Centralized management method for medical equipment IoT data collection terminal

Abstract

The present application proposes a centralized management method for medical device Internet of Things data acquisition terminals, which manages multiple acquisition terminals respectively connected to at least one medical device through a configuration center located in a host computer; the configuration center includes an acquisition terminal status monitoring module and a task registration and delivery module; the acquisition terminal status monitoring module registers the acquisition terminal through a registration configuration page, and displays the operating status of the acquisition terminal that has been registered to the host computer through a monitoring page; the task registration and delivery module dynamically delivers acquisition tasks to each acquisition terminal according to the real-time resource usage status of each acquisition terminal.

Description

Technical Field

[0001] The present application relates to medical Internet of Things technology, and in particular to a centralized management method for medical device Internet of Things data acquisition terminals. Background Art

[0002] Medical equipment IoT data collection technology is often used in medical scenarios such as ICUs, wards, rehabilitation rooms, and laboratories to collect, parse, and upload data generated by medical equipment. The usual data collection solution is to connect medical equipment with a data collection terminal, parse the original data protocol through a program, and finally upload it to the data center. In actual applications, multiple data collection terminals work at the same time. However, there is currently no effective mechanism for centralized management of all collection terminals, resulting in difficulties in operation and maintenance. The following problems exist:

[0003] (1) It is impossible to centrally configure the data collection terminals. Operation and maintenance personnel need to manually configure dozens or even hundreds of terminals one by one, which is inefficient and may cause configuration errors.

[0004] (2) The operating status of the data collection terminal cannot be monitored in real time, and fault problems cannot be discovered in time, which affects the use of business data.

[0005] (3) It is impossible to uniformly schedule and manage the resource usage of the data collection terminals, resulting in load imbalance, unreasonable resource usage, and increased failure rate. Summary of the invention

[0006] In view of the above problems, the present application aims to propose a centralized management method for data acquisition terminals of the Internet of Things for medical devices, which can display the working status of the data acquisition terminals in real time. The present application aims to propose a centralized management method for data acquisition terminals of the Internet of Things for medical devices, which can assign tasks to the data acquisition terminals according to their capabilities and make the best possible use of each data acquisition terminal in the Internet of Things for medical devices.

[0007] The centralized management method for medical device IoT data collection terminals of the present application manages multiple collection terminals respectively connected to at least one medical device through a configuration center located in a host computer; The configuration center includes the acquisition terminal status monitoring module and the task registration and delivery module; The acquisition terminal status monitoring module registers the acquisition terminal through the registration configuration page, and displays the operating status of the acquisition terminal that has been registered to the host computer through the monitoring page; The task registration and delivery module dynamically delivers collection tasks to each collection terminal according to the real-time resource usage status of each collection terminal.

[0008] Preferably, the monitoring page can be further accessed to a details page, where the operation information is displayed in detail in units of tasks.

[0009] Preferably, the data of the monitoring page and the details page are derived from the collection terminal, including: error information, real-time resource usage, collection terminal operating system version number, collection terminal hardware version number, collection terminal software version number, and at least one of the collection terminal network address.

[0010] Preferably, error information and real-time resource usage are refreshed regularly; The acquisition terminal operating system version number, acquisition terminal hardware version number, acquisition terminal software version number, and acquisition terminal network address are refreshed when the acquisition terminal goes online.

[0011] Preferably, the terminal status monitoring module can restart the acquisition terminal.

[0012] Preferably, if the collection terminal has sufficient resources when executing the task, the task registration and delivery module allocates the task to the collection terminal; If the collection terminal is short of resources when executing a task, the task registration and delivery module will delete the task from the collection terminal and assign it to another collection terminal.

[0013] Preferably, when generating configuration, the task registration and delivery module distributes the configuration according to the resources of each acquisition terminal; the resources include computing resources and storage resources.

[0014] Preferably, when resources are sufficient, the task registration and delivery module adopts a balanced strategy to deliver the task to the acquisition terminal with the most sufficient resources at present; If there is currently no collection terminal with sufficient resources to match a task, the task registration and delivery module adopts an efficient space utilization strategy to enable as many medical devices as possible to be collected by the collection terminal.

[0015] Preferably, the task registration and delivery module uses the CPU and memory of the acquisition terminal as a two-dimensional vector, and uses the module length of the two-dimensional vector as the basis for selecting the acquisition terminal when delivering the task.

