A method and apparatus for adjusting parameters of a blood collection device

Abstract

The application discloses a blood collection equipment parameter adjustment method and device, and belongs to the field of blood collection. The method comprises the steps of generating initial collection parameters, acquiring blood stratification images, identifying platelet interface information, and dynamically adjusting collection parameters. The real-time thickness of the platelet interface is identified through a deep learning model, so that the platelet interface identification accuracy is effectively improved. The collection parameters are adjusted according to the real-time thickness of the platelet interface, the alarm occurrence probability in the collection process is reduced, the platelet product qualification rate and the collection efficiency are improved, and meanwhile, the platelet collection time can be reasonably shortened, and the blood donation experience is improved.

Description

Technical Field

[0001] This invention relates to the field of blood collection, and in particular to a method and apparatus for adjusting parameters of a blood collection device. Background Technology

[0002] Platelet apheresis devices can be classified according to platelet separation method, extracorporeal circulation mode, and consumable structure. Based on platelet separation method, they can be divided into blanched membrane method and platelet-rich plasma method; based on extracorporeal circulation mode, they can be divided into intermittent and continuous methods; and based on consumable structure, they can be divided into single-needle and double-needle methods. Currently, the main platelet apheresis devices available domestically and internationally include Terumo Bistrom's Trima Accel (USA), Blood Technologies' MCS+ (USA), Fresenius Kabi's Amicore, Amicus, and COM.TEC (Germany), and Nanger's NGL XCF 3000 (China). Among these, Terumo Bistrom's Trima Accel device uses the blanched membrane method, while the other devices use the platelet-rich plasma method.

[0003] During platelet collection, due to individual differences among donors, some may not adapt to the estimated collection parameters, potentially leading to frequent alarms and collection interruptions, thus affecting the final platelet product results. Conversely, some donors may adapt to faster collection parameters; in this case, consistently using the estimated parameters would prolong the overall collection time and reduce the donor experience. Therefore, dynamically adjusting collection parameters based on the actual collection situation is crucial.

[0004] In existing technologies, lasers are emitted from a light source towards the blood stratification points, and optical sensors receive the light signals after they pass through the blood. Based on the differences in the optical properties of different blood components, the separation interface information is obtained. Although optical detection methods can obtain separation interface information simply and efficiently, they cannot accurately quantify and assess platelet interface thickness. Furthermore, this method has weak anti-interference capabilities and is easily affected by other factors, which can lead to inaccurate interface recognition and consequently affect the accuracy of subsequent parameter adjustments. Summary of the Invention

[0005] In order to overcome the shortcomings of the prior art, one of the objectives of this invention is to provide a method for adjusting the parameters of a blood collection device that is highly resistant to interference and has high accuracy.

[0006] In order to overcome the shortcomings of the prior art, one of the objectives of this invention is to provide a blood collection equipment parameter adjustment device with strong anti-interference ability and high accuracy.

[0007] One of the objectives of this invention is achieved through the following technical solution:

[0008] A method for adjusting parameters of a blood collection device includes the following steps:

[0009] Generate initial collection parameters: Establish a blood donation database, which includes blood donor information and collection parameter information. When a blood donor donates blood, query the database to see if the blood donor's blood donation record exists. If it exists, use the historical collection parameters from the blood donation record and combine them with the donor's own information to calculate the initial collection parameters. If it does not exist, refer to the collection parameters of historical blood donors with similar blood donor information and combine them with the donor's own information to generate the initial collection parameters.

[0010] Obtaining blood stratification images: Obtaining images of the blood stratification interface within the centrifuge chamber;

[0011] Identifying platelet interface information: The blood layer interface image is input into a deep learning model. The deep learning model extracts feature information from the blood layer interface image, fuses the feature information to obtain multiple information fusion nodes, and passes the feature images in the multiple information fusion nodes as input to a region candidate network. The region candidate network is combined with the generator to calculate the final platelet region prediction box, thereby obtaining platelet interface information.

[0012] Dynamically adjust collection parameters: Based on real-time platelet interface information, when the platelet interface thickness is greater than the preset value, increase the plasma and / or platelet collection speed; when the platelet interface thickness is less than the preset value, increase the anticoagulant rate and blood collection speed; when the platelet interface thickness is within the preset range, maintain the current collection parameters.

