Home non-invasive anemia monitoring and early warning method and system in chronic kidney disease scenarios

Abstract

The present application relates to the field of biomedical engineering technology, and in particular to a method and system for non-invasive home anemia monitoring and early warning in the context of chronic kidney disease. The system includes a hemoglobin detection device, a dialysis machine, and a processing device. The physiological data of chronic kidney patients are collected by the hemoglobin detection device, the chronic kidney patients are dialyzed and the dialysis data are recorded by the dialysis machine, and the predicted anemia risk level of the chronic kidney patients is determined by the processing device based on the physiological data, dialysis data, and vital signs of the chronic kidney patients, and anemia early warning is performed based on the predicted anemia risk level and physiological data. The present application realizes home monitoring of anemia in chronic kidney patients without the need for them to travel back and forth to the hospital regularly, which reduces the economic and physical burden of chronic kidney patients and significantly improves the quality of life of chronic kidney patients.

Description

Technical Field

[0001] The present application relates to the field of biomedical engineering technology, and in particular to a method and system for non-invasive home anemia monitoring and early warning in the context of chronic kidney disease. Background Art

[0002] As the number of chronic kidney disease (CKD) patients continues to grow, home dialysis has become an important treatment option for end-stage renal disease (ESRD) patients due to its flexibility and respect for patients' autonomy. To monitor the status of home dialysis patients, home dialysis monitoring systems are installed. These systems monitor physiological parameters (such as blood pressure, temperature, respiration, and blood sugar) in real time, providing a real-time understanding of the patient's physical condition.

[0003] However, patients with chronic kidney disease often suffer from renal anemia, which is primarily caused by insufficient renal erythropoietin (EPO) secretion, impaired iron metabolism, and chronic inflammation, leading to impaired hemoglobin (Hb) synthesis. As renal function declines, the incidence of renal anemia gradually increases and the severity of anemia gradually worsens. However, existing home dialysis monitoring systems are unable to monitor anemia in home dialysis patients, forcing them to visit the hospital for regular blood tests, which not only increases the financial burden on patients, but also increases their physical and psychological burden.

[0004] Therefore existing technology still needs to be improved and improved. Summary of the Invention

[0005] The technical problem to be solved by this application is to provide a home non-invasive anemia monitoring and early warning method and system in the context of chronic kidney disease in response to the shortcomings of the existing technology.

[0006] In order to solve the above technical problems, the first aspect of the present application provides a home non-invasive anemia monitoring and early warning system in the chronic kidney disease scenario, wherein the home non-invasive anemia monitoring and early warning system in the chronic kidney disease scenario specifically includes a hemoglobin detection device, a dialysis machine, and a processing device;

[0007] The hemoglobin detection device is used to collect physiological data of chronic kidney disease patients, wherein the physiological data includes hemoglobin data and perfusion index data;

[0008] The dialysis machine is used to perform dialysis on chronic kidney patients and record dialysis data;

[0009] The processing device is used to determine the predicted anemia risk level of the chronic kidney patient based on the physiological data, the dialysis data and the vital sign data of the chronic kidney patient, and to provide anemia warning based on the predicted anemia risk level and the physiological data.

[0010] The non-invasive home anemia monitoring and early warning system for chronic kidney disease, wherein the step of determining the predicted anemia risk level of a chronic kidney patient based on the physiological data, the dialysis data, and the vital signs data of the chronic kidney patient specifically includes:

[0011] Extracting features from the physiological data using a pre-trained feature extraction model to obtain hemoglobin time series features and perfusion index fluctuation features;

[0012] The hemoglobin time series characteristics, the perfusion index fluctuation characteristics, the dialysis data, and the physical sign data are input into a pre-trained anemia prediction model, and the predicted anemia risk level of the chronic kidney disease patient is determined by the anemia prediction model.

[0013] In the home non-invasive anemia monitoring and early warning system for chronic kidney disease scenarios, the feature extraction of the physiological data using a pre-trained feature extraction model to obtain hemoglobin time series features and perfusion index fluctuation features specifically includes:

[0014] generating initial hemoglobin baseline data based on the hemoglobin data;

[0015] Dynamically adjusting the initial hemoglobin baseline data according to the perfusion index data to obtain hemoglobin baseline data;

[0016] The hemoglobin baseline data and the perfusion index data are input into a pre-trained feature extraction model, and the pre-trained feature extraction model outputs the hemoglobin time series feature and the perfusion index fluctuation feature.

[0017] The home non-invasive anemia monitoring and early warning system for chronic kidney disease scenarios, wherein, before determining the predicted anemia risk level of the chronic kidney patient based on the physiological data, the dialysis data, and the vital sign data of the chronic kidney patient, further comprises:

[0018] Acquiring altitude data of the chronic kidney disease patient;

[0019] The hemoglobin and the perfusion index are corrected based on the altitude data.