[0016] The centralized management method for medical equipment Internet of Things data acquisition terminals of the present application can centrally manage multiple acquisition terminals intuitively through the configuration center on the host computer and the corresponding operation pages. The acquisition terminals can be registered through the host computer, the status of the acquisition terminals can be monitored, the acquisition terminals can be restarted, and tasks can be dynamically assigned to the acquisition terminals according to the resource utilization of the acquisition terminals. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a network architecture diagram of the medical device Internet of Things involved in this application; Figure 2 It is the workflow diagram of the configuration center; Figure 3 This is a schematic diagram of the monitoring page of the configuration center; Figure 4 It is a schematic diagram of the collection terminal resource distribution on the details page of the monitoring page of the configuration center; Figure 5 This is a schematic diagram of the control interface for the configuration center to remotely restart the acquisition terminal; Figure 6 A schematic diagram of distributing configurations to the configuration center; Figure 7 A flowchart for configuring the center's task distribution; Figure 8 This is a schematic diagram of the algorithm used by the configuration center to allocate tasks using an efficient space utilization strategy; Fig. 9 Flowchart of the algorithm for the two-dimensional bin packing problem used by the configuration center. DETAILED DESCRIPTION

[0018] Below, the present application is described in detail with reference to the accompanying drawings.

[0019] like Figure 1 Shown is a network architecture diagram of the medical device Internet of Things involved in this application.

[0020] exist Figure 1 In the network architecture of medical device data collection, from bottom to top, they are: medical device data source layer, raw data access layer, data parsing and encapsulation layer, standardized data transmission layer, and big data persistence layer.

[0021] Among them, the host computer mainly realizes the reasonable allocation of acquisition terminal hardware resources and medical equipment data processing by controlling the mapping relationship of the "raw data access layer".

[0022] In addition, the rest of the functions of the upper computer are mainly concentrated in equipment maintenance. The main functions include: collecting program running errors, restarting the collection terminal, remote login, real-time traffic statistics, collecting terminal geographic location records, information registration responsible persons, fuzzy search, etc., forming a complete and centralized data collection full life cycle management system.

[0023] The operation process of the host computer software is as follows: Figure 2 As shown, after the administrator logs in, he can directly enter the monitoring homepage (such as Figure 3 As shown in the figure), the monitoring page can observe the running status of the acquisition terminal that has been registered to the host computer, and can also add unregistered acquisition terminals. The observation granularity is a single acquisition terminal.

[0024] After entering the details page, you can see more detailed operation information, such as Figure 4As shown in the figure, the observation granularity is a single task. The data on the monitoring page and its details page comes from the collection terminal, mainly including: error information (regular refresh), real-time resource usage (regular refresh), collection terminal operating system version number (online refresh), collection terminal hardware version number (online refresh), collection terminal software version number (online refresh), collection terminal network address (online refresh), etc.

[0025] In addition, the acquisition terminal can also be restarted to minimize the difficulty of maintenance. The principle is to send a restart command to the remote acquisition software, and the acquisition software calls the virtual container control system control command to trigger the restart. Figure 5 The restart operation page is shown.

[0026] The core function of the task registration and dispatching module is to dynamically dispatch the collection tasks according to the real-time resource usage of each collection terminal. In layman's terms, if the resource utilization rate fed back to the host computer by the collection terminal is sufficient, the task will be assigned to this collection terminal. Otherwise, the task will be deleted from this collection terminal and assigned to another collection terminal. There are two considerations when designing this function: 1. The utilization rate of the collection terminal's calculation and storage resources changes dynamically. For example, if the collection terminal is disconnected from the data center for a long time and accumulates data, a single collection terminal has insufficient storage space. After replacing the thread execution terminal, data can continue to be stored; 2. If a collection terminal goes down and multiple channel data cannot be processed suddenly, the task can be assigned to the remaining terminals for execution to avoid data loss.

[0027] When one or more configurations are generated, the management module distributes the configurations based on the computing storage resources of each terminal, involving a task distribution algorithm based on the computing storage resources of the acquisition terminal.

[0028] The algorithm idea is as follows: Since it is developed in Java, the garbage collection mechanism (GC) leads to higher requirements for hardware memory; Since the acquisition terminal is responsible for data collection, analysis, and packaging, "objects" will be created and destroyed at a high rate during the process, which is also a CPU-intensive program.

[0029] The algorithm distributes tasks based on real-time memory and CPU usage, such as Figure 7 shown.

[0030] When resources are sufficient, the goal is to give each task as many resources as possible, and adopt a balanced strategy to send tasks in sequence to the acquisition terminal with the most sufficient resources at the moment; If there is no collection terminal with sufficient resources to match a task, an efficient space utilization strategy is adopted, with the goal of collecting as many medical devices as possible. The algorithm is described as follows Figure 8 shown.

[0031] Residue vector: x = [ x 1 x 2 x 3 … x n ] , x represents the remaining amount of CPU and memory of each acquisition terminal. The memory of acquisition terminals with the same configuration still has some differences during operation.

[0032] Increment vector: y = [ y 1 y 2 y 3 … y m ] ,y represents the amount of CPU and memory required by the new medical equipment. y comes from the long-term feedback of the status monitoring module.

[0033] Sufficient margin, using the minimum resource occupancy priority algorithm to distribute tasks to each service to ensure the best performance; In the case of insufficient margin, the constrained two-dimensional packing problem is introduced into the task distribution algorithm to ensure that the number of connected devices is close to the maximum value under the constraints of computing and storage resources.