[0013] Furthermore, the blood collection device parameter adjustment method also includes a blood collection speed adjustment step, which is located after the dynamic adjustment of the collection parameters. When the blood donor successfully collects blood for N minutes with the current collection parameters, the blood collection speed is increased and the collection parameters are updated; when the blood donor experiences P alarms within L minutes with the current collection parameters, the blood collection speed is decreased and the collection parameters are updated.

[0014] Furthermore, in the blood collection speed adjustment step, the blood collection speed is preset with an upper limit value and a lower limit value, the blood collection speed does not exceed the upper limit value, and the blood collection speed does not fall below the lower limit value.

[0015] Furthermore, in the step of generating initial collection parameters, the blood donor information includes multiple items such as name, gender, weight, height, ID number, platelet count, and hematocrit.

[0016] Furthermore, in the step of generating initial collection parameters, the collection parameter information includes plasma collection rate, platelet collection rate, blood collection rate, anticoagulant rate, and reinfusion rate.

[0017] Furthermore, in the step of acquiring blood layered images, a speed sensor is placed at the centrifuge shaft, and a strobe light source and a camera are respectively connected to the speed sensor. The camera and the strobe light source are triggered simultaneously by the speed signal, thereby achieving synchronous imaging and illumination at any centrifugation speed.

[0018] Furthermore, in the step of identifying platelet interface information, the deep learning model includes a feature extraction backbone network, which includes 5 sub-modules, and each sub-module includes 10 residual network calculations.

[0019] Furthermore, in the step of identifying platelet interface information, the feature information extracted from multiple layers in the feature extraction backbone network is fused with the feature information of the next layer after being upsampled twice and processed by the ESAM module to obtain an information fusion node.

[0020] Furthermore, the identification of platelet interface information also includes an image preprocessing step, which is performed before the blood layer interface image is input into the deep learning model. Specifically, the image preprocessing step involves preprocessing the original blood layer interface image to improve the image signal-to-noise ratio.

[0021] The second objective of this invention is achieved by the following technical solution:

[0022] A blood collection device parameter adjustment apparatus is provided for implementing any of the above-described blood collection device parameter adjustment methods. The blood collection device parameter adjustment apparatus includes a camera, a light source, a speed sensor, and a processor. The speed sensor is installed on the centrifuge shaft. The camera and the light source are both connected to the speed sensor. The speed signal of the speed sensor simultaneously triggers the camera and the light source, thereby achieving synchronous imaging and illumination at any centrifugal speed. The camera is communicatively connected to the processor. The processor recognizes the image acquired by the camera and adjusts the collection parameters of the blood collection device according to the recognition result.

[0023] Compared to existing technologies, the blood collection device parameter adjustment method of this invention, through steps such as generating initial collection parameters, acquiring blood layer images, identifying platelet interface information, and dynamically adjusting collection parameters, effectively improves the accuracy of platelet interface recognition by using a deep learning model to identify the real-time thickness of the platelet interface; adjusting collection parameters according to the real-time thickness of the platelet interface reduces the probability of alarms during collection, improves the platelet product qualification rate and collection efficiency; and can also reasonably shorten the platelet collection time, enhancing the blood donation experience. Attached Figure Description

[0024] Figure 1 This is a flowchart of the blood collection device parameter adjustment method of the present invention;

[0025] Figure 2 This is a schematic diagram of the network structure of a deep learning model;

[0026] Figure 3 This is a schematic diagram of the ESAM module structure;

[0027] Figure 4 This is a schematic diagram of the parameter adjustment device for a blood collection equipment.

[0028] In the diagram: 10, centrifuge plate; 20, rotating shaft; 30, speed sensor; 40, camera; 50, light source; 60, auxiliary light source. Detailed Implementation

[0029] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] It should be noted that when a component is said to be "fixed to" another component, it can be directly on the other component or fixed through another intermediate component. When a component is said to be "connected to" another component, it can be directly connected to the other component or may also have another intermediate component present. When a component is said to be "set to" another component, it can be directly set on the other component or may also have another intermediate component present.