[0020] The home non-invasive anemia monitoring and early warning system for chronic kidney disease scenarios, wherein, after correcting the hemoglobin and the perfusion index according to the altitude data, further comprises:

[0021] comparing the perfusion index with a preset perfusion index threshold;

[0022] If the perfusion index is less than the perfusion index threshold, the perfusion index is manually confirmed. If the confirmation is passed, the perfusion index is retained; if the confirmation is not passed, the perfusion index is discarded.

[0023] The home non-invasive anemia monitoring and early warning system for chronic kidney disease scenarios, wherein the anemia early warning based on the predicted anemia risk level and the physiological data specifically includes:

[0024] Matching the hemoglobin data, perfusion index data, and predicted anemia risk level with a preset warning level condition library, wherein the preset warning level condition library includes first-level warning conditions, second-level warning conditions, and third-level warning conditions;

[0025] When the hemoglobin data, perfusion index data, and predicted anemia risk level meet the first-level warning conditions, a first-level warning is triggered to generate a chronic kidney disease patient prompt message;

[0026] When the hemoglobin data, perfusion index data and predicted anemia risk level meet the secondary warning conditions, a secondary warning is triggered to generate a prompt message for the chronic kidney disease patient's medical and / or family members;

[0027] When the hemoglobin data, perfusion index data, and predicted anemia risk level meet the third-level warning conditions, a third-level warning is triggered to automatically locate and activate the emergency channel;

[0028] Among them, the first-level warning condition is that the hemoglobin at the current time is less than the first preset threshold and greater than the second preset threshold; the first-level warning condition is that the predicted anemia risk level is greater than or equal to the moderate anemia risk level, and when the predicted anemia risk level is equal to the moderate anemia risk level, the predicted probability of the predicted anemia risk level is greater than the preset probability threshold; the third-level warning condition is that the hemoglobin at the current time is less than the second preset threshold or the perfusion index at the current time is less than the preset perfusion index threshold.

[0029] The home non-invasive anemia monitoring and early warning system in the chronic kidney disease scenario, wherein the processing device is also used to generate a hemoglobin-ultrafiltration volume correlation heat map based on physiological data, generate an anemia risk trend curve based on the predicted anemia risk level and several historical predicted anemia risk levels, and / or generate a correlation map between dialysis efficacy and anemia risk based on the predicted anemia risk level, so as to demonstrate the impact of dialysis on hemoglobin levels.

[0030] The non-invasive home anemia monitoring and early warning system for chronic kidney disease scenarios, wherein the non-invasive home anemia monitoring and early warning system for chronic kidney disease scenarios further includes:

[0031] The cloud device is used to synchronize physiological data, dialysis data and predicted anemia risk level, and push the predicted anemia risk level to medical staff and / or family members.

[0032] A second aspect of the present application provides a method for non-invasive home anemia monitoring and early warning in the context of chronic kidney disease, which utilizes the above-mentioned non-invasive home anemia monitoring and early warning system in the context of chronic kidney disease. The method specifically includes:

[0033] Collecting physiological data of a chronic kidney disease patient, wherein the physiological data includes hemoglobin data and perfusion index data;

[0034] Perform dialysis on chronic kidney disease patients and record dialysis data;

[0035] The predicted anemia risk level of the chronic kidney patient is determined based on the physiological data, the dialysis data and the vital sign data of the chronic kidney patient, and anemia warning is performed based on the predicted anemia risk level and the physiological data.

[0036] The method for non-invasive home anemia monitoring and early warning in the context of chronic kidney disease, wherein the method further comprises:

[0037] The physiological data, dialysis data, and predicted anemia risk level are synchronized to a cloud device, so that the predicted anemia risk level is pushed to a medical care end and / or a family member end via the cloud device.

[0038] Beneficial effects: Compared with the prior art, the embodiment of the present application provides a method and system for non-invasive home anemia monitoring and early warning in the chronic kidney disease scenario, the system includes a hemoglobin detection device, a dialysis machine and a processing device, the physiological data of the chronic kidney patient is collected by the hemoglobin detection device, the chronic kidney patient is dialyzed and the dialysis data is obtained by the dialysis machine, the predicted anemia risk level of the chronic kidney patient is determined by the processing device based on the physiological data, the dialysis data and the vital signs data of the chronic kidney patient, and an anemia early warning is performed based on the predicted anemia risk level and the physiological data. The embodiment of the present application collects the physiological data of the chronic kidney patient by the hemoglobin detection device, and performs anemia risk level prediction and anemia early warning by the processing device, thereby realizing home monitoring of anemia in chronic kidney patients, eliminating the need for them to travel back and forth to the hospital regularly, reducing the economic and physical burden of chronic kidney patients, and significantly improving the quality of life of chronic kidney patients. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0040] Figure 1 This is a functional block diagram of the home non-invasive anemia monitoring and early warning system for chronic kidney disease scenarios provided in an embodiment of the present application.

