CN111180058B - An efficiency optimization method for automatic staining of urine cells - Google Patents

Abstract

An efficiency optimization algorithm for automatic dyeing of urine cells is characterized in that through research on an automatic dyeing scheduling process of urine cells, the effective utilization rate of a dye vat is maximized by distributing dye vat dye solutions, then different dyeing schemes are designed for the same dyeing method according to the distributed dye vat dye solutions, and finally a dyeing time registry is distributed and calculated for each dyeing scheme. The method can dye a plurality of tasks in the same dyeing mode, and remarkably improves the dyeing efficiency.

Description

A method for optimizing the efficiency of automatic staining of urine cells

Technical field:

The present invention relates to the fields of task scheduling, medical cell staining and servo control, and in particular to multi-task scheduling optimization technology.

Background technology:

In recent years, the number of diabetes patients in my country is increasing year by year. For the early diagnosis of diabetes, staining and analyzing urine cells in urine sediments is a very effective method. Traditional cell staining methods are usually performed manually, which is inefficient and has poor consistency. Some hospitals in China have also begun to gradually use automatic cell staining machines to replace traditional manual staining. Compared with manual staining, automatic cell staining machines have efficient and stable performance. Automatic staining equipment is simple to operate and can be stained in batches. It is particularly suitable for large hospitals in China, and the staining results are highly standardized. Using automatic cell staining machines instead of manual staining has become an important trend in the development of medical technology today.

The appearance of fully automatic cell staining machines in my country is relatively late, and the corresponding technical research and development investment is also inferior to many foreign manufacturers. Some well-known foreign companies such as Japan's Sakura (SAKURA), the United States' Johnson (Johnson), Germany's Siemens (SIMNES) and Leica (Leica), the Netherlands' Philips (PHILIPS) and other companies have long been involved in the field of medical device automation. These companies have quite mature technology in producing automation equipment, and the products they produce have a high degree of automation. At present, most of the fully automatic cell staining machines used in most hospitals in my country are imported products. However, these devices are generally expensive, with prices ranging from hundreds of thousands to millions. These devices generally have strict requirements on the operating environment, and when using these devices, they generally need to use bundled staining reagents, which are usually expensive and cause pollution and volatility with the number and duration of use, and need to be replaced frequently. At present, these devices are generally used by some large tertiary hospitals in China. For some local hospitals and third-party medical testing institutions, it is difficult to bear such huge expenses, and because the number of samples processed is not large, they still rely on traditional manual work to complete the corresponding experiments. Therefore, in recent years, some domestic hospitals and scientific research institutions have also begun to conduct relevant research on fully automatic dyeing machines. Compared with foreign fully automatic dyeing machines, the fully automatic dyeing machines developed by some domestic hospitals and scientific research institutions are mostly for specific purposes, and the dyeing efficiency is low. They generally use serial dyeing, that is, the next dyeing task is started only after one dyeing task is completely completed, which can better ensure the continuity of the dyeing steps of the same dyeing task. In order to optimize the staining efficiency, the literature 1 (L.S.Wang, Z.M.Yu, D.D.Zhang, Guofeng Qin, Zhenlei Xu, A Scheduling Algorithm for Urine Cell Dyeing Machine, ICCSE 2019, August 19-21, 2019.) proposed an accuracy-first staining scheduling algorithm (Accuracy-First Scheduling Algorithm, AFS). This algorithm is a dynamic scheduling algorithm, that is, the staining process has already started but a new staining slide has been added. The algorithm calculates the delay time for it to enter the slot for staining. The slide can be stained after waiting for the delay time, and it will not affect the previously placed slides. The algorithm has greatly improved the staining efficiency compared to the serial method, but the competition for the same staining resource (staining tank) in the staining process of this algorithm is obvious, that is, the same staining resource can only serve one staining task at the same time, especially multiple staining tasks of the same staining method compete for each staining resource.