[0031] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0032] Please see Figure 1 This application discloses a method for adjusting parameters of a blood collection device, comprising the following steps:

[0033] Generate initial collection parameters: Establish a blood donation database, which includes blood donor information and collection parameter information. When a blood donor donates blood, check if there is a blood donation record in the database. If there is, use the historical collection parameters on the blood donation record and combine them with the donor's own information to calculate the initial collection parameters. If there is no record, refer to the collection parameters of historical blood donors with similar information and combine them with the donor's own information to generate the initial collection parameters.

[0034] Obtaining blood stratification images: Obtaining images of the blood stratification interface within the centrifuge chamber;

[0035] Identifying platelet interface information: The blood layer interface image is input into the deep learning model. The deep learning model extracts feature information from the blood layer interface image, fuses the feature information to obtain multiple information fusion nodes, and passes the feature images in the multiple information fusion nodes as input to the region candidate network. The region candidate network and the generator are combined to calculate the final platelet region prediction box, thereby obtaining the platelet interface information.

[0036] Dynamically adjust collection parameters: Based on real-time platelet interface information, when the platelet interface thickness is greater than the preset value, increase the plasma and / or platelet collection speed; when the platelet interface thickness is less than the preset value, increase the anticoagulant rate and blood collection speed; when the platelet interface thickness is within the preset range, maintain the current collection parameters.

[0037] The specific steps for generating initial acquisition parameters are as follows:

[0038] Donor information includes several of the following: name, gender, weight, height, ID number, platelet count, and hematocrit. Name and ID number are used to identify the donor. Platelet count and hematocrit are used to generate collection parameters for donors with donation records. Gender, weight, height, platelet count, and hematocrit are also used to generate collection parameters for similar donors. Collection parameters include plasma collection rate, platelet collection rate, blood collection rate, anticoagulant rate, and reinfusion rate. Anticoagulant rate refers to the rate at which anticoagulant is drawn; reinfusion rate refers to the rate at which blood is reinfused to the donor; and blood collection rate refers to the rate at which blood is collected from the donor. Plasma collection rate and platelet collection rate refer to the rate at which plasma and platelets are collected after centrifugation in the centrifuge.

[0039] The specific steps for obtaining blood layered images are as follows:

[0040] Since the centrifugation speed during plasma and platelet collection can reach up to 3000 rpm, the exposure time must be minimized to reduce blurring and ensure good imaging results. Please continue reading. Figure 4 This application places a light source 50 and a camera 40 above the centrifuge tray 10. The light source 50 is a strobe light source; the camera 40 is a high-speed camera. A speed sensor 30 is placed at the centrifuge shaft 20. Both the light source 50 and the camera 40 are connected to the speed sensor 30. The speed signal simultaneously triggers the camera 40 and the light source 50, thereby achieving synchronous imaging and illumination at any centrifugation speed. An auxiliary light source 60 is placed below the centrifuge tray 10 to enhance the contrast of the blood stratification image.

[0041] The platelet interface information identification step is used to quickly and accurately identify platelet interface information from images, so as to dynamically adjust the acquired parameter data based on real-time image information.

[0042] This patent employs an improved Faster R-CNN deep learning neural network model to acquire platelet interface information. The improved Faster R-CNN deep learning neural network model is as follows: Figure 2 As shown, the specific process is as follows:

[0043] (1) Preprocess the original image of the blood layer interface to improve the image signal-to-noise ratio;

[0044] (2) Input the preprocessed image into the feature extraction backbone network region of the improved Faster R-CNN deep learning model to extract feature information from the image. The feature extraction backbone network consists of 5 sub-modules, and each sub-module contains 10 residual network calculations.

[0045] (3) Copy the feature information obtained from the last layer in step 2 to information node M5, and then perform a 2x upsampling and ESAM (Efficient Sharpness-Aware Minimization) module (e.g.) on the information in M5. Figure 3 The information in M4 is processed (as shown) and fused with the fourth layer feature information from step 2 to obtain information node M4. The information in M4 is then processed by doubling upsampling and the ESAM module, and fused with the third layer feature information from step 2 to obtain information node M3. The information in M3 is then processed by doubling upsampling and the ESAM module, and fused with the second layer feature information from step 2 to obtain information node M2. Next, the information in M2 is copied to information node P2. The information in M3 is fused with the information in P2 to obtain information node P3. The information in M4 is fused with the information in P3 to obtain information node P4. The information in M5 is fused with the information in P4 to obtain information node P5. The information in P5 is then downsampled by doubling to obtain information node P6.