[0041] Figure 2 This is a functional block diagram of an embodiment of a home non-invasive anemia monitoring and early warning system for chronic kidney disease scenarios provided in an embodiment of the present application.

[0042] Figure 3 The figure shows an example flow chart of the process of obtaining the predicted anemia risk level.

[0043] Figure 4 Flowchart showing the principle of the process for obtaining the predicted anemia risk level. DETAILED DESCRIPTION

[0044] The present application provides a method and system for non-invasive home anemia monitoring and early warning in the context of chronic kidney disease. To make the purpose, technical solutions, and effects of this application more clear and explicit, the present application is further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are intended only to explain this application and are not intended to limit this application.

[0045] It will be understood by those skilled in the art that, unless expressly stated otherwise, the singular forms "a", "an", "said" and "the" used herein may also include the plural forms. It should be further understood that the term "comprising" used in the specification of the present application refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof. It should be understood that when we refer to an element as being "connected" or "coupled" to another element, it may be directly connected or coupled to the other element, or there may be intermediate elements. In addition, "connected" or "coupled" as used herein may include wireless connections or wireless couplings. The term "and / or" used herein includes all or any units and all combinations of one or more associated listed items.

[0046] It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those skilled in the art to which this application belongs. It should also be understood that terms such as those defined in common dictionaries should be understood to have meanings consistent with their meanings in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless specifically defined as herein.

[0047] It should be understood that the sequence numbers and sizes of the steps in this embodiment do not imply the order of execution. The order of execution of each process is determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of this application.

[0048] Research has found that as the number of chronic kidney disease (CKD) patients continues to expand, home dialysis has become an important treatment option for end-stage renal disease (ESRD) patients due to its flexibility and respect for patient autonomy. However, renal anemia, a core complication of CKD, is particularly prevalent in CKD patients. This condition is primarily caused by insufficient renal erythropoietin (EPO) secretion, impaired iron metabolism, and chronic inflammation, leading to impaired hemoglobin (Hb) synthesis. As renal function declines, the incidence and severity of anemia gradually increase.

[0049] Studies have found that anemia is an independent risk factor for cardiovascular events in CKD patients. For every 10g / L increase in Hb, the relative risk of cardiovascular events decreases by approximately 17%. Therefore, effective home anemia monitoring and early warning in the setting of chronic kidney disease is crucial for reducing the incidence of cardiovascular events and improving the quality of life and safety of dialysis patients. However, existing home dialysis monitoring systems generally only monitor basic physiological parameters (such as blood pressure, temperature, respiration, and blood sugar) in real time to provide a real-time understanding of the patient's physical condition. However, they are unable to monitor anemia in home dialysis patients, forcing them to visit the hospital regularly for blood tests. This increases the financial burden on patients, as well as their physical and psychological burden.

[0050] In order to solve the above problems, the embodiment of the present application provides a home non-invasive anemia monitoring and early warning system in the chronic kidney disease scenario, the system includes a hemoglobin detection device, a dialysis machine and a processing device, the hemoglobin detection device is used to collect the physiological data of the chronic kidney patient, the dialysis machine is used to dialyze the chronic kidney patient and record the dialysis data, the processing device is used to determine the predicted anemia risk level of the chronic kidney patient based on the physiological data, the dialysis data and the vital signs data of the chronic kidney patient, and an anemia early warning is performed based on the predicted anemia risk level and the physiological data. The embodiment of the present application collects the physiological data of the chronic kidney patient through the hemoglobin detection device, records the dialysis data through the dialysis machine, and then performs anemia risk level prediction and anemia early warning through the processing device, thereby realizing home monitoring of anemia in chronic kidney patients, eliminating the need for them to travel back and forth to the hospital regularly, reducing the economic and physical burden of chronic kidney patients, and significantly improving the quality of life of chronic kidney patients.

[0051] The application content will be further explained below through description of embodiments in conjunction with the accompanying drawings.