Summary of the invention:

The object of the present invention is to provide a method for improving the parallel dyeing efficiency of the same dyeing method.

The idea of the present invention is mainly to allocate multiple copies of dyeing resources for which competition is most intense among parallel dyeing tasks, reduce competition among dyeing tasks, and thus improve dyeing efficiency. Biological cell dyeing can be compared to a computer task scheduling system, where each dyeing group is regarded as a different process and the dyeing tank is regarded as an external resource of the computer. Different dyeing groups will compete for the dyeing tank during the dyeing process. As shown in the schematic diagram of the dyeing task scheduling scenario in Figure 1, different dyeing tasks will compete for the dyeing tank, and the dyeing steps may not be exactly the same for different dyeing schemes. According to the dyeing process and the composition of the dyeing machine, we can find that the entire dyeing process must follow the following principles: the dyeing of the dyeing tank is allocated and fixed in advance, the resources are exclusive and cannot be deprived, the dyeing process of the same dyeing task is continuous, and the robot arm is exclusive.

Technical Solution

The method first allocates the dyeing tank dye liquid based on the EA36 Papanicolaou dyeing method to maximize the effective utilization rate, then designs different dyeing schemes for the dyeing method according to the allocated dyeing tank dye liquid, and finally allocates and calculates the dyeing time registry for each dyeing scheme, thereby determining the dyeing start time and end time for each dyeing step.

FIG2 is a flow chart of the dyeing efficiency optimization method, and the main steps of the method will be described below.

1) Initialization. This process completes the recording of each dyeing step and its dyeing time of the dyeing method, and records the number and distribution of dyeing tanks of the target dyeing machine.

2) Calculate the number of idle dyeing slots n _idle . This process mainly records the idle dyeing slots of the dyeing machine except the film taking, drying and washing slots.

3) Counting the dyeing time t ₁ to t _k . This process counts the dyeing time of each dyeing step, excludes the time data with the same time, and then sorts them from large to small and puts them in t ₁ to t _k .

个染槽数，其中1≤x<i；为染色时间未超过t_i的染色步骤则每个步骤分配1个染槽；然后计算总的独立染液数n_i。4) Calculate the number of independent dyeing solutions n _i when the unit dyeing time is _ti . The work completed in this step is actually to allocate multiple dyeing tanks for the dyeing steps with dyeing time exceeding _ti , that is, to allocate multiple dyeing tanks for the dyeing steps with dyeing time from t ₁ to _ti (excluding washing).

The number of dyeing tanks is 1≤x<i; for dyeing steps whose dyeing time does not exceed _ti , each step is assigned 1 dyeing tank; then the total number of independent dye solutions _ni is calculated.

5) Distribute the dye liquor according to the unit dyeing time _ti . In this process, the dye liquor in the dye tank is distributed sequentially according to the dyeing steps, and multiple dye liquors of the same dyeing step are distributed adjacently.

6) Design a dyeing scheme based on the dye tank and dye solution. Since the dyeing steps with long dyeing time have been separated into multiple dye tanks, when allocating dye tanks to parallel tasks, the steps with multiple dye tanks should be allocated in turn according to the task queue, so as to ensure that there is no competition between multiple tasks for the same dyeing.

7) Staining time registration algorithm. This algorithm uses the idea of backtracking to calculate the entry time of the slide. The time registration status of each staining slot is recorded and saved through the global time registry. The time registry is a two-dimensional array of int type, with a size of n x m. Among them, the row represents n staining slots, and the column represents that each staining slot can record m/2 time periods, and one time period occupies two element spaces (staining start time and staining end time). A staining slot table is set in the system, which stores the occupancy status of each staining slot. If the data in the staining slot table is all 0, it means that there is no slide in the staining slot. Figure 3 is a flow chart of the staining time registration algorithm.

