PK72): [[S1001, S1002, S1003], [S1003, S1004, S1005]]}. After extracting the sample numbers for each node pair in this way, the system stores all node pairs and their corresponding sample number combinations into the sample number sequence set.

[0099] S502: Based on the arrangement order of each group of sample number data in the sample number sequence set, calculate and mark the position of the consecutive number difference, and detect the position of duplicate and reversed numbers in the number list to determine whether there are skipped numbers, duplicates or misaligned sequences, and obtain the position of abnormal distribution of number continuity.

[0100] Based on the arrangement order of sample number data in each group of sample number sequences in the node's sample number sequence set, the difference between consecutive numbers is calculated and their positions are marked. The system also detects duplicate and reversed numbers in the number list to determine if there are skipped numbers, duplicates, or misaligned sequences. First, the system parses the number values ​​in each group of sample numbers, extracting the purely numerical portion and converting it into a comparable list of values. For example, [S1001, S1002, S1005] is converted to [1001, 1002, 1005]. Then, the system performs continuity checks on the sample numbers, using a difference analysis method to handle the number spacing. The difference between two adjacent numbers is calculated; if the difference is not 1, it is marked as a skipped number. For example, in the number [1001, 1002, 1005], the difference between 1002 and 1005 is 3, indicating two skipped numbers, recorded as 1002→1005. Duplicate number detection is performed by comparing the length of the original list with the length of the set after removing duplicates. For example, [1001, 1002, 1002, 1003] becomes [1001, 1002, 1003] after deduplication. The reduced length indicates a duplicate, and the duplicate position is recorded as the 2nd to 3rd position. For misalignment detection, the current number list is compared position by position with its ascending sorted result. For instance, if the current list is [1003, 1001, 1002], and the ascending sorted list is [1001, 1002, 1003], the system compares the index differences and determines the reversed position as the 1st and 2nd positions. In cases of misalignment, the system records the serial number continuity anomalies in a structured format, including skip number markers, duplicate position indexes, and reverse position markers, while also providing anomaly type labels. The system adopts the following classification criteria: if the difference is >1, it is a skip number; if the difference is =0, it is a duplicate; if the current number value is less than the previous number, it is a reverse. 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 index, forming a set of serial number continuity anomaly distribution position records.

[0101] S503: Based on the skipped, repeated, or misaligned numbers marked in the continuous abnormal distribution of the numbers, extract the corresponding node pair index and locate the graphic rendering area. In the graphic structure of the drug IND stage task, update the abnormal block border identification information for each abnormal node pair superimposed with an abnormal symbol, and generate a visual structure set of the drug IND stage task layer.

[0102] Based on the skipped, repeated, or misaligned numbers marked in the distribution of serial number anomalies, the system extracts the corresponding node pair indexes and accesses the layer structure. First, it matches the node pairs with the graphics index table to locate the graphic region of the node pair within the task's graphics structure. The graphics structure uses node pairs as the basic unit, with each node pair occupying a specific rectangular area in the graphics rendering structure. The system adds identification information to the graphic regions of all node pairs with anomalies. Anomaly types are overlaid using symbol layers; for example, skipped numbers are indicated by an "!" icon, repeated numbers by a "△" icon, and misaligned numbers by a symbol layer. The icon is located in the upper right corner of the tile area, and the border color is adjusted accordingly. The border color of the jump number is set to red (#FF0000), the repeating border is orange (#FFA500), and the misaligned border is purple (#800080). The rendering process of the graphic area is handled by the anomaly icon rendering module. The system combines the anomaly symbol primitives into texture materials according to the anomaly type and performs a local layer update operation on the target area. After the update, the layer content includes the visible status anomaly icon, and finally forms a visual structure set of task layers in the drug IND stage, which includes the status rendering content and layer coordinate information of all anomaly node pairs.

[0103] Please see Figure 7 A visualization management system for the progress of drug IND stage research, the system includes:

[0104] The sequence offset identification module obtains the sequential number of the preclinical efficacy trial node in the drug IND stage research task flow in the original plan, and compares the numbering according to the plan and execution order to generate a task node sequence offset set.

[0105] The delay layer construction module determines the offset direction of the preclinical efficacy trial nodes in the task node order offset set, identifies the continuous offset node group and filters the delayed nodes, constructs a closed boundary layer, and generates a delay segment layer structure set.