[0052] This embodiment provides a home non-invasive anemia monitoring and early warning system for chronic kidney disease. Figure 1 As shown, the home non-invasive anemia monitoring and early warning system for chronic kidney disease scenarios includes a hemoglobin detection device 100, a dialysis machine 200, and a processing device 300. Both the hemoglobin detection device 100 and the dialysis machine 200 communicate with the processing device 300. The hemoglobin detection device 100 is used to collect physiological data from chronic kidney patients and transmit the collected physiological data to the processing device 300. The dialysis machine 200 is used to perform dialysis (such as home peritoneal dialysis) on chronic kidney patients and transmit the recorded dialysis data to the processing device 300. The processing device 300 is used to determine the chronic kidney patient's predicted anemia risk level based on the physiological data, dialysis data, and the chronic kidney patient's vital signs, and to provide anemia early warning based on the predicted anemia risk level and physiological data. The embodiments of the present application realize non-invasive home anemia monitoring and early warning in the chronic kidney disease scenario through the mutual cooperation of hemoglobin detection equipment, dialysis machines and processing equipment. Chronic kidney patients do not need to travel back and forth to the hospital regularly, which reduces the economic and physical burden of chronic kidney disease patients at home and significantly improves the quality of life of chronic kidney disease patients.

[0053] Specifically, the hemoglobin detection device 100 adopts a non-invasive hemoglobin detection device, and the hemoglobin, blood oxygen saturation, pulse rate and perfusion index of a chronic kidney disease patient can be measured through the hemoglobin detection device to obtain the physiological data of the chronic kidney disease patient, and the collected physiological data can be synchronously transmitted to the processing device. In addition, the hemoglobin detection device can adopt a lightweight wearable form, which is convenient for chronic kidney disease patients to achieve measurement anytime and anywhere, and does not interfere with their dialysis. Among them, the hemoglobin detection device can collect the physiological data of the chronic kidney disease patient in real time, or collect the physiological data of the chronic kidney disease patient once at a preset time interval, or collect the physiological data of the chronic kidney disease patient according to the collection instruction, etc. In the embodiment of the present application, since the chronic kidney disease patient needs to undergo dialysis every day, the hemoglobin detection device measures the physiological data of the chronic kidney disease patient once a day at a fixed time, and uploads the measured physiological data to the processing device every day.

[0054] The dialysis machine 200 is responsible for completing the core dialysis operations and recording real-time dialysis data such as dialysate residence time, ultrafiltration volume, urea clearance index (Kt / v), and urea clearance rate. The dialysis machine automatically uploads dialysis data to a processing device via its built-in IoT communication module. The dialysis machine also receives interventions from the processing device to ensure timely intervention. Furthermore, the dialysis machine can automatically pause treatment if an abnormality occurs (such as a blockage in the line or an abnormal ultrafiltration volume), and provide a voice prompt for the patient to check the dialysis machine.

[0055] The processing device 300 is used to predict anemia risk levels and provide anemia warnings. It includes at least one processor and memory. The logic instructions in the memory can be implemented as software functional units and, when sold or used as a standalone product, can be stored in a computer-readable storage medium. The memory, as a computer-readable storage medium, can be configured to store software programs or computer-executable programs, such as program instructions or modules corresponding to the methods in the embodiments of the present disclosure. The processor executes the software programs, instructions, or modules stored in the memory to perform functional applications and data processing, thereby implementing the methods in the embodiments described above.

[0056] The processing device determines the user's predicted anemia risk level based on the physiological data, dialysis data, and vital signs data, and issues an anemia warning based on the predicted anemia risk level and the physiological data. Physiological data is synchronized with the hemoglobin detection device and includes hemoglobin data and perfusion index data. Hemoglobin (Hb) is a core direct parameter for anemia, and its concentration is significantly negatively correlated with the severity of anemia. The perfusion index (PI) measures the intensity of peripheral microcirculatory blood flow, with a normal value range of 1.0-10.0, and can assist in determining the signal quality of hemoglobin. Dialysis data is synchronized from the dialysis machine to the processing device and may include dialysate residence time, ultrafiltration volume, and toxin clearance rate. The predicted anemia risk level is predicted based on the physiological data, dialysis data, and vital signs data, and is used to reflect the anemia status of chronic kidney disease patients.

[0057] In one embodiment, Figure 3 and Figure 4 As shown, the method of determining the predicted anemia risk level of a chronic kidney patient based on the physiological data, the dialysis data, and the physical sign data of the chronic kidney patient specifically includes:

[0058] H10. Extract features from the physiological data using a pre-trained feature extraction model to obtain hemoglobin time series features and perfusion index fluctuation features;

[0059] H20. Input the hemoglobin time series characteristics, the perfusion index fluctuation characteristics, the dialysis data, and the physical sign data into a pre-trained anemia prediction model, and determine the predicted anemia risk level of the chronic kidney disease patient through the anemia prediction model.