By adopting the above scheme, the beneficial effects of the present invention are as follows:

The present invention studies the relationship between dyeing resources and dyeing tasks in the dyeing scheduling process, and reasonably allocates multiple dye tanks to several dyeing solutions with the most intense competition between dyeing steps of each dyeing task, thereby greatly improving the dyeing efficiency.

Explanation of the attached table:

Table 1 is the EA36 Papanicolaou staining method

Table 2 shows the distribution of dyeing tanks used in the dyeing machines

Table 3 is the initial distribution plan of dyeing bath dyeing liquid

Table 4 is the final distribution plan of dyeing bath and dyeing liquid

Table 5 is the final staining scheme and staining time registration table

Description of the drawings:

Figure 1 is a schematic diagram of the dyeing task scheduling scenario

Figure 2 is a flow chart of the dyeing scheduling efficiency optimization method

Figure 3 is a flow chart of the staining time registration algorithm

Figure 4 is the test result of EA36 Papanicolaou staining method

Specific implementation method:

Based on the existing dyeing device and EA36 Pap staining method, the specific embodiments of the present invention are described as follows:

Step (1) designs an initial distribution scheme of dyeing solution in the dyeing tank according to the dyeing tank distribution of the existing dyeing device (see Table 2) and the EA36 Papanicolaou dyeing method (see Table 3).

Step (2) solves the unit dyeing time as 60s based on the number of idle dyeing tanks and the time taken for the dyeing step.

In step (3), according to the unit staining time, it can be determined that 3 staining tanks need to be allocated for hematoxylin and orange solution, and 2 staining tanks need to be allocated for 95% alcohol in step 11.

Step (4) determines the final dye tank dye solution allocation plan according to the dye tank allocation method (see Table 4).

Step (5) The dyeing scheme is designed according to the dyeing steps and the final distribution scheme of the dyeing tank dye solution:

Q ₁ :C1->C2->C3->C19->C4->C20->C7->C8->C21->C3->C2->C9->C11->C12

->C14->C15->C18->C23->C24->C25

Q ₂ :C1->C2->C3->C19->C5->C20->C7->C8->C21->C3->C2->C10->C11->C1

2->C14->C16->C18->C23->C24->C25

Q ₃ :C1->C2->C3->C19->C6->C20->C7->C8->C21->C3->C2->C9->C11->C12

->C14->C17->C18->C23->C24->C25

Queue Q _3k+ 1 is the same as Q ₁ , queue Q _3k+2 is the same as Q ₂ , and queue Q _3k is the same as Q ₃ .

Step (6) can calculate the staining time registry according to the staining scheme (see Table 5).

For the EA36 dyeing method, the above test methods are performed according to the serial dyeing method, the AFS dyeing method, and the optimized method of the present invention, and the dyeing time and the number of dyeing task groups are in a positive linear positive correlation, as shown in Figure 4. Assuming that the dyeing time of the three test methods is t ₁ , t ₂ , and t ₃ , respectively, and the number of dyeing task groups is k, then for the three test methods, the following functions can be obtained respectively:

1) The relationship between the dyeing time and the number of dyeing task groups in the serial dyeing test method is shown in formula 1.

t ₁ =950×k (1)

2) Parallel dyeing test method The relationship between dyeing time and the number of dyeing task groups is shown in formula 2.

t ₂ =180×k+770 (2)

3) The relationship between the dyeing time and the number of dyeing task groups in the optimized parallel dyeing test method is shown in Formula 3.

t ₃ =60×h+890 (3)

According to Formula 1 and Formula 2, the dyeing efficiency improvement value v ₂₁ of the parallel dyeing method relative to the serial dyeing method can be obtained as shown in Formula 4.

According to Formula 4, when the number of dyeing groups k is infinite, the efficiency of the parallel dyeing method is expected to be improved by 5.2 times compared with the serial dyeing method.

According to Formula 1 and Formula 3, the dyeing efficiency improvement value v ₃₁ of the parallel dyeing method compared with the serial dyeing method after the scheduling algorithm is optimized can be obtained as shown in Formula 5.