[0106] The task urgency mapping module obtains the planned completion date, current system time, and remaining execution time of the preclinical efficacy trial nodes in the delayed segment layer structure, calculates the task urgency gradient value, matches the grayscale brightness interval to render node blocks, and generates a task urgency grayscale encoding 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 grayscale encoding layer, calculates the time overlap rate to determine dense node pairs, and obtains the node dense region structure set.

[0108] The abnormal node annotation module obtains the sample number sequence bound to each node pair in the node-dense region structure set, compares the continuity of sample numbers to identify abnormal node pairs, and generates a visual structure set of the task layer for the drug IND stage.

[0109] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for visual management of drug IND phase study progress, characterized in that, The method comprises the following steps: S1: obtaining the sequence number of the preclinical pharmacodynamic test node in the original plan in the drug IND stage research task flow, and performing number comparison according to the planned and executed sequence to generate a task node sequence offset set; S2: judging the offset direction of the preclinical pharmacodynamic test node in the task node sequence offset set, identifying the offset continuous node group and screening the delayed node, constructing the closed boundary layer, and generating a delay section layer structure set; S3: obtaining the planned completion date of the preclinical pharmacodynamic test node in the delay section layer structure set, the current system time and the remaining execution time calculation task urgency gradient value, matching the gray brightness interval rendering node block to generate a task urgency gray coding layer; S4: obtaining the starting and ending time of the pharmacokinetic detection node corresponding to the preclinical pharmacodynamic test node in the task urgency gray coding layer, and counting the time overlap rate to judge the dense node pair to obtain a node dense area structure set; S5: obtaining the sample number sequence bound to each node pair in the node dense area structure set, comparing the sample number continuity to identify abnormal node pairs, and generating a drug IND stage task layer visualization structure set.

2. The drug IND phase study progress visual management method according to claim 1, characterized in that: The task node sequence offset set comprises sequence number mapping relationship, sequence dislocation direction identifier and node execution sequence index, the delay section layer structure set comprises delay aggregation section range identifier, layer closed boundary box line parameter and task number offset screening record, the task urgency gray coding layer comprises task urgency level interval mapping table, gray brightness parameter binding configuration and node block visual rendering parameter, the node dense area structure set comprises task overlap rate comparison result, dense node pairing index and time section conflict marker label, and the drug IND stage task layer visualization structure set is specifically abnormal node identifier pattern, block boundary pattern state style and sample number continuity abnormal annotation. 3.The method of claim 1, wherein: The specific steps of obtaining the sequence number of the preclinical pharmacodynamic test node in the original plan in the drug IND stage research task flow, and performing number comparison according to the planned and executed sequence to generate a task node sequence offset set are as follows: S101: obtaining the node list of the preclinical pharmacodynamic test node in the drug IND stage research task flow, extracting the sequence number of the preclinical pharmacodynamic test node in the original task decomposition table, recording the corresponding relationship between the sequence number and the node unique identifier, and generating a node sequence number mapping result; S102: calling the node unique identifier of the preclinical pharmacodynamic test node in the node sequence number mapping result, extracting the registered execution time in the task progress record data, sorting all preclinical pharmacodynamic test nodes from early to late according to the registered execution time, sequentially marking the sequence number as the current sequence number, and constructing the corresponding relationship between the sequence number and the node unique identifier to generate a node actual execution sequence. S103: According to the node sequence number mapping result and the node actual execution sequence, the planned sequence number of each preclinical pharmacology test node is positionally corresponded with the current sequence number, the number difference value between the two numbers is calculated and the direction attribute is recorded, the number offset value of each node in the sequence arrangement is integrated, and a task node sequence offset set is generated.

4. The method of claim 3, wherein the method further comprises: The specific steps of judging the offset direction of the preclinical pharmacology test node in the task node sequence offset set, identifying the offset continuous node group and screening the delay node, constructing the closed boundary layer, and generating the delay section layer structure set are as follows: S201: Based on the number offset value of the preclinical pharmacology test node recorded in the task node sequence offset set, the offset value and the positive and negative direction mark of each node are extracted, the offset direction of the node is identified as advanced or delayed according to the positive and negative state of the offset value, and the identification result is annotated according to the node current execution time to generate a node offset direction identification sequence; S202: The execution time data of adjacent nodes in the node offset direction identification sequence is called, the time interval between node pairs is compared in time sequence, node pairs with a time interval not exceeding the node delay aggregation judgment value are grouped into the same group, and the number offset value corresponding to the continuous node in the group is extracted to obtain a delay node group offset value sequence; S203: According to the number offset value of the node in the delay node group offset value sequence, the number offset value is compared with the set delay judgment threshold value, the continuous node segments with the number offset value exceeding the delay judgment threshold value are screened, and the first node and the last node in each continuous node segment are located to construct a closed boundary range, the boundary coordinates are obtained, and a region layer frame is established to generate a delay section layer structure set.