[0060] Specifically, in step H10, Figure 4 As shown, the feature extraction model and the anemia prediction model are configured to form a prediction model. The input data of the prediction model includes hemoglobin data, perfusion index dialysis data, and the vital sign data. The feature extraction model is used to extract features from the hemoglobin data and perfusion index in the physiological data to obtain hemoglobin time series features and perfusion index fluctuation features. The feature extraction model can use a bidirectional LSTM or Bi-LSTM neural network to extract hemoglobin time series features from the hemoglobin data and perfusion index fluctuation features from the perfusion index data.

[0061] Furthermore, the physiological data, dialysis data, and vital sign data can be data acquired at the time of prediction or data acquired during a preset period of time before the prediction time. In this embodiment of the present application, the physiological data, dialysis data, and vital sign data all include all data acquired during the preset period of time before the prediction time. For example, if the preset period is seven days and physiological data, dialysis data, and vital sign data are acquired daily, then the physiological data include the seven physiological data acquired closest to the prediction time, the dialysis data include the seven dialysis data acquired closest to the prediction time, and the vital sign data include the seven vital sign data acquired closest to the prediction time. To this end, before predicting the anemia risk level based on the physiological data, dialysis data, and vital sign data, they must first be aligned according to the acquisition time to ensure their temporal sequence. The vital sign data can be acquired using a hemoglobin test device or other vital sign data acquisition device and may include age, EPO (erythropoietin) dosage, gender, and the like.

[0062] In one embodiment, after aligning the physiological data, dialysis data, and vital sign data, the physiological data may be preprocessed to improve the accuracy of the physiological data. Accordingly, before determining the predicted anemia risk level of the chronic kidney disease patient based on the physiological data, the dialysis data, and the vital sign data of the chronic kidney disease patient, the method further includes:

[0063] Acquiring altitude data of the chronic kidney disease patient;

[0064] The hemoglobin and the perfusion index are corrected based on the altitude data.

[0065] Specifically, the altitude data refers to the altitude at which the chronic kidney disease patient is located, which may be pre-stored in the processing device. That is, the processing device will pre-acquire the altitude at which the chronic kidney disease patient is located and store it. At the same time, when the altitude at which the chronic kidney disease patient is located changes, the altitude data stored in the processing device will also be updated accordingly. After acquiring the altitude data, the hemoglobin and the perfusion index are corrected according to the altitude data, wherein the correction may adopt a preset correspondence between the altitude severity and the adjustment threshold to determine the correction threshold corresponding to the altitude data, and then correct the hemoglobin and the perfusion index according to the correction threshold; or it may be a preset correction formula, and then correct the hemoglobin and the perfusion index according to the altitude data and the correction formula, etc.

[0066] Furthermore, in practical applications, the perfusion index in the physiological data may contain abnormalities. To prevent the impact of abnormalities on prediction accuracy, the perfusion index may be confirmed after verification (i.e., an abnormality check). Accordingly, after correcting the hemoglobin and the perfusion index according to the altitude data, the following steps may also be performed:

[0067] comparing the perfusion index with a preset perfusion index threshold;

[0068] If the perfusion index is less than the perfusion index threshold, the perfusion index is manually confirmed. If the confirmation is passed, the perfusion index is retained; if the confirmation is not passed, the perfusion index is discarded.

[0069] Specifically, a perfusion index threshold is pre-set and serves as a basis for determining whether a perfusion index is abnormal. For example, the pre-set perfusion index threshold is 0.3. When the perfusion index is less than the perfusion index threshold, the perfusion index is determined to be abnormal and displayed to a pre-set confirmation personnel for manual confirmation. If the manual confirmation passes, the perfusion index is retained. If the manual confirmation fails, the perfusion index is deemed invalid and discarded. Conversely, when the perfusion index is greater than or equal to the perfusion index threshold, the perfusion index is determined to be normal and retained without manual confirmation.

[0070] In one embodiment, extracting features from the physiological data using a pre-trained feature extraction model to obtain hemoglobin time series features and perfusion index fluctuation features specifically includes:

[0071] generating initial hemoglobin baseline data based on the hemoglobin data;

[0072] Dynamically adjusting the initial hemoglobin baseline data according to the perfusion index data to obtain hemoglobin baseline data;

[0073] The hemoglobin baseline data and the perfusion index data are input into a pre-trained feature extraction model, and the pre-trained feature extraction model outputs the hemoglobin time series feature and the perfusion index fluctuation feature.

[0074] Specifically, the initial hemoglobin baseline data is calculated based on the hemoglobin data, for example, the initial hemoglobin baseline data is calculated based on hemoglobin data for seven consecutive days. After obtaining the initial hemoglobin baseline data, a smoothing filter (e.g., a moving average method) can be applied to the initial hemoglobin baseline data to reduce the initial hemoglobin baseline data.