[0046] (4) The information in nodes P2, P3, P4, P5 and P6 is passed as input to the region candidate network and combined with the generator to calculate the final platelet region prediction box, thereby obtaining the platelet interface information.

[0047] The specific steps for dynamically adjusting the acquisition parameters are as follows:

[0048] After obtaining platelet interface information, the relevant collection parameters are dynamically adjusted based on real-time interface information. If the platelet interface is detected to be too thick, indicating a high platelet concentration at the current separation interface, the plasma and platelet collection rates can be appropriately increased, thereby shortening the platelet collection time. If the platelet interface is detected to be too thin, indicating a low platelet concentration at the current separation interface, the blood collection rate and anticoagulant rate can be appropriately increased to help the platelet interface accumulate further. If the platelet interface is detected to be within the normal range, collection continues with the current collection parameters.

[0049] Simultaneously, the equipment's operating status is monitored in real time during platelet collection, and the collection parameters are dynamically adjusted based on the equipment's operating status. Therefore, the blood collection equipment parameter adjustment method also includes a blood collection speed adjustment step, which follows the dynamic adjustment of collection parameters. When a blood donor successfully collects blood for N minutes using the current collection parameters, the blood collection speed is increased, and the collection parameters are updated; when a blood donor experiences P alarms within L minutes using the current collection parameters, the blood collection speed is decreased, and the collection parameters are updated.

[0050] Specifically, if a donor has successfully collected blood for 5 minutes using the current collection parameters (meaning no alarms occurred during this time), indicating the donor is comfortable with the current parameters, the blood collection speed can be appropriately increased unless the program's maximum speed limit is reached, and other collection parameters are updated simultaneously. If the donor experiences three consecutive alarms within 3 minutes using the current parameters, indicating multiple interruptions to the collection process and potentially low platelet concentration, the blood collection speed can be appropriately reduced unless the program's minimum speed limit is reached, and other collection parameters are updated simultaneously to maintain continuity of blood donation. If neither of these conditions is met, collection continues using the current parameters.

[0051] After platelet collection is completed, the collection parameters and data will be updated in the blood donation database for future reference.

[0052] Please continue reading. Figure 4 This application also discloses a blood collection device parameter adjustment device for implementing the above-mentioned blood collection device parameter adjustment method. The blood collection device parameter adjustment device includes a camera 40, a light source 50, a speed sensor 30, and a processor. The speed sensor 30 is installed on the centrifuge shaft 20. The camera 40 and the light source 50 are both connected to the speed sensor 30. The speed signal of the speed sensor 30 simultaneously triggers the camera 40 and the light source 50, thereby achieving synchronous imaging and illumination at any centrifugation speed. The camera 40 is communicatively connected to the processor. The processor recognizes the image acquired by the camera 40 and adjusts the collection parameters of the blood collection device according to the recognition result.

[0053] Compared to existing technologies, the blood collection device parameter adjustment method of this invention, through steps such as generating initial collection parameters, acquiring blood layer images, identifying platelet interface information, and dynamically adjusting collection parameters, effectively improves the accuracy of platelet interface recognition by using a deep learning model to identify the real-time thickness of the platelet interface; adjusting collection parameters according to the real-time thickness of the platelet interface reduces the probability of alarms during collection, improves the platelet product qualification rate and collection efficiency; and can also reasonably shorten the platelet collection time, enhancing the blood donation experience.

[0054] The above embodiments merely illustrate several implementation methods of the present invention, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the invention patent. It should be noted that, for those skilled in the art, several modifications and improvements can be made without departing from the concept of the present invention. These are all equivalent modifications and improvements made to the above embodiments based on the essential technology of the present invention, and all of these fall within the protection scope of the present invention.