5. The method of claim 4, wherein the method further comprises: The specific steps of obtaining the planned completion date of the preclinical pharmacology test node in the delay section layer structure set, the current system time, and the remaining execution time to calculate the task urgency gradient value, matching the gray brightness interval rendering node block, and generating the task urgency gray coding layer are as follows: S301: The planned completion date, current time and remaining execution time of each preclinical pharmacology test node in the delay section layer structure set are obtained, the task urgency gradient value corresponding to each node is calculated, and a node urgency gradient value sequence is generated; S302: Based on each task urgency gradient value in the node urgency gradient value sequence, a task urgency level division interval table is called, each gradient value is identified, the corresponding level section is matched with a preset gray brightness range table, the gray brightness parameter corresponding to the current gradient value is extracted, and a task urgency brightness mapping parameter set is obtained; S303: According to the brightness parameter corresponding to the preclinical pharmacology test node in the task urgency brightness mapping parameter set, the node graph region index table is matched, the gray brightness parameter is applied to the brightness rendering channel of the node graph region, the graph brightness rendering configuration of each node is established, and a task urgency gray coding layer is generated.

6. The method of claim 5, wherein the method further comprises: The specific steps of obtaining the start and end times of the pharmacokinetics detection node to which the preclinical pharmacological test node in the task urgency gray coding layer corresponds, and statistically obtaining the node execution time overlap rate and judging the dense node pairs to obtain the node dense area structure set are as follows: S401: Obtain the pharmacokinetics detection node corresponding to each preclinical pharmacological 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 rate of each node pair, and generate a node pair overlap rate value set; S402: According to the task time period overlap rate value of the node pair in the node pair overlap rate value set, according to the task dense determination threshold, the overlap rate of each node pair is compared, and the node pairs with an overlap rate greater than the task dense determination threshold are marked, and the node pair index number record is extracted as a task dense structure unit, and a node dense area structure set is generated.

7. The method of claim 6, wherein the method further comprises: The specific steps of obtaining the sample number sequence bound to each node pair in the node dense area structure set, comparing the sample number continuity to identify abnormal node pairs and mark the graph abnormal state, and generating the drug IND stage task layer visualization structure set are as follows: S501: Obtain the sample number sequence bound to each node pair in the node dense area structure set, extract the sample number list of the preclinical pharmacological test node and the pharmacokinetics detection node respectively, and generate a node pair sample number sequence set; S502: According to the arrangement order of each sample number data in the node pair sample number sequence set, calculate the continuous number difference and position identification, and detect the repeated number and reverse sequence number position existing in the number list, judge whether there is a number skipping, repetition or sequence dislocation situation, and obtain the number continuity abnormal distribution position; S503: According to the number skipping, repetition or dislocation situation marked in the number continuity abnormal distribution position, extract the corresponding node pair index and locate the graph rendering area, superimpose the state abnormal symbol for each abnormal node pair in the drug IND stage task graph structure respectively, update the abnormal graph block border identification information, and generate the drug IND stage task layer visualization structure set.

8. A drug IND phase study progress visual management system, characterized in that, According to any one of claims 1-7, the drug IND stage research progress visualization management method is executed, and the system comprises: The order offset identification module obtains the sequence number of the preclinical pharmacological test node in the original plan in the drug IND stage research task process, and compares the numbers according to the planned and executed sequence to generate a task node order offset set; The delay layer construction module judges the offset direction of the preclinical pharmacological test node in the task node order offset set, identifies the offset continuous node group and screens the delay node, 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 pharmacological test node in the delay section layer structure set, the current system time, and the remaining execution time to calculate the task urgency gradient value, matches the gray brightness interval to render the node block, and generates a task urgency gray coding layer. The task density recognition module obtains the start and end times of the pharmacokinetics detection node to which the preclinical efficacy test node in the task urgency gray scale coding layer corresponds, and counts the time overlap rate to determine the dense node pairs to obtain a node dense area structure set; The abnormal node marking module obtains the sample number sequence bound to each node pair in the node dense area structure set, compares the sample number continuity to identify abnormal node pairs, and generates a drug IND stage task graph layer visual structure set.

Images