[0075] After obtaining the initial hemoglobin baseline data, the perfusion index data is used to dynamically adjust the initial hemoglobin baseline data to obtain the hemoglobin baseline data, so that the hemoglobin baseline data can more accurately reflect the actual situation of the chronic kidney disease patient. The adjustment formula for dynamically adjusting the initial hemoglobin baseline data using the perfusion index data can be:

[0076] baseline(t)=Smooth(Hb(t))×Adjust(PI(t));

[0077] Wherein, baseline(t) represents the hemoglobin baseline data, Hb(t) represents the initial hemoglobin baseline data, Smooth() represents the smoothing operation, Adjust() represents determining the adjustment value according to the perfusion index data, and PI(t) represents the perfusion index data.

[0078] It should be noted that the adjustment value corresponding to the perfusion index data can be determined using an existing method, for example, according to a preset correspondence between the perfusion index data and the adjustment value.

[0079] After obtaining the hemoglobin baseline data, the feature extraction model is used to extract features from the hemoglobin baseline data and the perfusion index data, respectively, to obtain hemoglobin time series features and perfusion index fluctuation features. The hemoglobin time series features and perfusion index fluctuation features are both used to reflect the time series information of changes in hemoglobin concentration. The hemoglobin time series features may include the mean hemoglobin value and the rate of hemoglobin decline. The mean hemoglobin value is used to reflect the overall level of hemoglobin concentration, and the rate of hemoglobin decline is used to assess the speed and trend of hemoglobin changes. The perfusion index fluctuation features may include the perfusion index fluctuation pattern, which is used to describe the stability or fluctuation of the perfusion index.

[0080] In step H20, the anemia prediction model predicts the anemia risk level based on the hemoglobin time series characteristics, perfusion index fluctuation characteristics, dialysis data, and physical sign data. Among them, the dialysis data can include dialysate residence time, ultrafiltration volume, and toxin clearance rate. The dialysate residence time is used to reflect the dialysis duration of chronic kidney disease patients, the ultrafiltration volume is used to reflect the patient's dialysis dose, and the toxin clearance rate is used to reflect the adequacy of dialysis. The age in the physical sign data will affect the hemoglobin level and anemia risk, and the EPO dose is directly related to the effect of anemia treatment. To this end, combining the hemoglobin time series characteristics, perfusion index fluctuation characteristics, dialysis data, and physical sign data to predict the anemia risk level can more accurately analyze the dynamic changes of hemoglobin in chronic kidney disease patients and improve the prediction accuracy of the anemia risk level.

[0081] The anemia prediction model is a pre-trained neural network model used to predict the anemia risk level of patients with chronic kidney disease. The anemia prediction model uses a random forest model, a convolutional neural network model, or the like. In an embodiment of the present application, the anemia prediction model uses a random forest model. The pre-training process of the random forest model can be to prepare a training data set and use a feature extraction model to extract features from the training data in the training data set to obtain a target training data set. Each target training data set in the target training data set includes hemoglobin time series features, perfusion index fluctuation features, dialysis data, and vital sign data. At the same time, each target training data set corresponds to an anemia risk level label. Then, the target training data set is used to train the initial random forest model to learn the contribution of hemoglobin time series features, perfusion index fluctuation features, dialysis data, and vital sign data to anemia risk. Through cross-validation and hyperparameter tuning, the optimal model configuration (such as the number of trees, maximum depth, etc.) is selected to obtain a trained random forest model.

[0082] Furthermore, when anemia risk levels are predicted using the anemia prediction model, the corresponding predicted probability is also predicted. The predicted probability is used to reflect the reliability of the predicted anemia risk level. Predicted anemia risk levels can include no anemia risk level, mild anemia risk level, moderate anemia risk level, and severe anemia risk level. This is because the World Health Organization classifies different hemoglobin concentrations into three anemia levels: mild anemia (male: 110-129, female: 110-119), moderate anemia (80-109g / L), and severe anemia (<80g / L). To ensure that the predicted anemia risk level matches the anemia levels classified by the World Health Organization, the three anemia risk level categories of mild anemia risk level, moderate anemia risk level, and severe anemia risk level are configured in the anemia prediction model. In addition, since there may be cases where the patient is not anemic during anemia testing, a no-anemia risk level is configured in the anemia prediction model to improve the comprehensiveness of the anemia prediction model. Among them, when the hemoglobin concentration of men is <130g / L, the hemoglobin concentration of non-pregnant women is <120g / L, and the hemoglobin concentration of pregnant women is <110g / L, they will be judged as anemic. The no-anemia risk level corresponds to a hemoglobin concentration of >=130g / L for men, >=120g / L for non-pregnant women, and >=110g / L for pregnant women.