According to Formula 5, when the number of dyeing groups k is infinite, the efficiency of the parallel dyeing method after scheduling algorithm optimization is expected to be improved by 15.8 times compared with the serial dyeing method.

According to Formula 2 and Formula 3, the improvement value v ₃₂ of the parallel dyeing efficiency after optimization compared with the parallel dyeing method before optimization can be obtained as shown in Formula 6.

According to Formula 6, it can be predicted that when the number of dyeing groups k is infinite, the efficiency of the parallel dyeing method after scheduling algorithm optimization is expected to be improved by 3 times compared with the parallel dyeing method before optimization.

Before optimization, the dyeing method used 19 dyeing tanks, so the utilization rate of the dyeing tanks before optimization was 19/26≈73%. After optimization, the dyeing method used 24 dyeing tanks, so the utilization rate of the dyeing tanks after optimization was increased to 24/26≈92%.

Table 1 EA36 Papanicolaou staining method

Table 2 Distribution of dyeing tanks in dyeing machines

Table 3 Initial distribution plan of dyeing bath dyeing liquid

Table 4 Final distribution plan of dyeing bath and dyeing liquid

Table 5 Final staining scheme and staining time registration table

Claims (1)

1. An efficiency optimization method for automatic staining of urine cells is characterized in that,

firstly, distributing dye vat dye liquor based on an EA36 Papanicolaou dyeing method, designing different dyeing schemes for the dyeing method according to the distributed dye vat dye liquor, and finally distributing and calculating a dyeing time registry for each dyeing scheme so as to determine the dyeing starting time and the dyeing ending time for each dyeing step; the method specifically comprises the following steps:

1) Initialization of

Recording the dyeing steps and the dyeing time of the EA36 Papanicolaou dyeing method, and recording the number and the distribution of dyeing tanks of a target dyeing machine;

2) Counting and calculating the number n of idle dyeing tanks _idle

Recording the idle dyeing tanks except for the slice taking, drying and washing tanks of the dyeing machine;

3) Counting the dyeing time t ₁ ～t _k

Counting the dyeing time of each dyeing step, excluding time data with the same time, and then placing the time data in t according to the sequence from big to small ₁ ～t _k In (a) and (b);

4) Calculating the dyeing time of the unit to be t _i Number of independent dye solutions n _i

Dyeing time exceeds t _i A plurality of dyeing tanks are allocated in the dyeing step of (1), and the washing step is not included, namely the dyeing time is t ₁ To t _i Is assigned to the dyeing step of (a)

The number of the dyeing tanks is equal to or less than 1 and x<i；

Dyeing time is not more than t _i Each step is allocated with 1 dye vat;

then the total independent dye liquor number n is calculated _i ；

5) According to the unit dyeing time t _i Dispensing dye liquor

Sequentially distributing the dye solutions of the dye tanks according to the dyeing steps, and adjacently distributing multiple dye solutions of the same dyeing step;

6) Designing a dyeing scheme according to the dye liquor of the dye vat

Because the dyeing step with long dyeing time is independent of a plurality of dyeing tanks, when the dyeing tanks are distributed for parallel tasks, the steps with the plurality of dyeing tanks are distributed in turn according to the task queues, so that the same dyeing is not competing among the plurality of tasks;

7) Dyeing time registration algorithm

Calculating the entry time of the slide by using the concept of a backtracking method;

recording and storing the time registration condition of each dye vat through a global time registry; the time registry is a two-dimensional array of int types, the size of which is n x m; wherein, the row represents n dyeing tanks, the column represents that each dyeing tank can record m/2 time periods, one time period occupies two element spaces, and the two element spaces are dyeing start time and dyeing end time;

setting a dye vat table in the system, wherein the table stores the occupation condition of each dye vat; if the staining bath table data are all 0, it indicates that there is no slide in the staining bath.

Images