Claims

1. A method for adjusting parameters of a blood collection device, characterized in that, Includes the following steps: Generate initial collection parameters: Establish a blood donation database, which includes blood donor information and collection parameter information. When a blood donor donates blood, query the database to see if the blood donation record of the donor exists. If it exists, use the historical collection parameters on the blood donation record and combine them with the donor's own information to calculate the initial collection parameters. If the donor does not exist, the initial collection parameters are generated by referring to the collection parameters of historical blood donors with similar information and combining them with the donor's own information. Obtaining blood stratification images: Obtaining images of the blood stratification interface within the centrifuge chamber; Identifying platelet interface information: The blood layer interface image is input into a deep learning model. The deep learning model extracts feature information from the blood layer interface image, fuses the feature information to obtain multiple information fusion nodes, and passes the feature images in the multiple information fusion nodes as input to a region candidate network. The region candidate network is combined with the generator to calculate the final platelet region prediction box, thereby obtaining platelet interface information. Dynamically adjust collection parameters: Based on real-time platelet interface information, when the platelet interface thickness is greater than the preset value, increase the plasma and / or platelet collection speed; when the platelet interface thickness is less than the preset value, increase the anticoagulant rate and blood collection speed; when the platelet interface thickness is within the preset range, maintain the current collection parameters.

2. The method for adjusting parameters of a blood collection device according to claim 1, characterized in that: The blood collection device parameter adjustment method further includes a blood collection speed adjustment step, which is located after the dynamic adjustment of the collection parameters. When the blood donor successfully collects blood for N minutes with the current collection parameters, the blood collection speed is increased and the collection parameters are updated. When the blood donor experiences P alarms within L minutes with the current collection parameters, the blood collection speed is decreased and the collection parameters are updated.

3. The method for adjusting parameters of a blood collection device according to claim 2, characterized in that: In the blood collection speed adjustment step, the blood collection speed is preset with an upper limit and a lower limit. The blood collection speed shall not exceed the upper limit and shall not be lower than the lower limit.

4. The method for adjusting parameters of a blood collection device according to claim 1, characterized in that: In the step of generating initial collection parameters, the blood donor information includes multiple items such as name, gender, weight, height, ID number, platelet count, and hematocrit.

5. The method for adjusting parameters of a blood collection device according to claim 1, characterized in that: In the step of generating initial collection parameters, the collection parameter information includes plasma collection rate, platelet collection rate, blood collection rate, anticoagulant rate, and reinfusion rate.

6. The method for adjusting parameters of a blood collection device according to claim 1, characterized in that: In the step of acquiring blood layered images, a speed sensor is placed at the centrifuge shaft, and a stroboscopic light source and a camera are respectively connected to the speed sensor. The camera and the stroboscopic light source are triggered simultaneously by the speed signal, thereby achieving synchronous imaging and illumination at any centrifugation speed.

7. The method for adjusting parameters of a blood collection device according to claim 1, characterized in that: In the step of identifying platelet interface information, the deep learning model includes a feature extraction backbone network, which includes 5 sub-modules, and each sub-module includes 10 residual network calculations.

8. The method for adjusting parameters of a blood collection device according to claim 7, characterized in that: In the step of identifying platelet interface information, the feature information extracted from multiple layers in the feature extraction backbone network is fused with the feature information of the next layer after being upsampled twice and processed by the ESAM module to obtain an information fusion node.

9. The method for adjusting parameters of a blood collection device according to claim 1, characterized in that: The process of identifying platelet interface information also includes an image preprocessing step, which is performed before inputting the blood layer interface image into the deep learning model. Specifically, the image preprocessing step involves preprocessing the original blood layer interface image to improve the image signal-to-noise ratio.

10. A parameter adjustment device for a blood collection device, used to implement the parameter adjustment method for a blood collection device as described in any one of claims 1-9, characterized in that: The blood collection device parameter adjustment mechanism includes a camera, a light source, a speed sensor, and a processor. The speed sensor is installed on the centrifuge shaft. The camera and the light source are both connected to the speed sensor. The speed signal of the speed sensor simultaneously triggers the camera and the light source, thereby achieving synchronous imaging and illumination at any centrifugation speed. The camera is communicatively connected to the processor. The processor recognizes the image acquired by the camera and adjusts the collection parameters of the blood collection device according to the recognition result.

Images