[0083] In one embodiment, performing anemia warning according to the predicted anemia risk level and the physiological data specifically includes:

[0084] Matching hemoglobin data, perfusion index data, and predicted anemia risk level with a preset warning level condition library;

[0085] When the hemoglobin data, perfusion index data, and predicted anemia risk level meet the first-level warning conditions, a first-level warning is triggered to generate a chronic kidney disease patient prompt message;

[0086] When the hemoglobin data, perfusion index data and predicted anemia risk level meet the secondary warning conditions, a secondary warning is triggered to generate a prompt message for the chronic kidney disease patient's medical and / or family members;

[0087] When the hemoglobin data, perfusion index data and predicted anemia risk level meet the third-level warning conditions, the third-level warning is triggered to automatically locate and start the emergency channel.

[0088] Specifically, the preset warning level condition library includes first-level warning conditions, second-level warning conditions and third-level warning conditions. The first-level warning condition is that the hemoglobin at the current time is less than the first preset threshold and greater than the second preset threshold; the first-level warning condition is that the predicted anemia risk level is greater than or equal to the moderate anemia risk level, and when the predicted anemia risk level is equal to the moderate anemia risk level, the predicted probability of the predicted anemia risk level is greater than the preset probability threshold; the third-level warning condition is that the hemoglobin at the current time is less than the second preset threshold or the perfusion index at the current time is less than the preset perfusion index threshold.

[0089] Furthermore, the current hemoglobin value refers to the hemoglobin data collected at the closest collection time to the predicted anemia risk level. In other words, the current time refers to the collection time closest to the predicted anemia risk level. The first preset threshold, the second preset threshold, the preset probability threshold, and the preset perfusion index threshold are all pre-set, with the first preset threshold being greater than the second preset threshold. For example, the first preset threshold is the anemia threshold (e.g., 130 g / L), the preset probability threshold is 0.7, and the preset perfusion index threshold is 0.3.

[0090] It should be noted that after the processing device receives the hemoglobin data synchronized with the hemoglobin detection device, it will directly perform an anemia determination based on the hemoglobin data to determine the actual anemia level corresponding to the hemoglobin data, and display the different actual anemia levels in different colors to serve as a reminder to the patient. In other words, after synchronously receiving the hemoglobin data, the embodiment of the present application will calculate the actual anemia level based on the hemoglobin data, and at the same time, based on the hemoglobin data and a number of historical hemoglobin data, predict the predicted anemia risk level for the future time period. Finally, the actual risk level (i.e., comparing the hemoglobin at the current time with the first preset threshold and the second preset threshold) and the predicted anemia risk level (i.e., combining the physiological data, the dialysis data, and the vital signs data of the chronic kidney disease patient to predict the anemia risk level) are combined to form an anemia warning, thereby improving the timeliness and accuracy of the anemia warning.

[0091] In one embodiment, in order to more clearly reflect the anemia condition and the relationship between dialysis efficacy and anemia risk, the processing device is also used to generate a hemoglobin-ultrafiltration volume correlation heat map based on physiological data, generate an anemia risk trend curve based on the predicted anemia risk level and several historical predicted anemia risk levels, and / or generate a correlation map between dialysis efficacy and anemia risk based on the predicted anemia risk level to demonstrate the impact of dialysis on hemoglobin levels.

[0092] In one embodiment, Figure 2 As shown, the home non-invasive anemia monitoring and early warning system in the chronic kidney disease scenario can also include a cloud device 400, a medical terminal 500 and a family terminal 600. The cloud device 400 communicates with the processing device 300, and the medical terminal 500 and the family terminal 600 both communicate with the cloud device 400. As a data hub and remote collaboration platform, the cloud device 400 focuses on the efficient storage, analysis and cross-terminal distribution of full data. Among them, the cloud device 400 and the processing device 300 use encrypted transmission, that is, the processing device 300 uses encryption technology to encrypt the data to be transmitted to the cloud device, and the cloud device decrypts the received encrypted data for remote access by medical staff, so that medical staff can adjust the treatment plan in time, thereby ensuring the effectiveness and safety of the treatment and the security of patient data. In addition, the processing device 300 can also be connected to the emergency center to send anemia early warning information to the emergency center.