[0034] The algorithm flow chart is as follows Fig. 9 shown.

[0035] First, a near-optimal solution is obtained through the backtracking algorithm, and this near-optimal solution is genetically iterated. After a certain number of cycles, arrive The mapping tends to be stable, making the mapping result closer to the optimal solution.

[0036] In the backtracking algorithm, each time before inserting a task The sorting is based on an experience: it is always believed that the tasks inserted at a certain moment in the future will require a lot of resources, so the collection terminals with smaller resources are given priority.

[0037] It is a two-dimensional vector containing CPU and memory, and its modulus is used as the sorting basis. This sorting mechanism is reasonable when the terminal resources are collected>> the resources required for the task, and the experiment has verified it. In the actual application of the algorithm, the two vectors can be weighted according to whether the program execution tends to be CPU-intensive or memory-intensive, and then the weighted vector modulus is used for sorting.

[0038] A near-optimal solution is obtained through the backtracking algorithm. The time complexity of this algorithm is high, and the calculation time is in seconds, which is much lower than the frequency of resource adjustment, and the calculation time can be ignored.

[0039] This near-optimal solution is used as the input of the genetic algorithm. The basic idea is that a certain task that has been assigned is assigned to another collection terminal with idle resource resource1 and is available. If a slightly larger idle resource resource2 can be freed up, the direction of this task is changed. After several cycles, intuitively, as the number of iterations increases, the idle resources of each collection terminal will gradually be concentrated on one collection terminal.

[0040] When the spatial resources of each collection terminal no longer change, the genetic algorithm is considered to have converged.

[0041] The mapping results between the tasks and the acquisition terminals are output, and the host computer automatically redistributes the tasks.

[0042] Unless otherwise defined, all technical and / or scientific terms used in this application have the same meaning as commonly understood by one of ordinary skill in the art to which the invention relates. The materials, methods and examples mentioned in this application are illustrative only and not limiting.

[0043] Although the present invention has been described in conjunction with specific embodiments, within the spirit of the invention of this application, those skilled in the art may make appropriate substitutions, modifications and changes, and such substitutions, modifications and changes still fall within the scope of protection of this application.

Claims

1. A centralized management method for data collection terminals of the Internet of Things for medical devices, which manages multiple collection terminals respectively connected to at least one medical device through a configuration center located in a host computer; The configuration center includes the acquisition terminal status monitoring module and the task registration and delivery module; The acquisition terminal status monitoring module registers the acquisition terminal through the registration configuration page, and displays the operating status of the acquisition terminal that has been registered to the host computer through the monitoring page; The task registration and delivery module dynamically delivers collection tasks to each collection terminal according to the real-time resource usage status of each collection terminal.

2. The centralized management method for medical device Internet of Things data collection terminals according to claim 1 is characterized in that: On the monitoring page, you can further enter the details page, which displays the operation information in detail by task.

3. The centralized management method for medical device Internet of Things data collection terminals according to claim 1 is characterized in that: The data on the monitoring page and the details page are derived from the collection terminal, including: error information, real-time resource usage, collection terminal operating system version number, collection terminal hardware version number, collection terminal software version number, and at least one of the collection terminal network address.

4. The centralized management method for medical device Internet of Things data collection terminals according to claim 3 is characterized in that: Error information and real-time resource usage are refreshed regularly; The acquisition terminal operating system version number, acquisition terminal hardware version number, acquisition terminal software version number, and acquisition terminal network address are refreshed when the acquisition terminal goes online.

5. The centralized management method for medical device Internet of Things data collection terminals according to claim 1 is characterized in that: The terminal status monitoring module can restart the acquisition terminal.

6. The centralized management method for medical device Internet of Things data collection terminals according to claim 1 is characterized in that: If the collection terminal has sufficient resources to execute the task, the task registration and delivery module will assign the task to the collection terminal; If the collection terminal is short of resources when executing a task, the task registration and delivery module will delete the task from the collection terminal and assign it to another collection terminal.

7. The centralized management method for medical device Internet of Things data collection terminals according to claim 1 is characterized in that: When generating configuration, the task registration and delivery module distributes the configuration according to the resources of each acquisition terminal; the resources are computing resources and storage resources.

8. The centralized management method for medical device Internet of Things data collection terminals according to claim 7 is characterized in that: When resources are sufficient, the task registration and delivery module adopts a balanced strategy to deliver tasks to the acquisition terminal with the most sufficient resources. If there is currently no collection terminal with sufficient resources to match a task, the task registration and delivery module adopts an efficient space utilization strategy to enable as many medical devices as possible to be collected by the collection terminal.

9. The centralized management method for medical device Internet of Things data collection terminals according to claim 8 is characterized in that: The task registration and delivery module uses the CPU and memory of the acquisition terminal as a two-dimensional vector, and uses the module length of the two-dimensional vector as the basis for selecting the acquisition terminal when delivering the task.

Images