[0093] Cloud device 400 is used to receive physiological data, dialysis data, and predicted anemia risk levels, and stores the received data in a time-series database and a relational database, ensuring data security and accessibility. The cloud data platform then receives online prescription revisions made by medical staff through visualization panels on the medical side (such as risk heat maps and anemia risk timelines). These revisions are delayed up to 10 seconds before reaching the dialysis machine, ensuring timely and safe interventions. Simultaneously, based on anemia warnings from the treatment device, alarms are sent to family members and medical staff to ensure a timely response. The cloud device can also simultaneously notify the emergency center of relevant information to the medical staff. Finally, the cloud device supports exporting historical data as structured PDF reports. These PDF reports can include anemia trend analysis (e.g., "Hemoglobin concentration increased after EPO dose adjustment"), intervention records, and efficacy evaluations, providing standardized support for clinical review. This allows medical staff to monitor patient status and historical data at any time, enabling early prediction of anemia and timely adjustments to prescriptions and nutritional status to prevent anemia and reduce the severity of renal anemia in patients with chronic kidney disease.

[0094] The Medical Device 500 connects medical staff to monitor patients' vital signs in real time via cloud-based devices. It receives physiological and dialysis data from cloud-based devices, as well as predicts anemia risk levels. It can also revise prescriptions online and synchronize them with dialysis machines via the cloud, significantly improving response speed and service quality. Furthermore, medical staff can communicate with family members through the Medical Device, further enhancing the sense of security and safety of chronic kidney disease patients.

[0095] The family member terminal 600 is used to connect with the patient's family. For example, the family member terminal is configured with the contact information of two family members. One family member serves as the primary contact, responsible for daily communication with medical staff through the family member terminal and obtaining updates on the patient's health data, including notifications of normal and abnormal data. At the same time, the contact information of the other family member serves as a backup contact. In an emergency or if the primary family member cannot be reached, the backup contact can be quickly found, ensuring unimpeded information transmission and allowing chronic kidney patients to receive necessary help and support as soon as possible.

[0096] The above completes the description of the home non-invasive anemia monitoring and early warning system for chronic kidney disease scenarios. Based on the home non-invasive anemia monitoring and early warning system for chronic kidney disease scenarios, the present embodiment provides a home non-invasive anemia monitoring and early warning method for chronic kidney disease scenarios, the method comprising collecting physiological data of chronic kidney patients, wherein the physiological data includes hemoglobin data and perfusion index data;

[0097] Perform dialysis on chronic kidney disease patients and record dialysis data;

[0098] The predicted anemia risk level of the chronic kidney patient is determined based on the physiological data, the dialysis data and the vital sign data of the chronic kidney patient, and anemia warning is performed based on the predicted anemia risk level and the physiological data.

[0099] In one embodiment, the home non-invasive anemia monitoring and early warning method in the chronic kidney disease scenario further includes:

[0100] The physiological data, dialysis data, and predicted anemia risk level are synchronized to a cloud device, so that the predicted anemia risk level is pushed to a medical care end and / or a family member end via the cloud device.

[0101] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A non-invasive home anemia monitoring and early warning system for chronic kidney disease, characterized by: The home non-invasive anemia monitoring and early warning system in the chronic kidney disease scenario specifically includes hemoglobin detection equipment, dialysis machines, and processing equipment; The hemoglobin detection device is used to collect physiological data of chronic kidney disease patients, wherein the physiological data includes hemoglobin data and perfusion index data; The dialysis machine is used to perform dialysis on chronic kidney patients and record dialysis data; The processing device is used to determine a predicted anemia risk level of the chronic kidney disease patient based on the physiological data, the dialysis data, and the vital sign data of the chronic kidney disease patient, and to provide an anemia early warning based on the predicted anemia risk level and the physiological data; The step of determining the predicted anemia risk level of the chronic kidney disease patient based on the physiological data, the dialysis data, and the physical sign data of the chronic kidney disease patient specifically includes: generating initial hemoglobin baseline data based on the hemoglobin data; Dynamically adjusting the initial hemoglobin baseline data according to the perfusion index data to obtain hemoglobin baseline data; Inputting the hemoglobin baseline data and the perfusion index data into a pre-trained feature extraction model, and outputting the hemoglobin time series feature and the perfusion index fluctuation feature through the pre-trained feature extraction model; inputting the hemoglobin time series characteristics, the perfusion index fluctuation characteristics, the dialysis data, and the physical sign data into a pre-trained anemia prediction model, and determining the predicted anemia risk level of the chronic kidney disease patient using the anemia prediction model; The performing of anemia warning according to the predicted anemia risk level and the physiological data specifically includes: Matching the hemoglobin data, perfusion index data, and predicted anemia risk level with a preset warning level condition library, wherein the preset warning level condition library includes first-level warning conditions, second-level warning conditions, and third-level warning conditions; When the hemoglobin data, perfusion index data, and predicted anemia risk level meet the first-level warning conditions, a first-level warning is triggered to generate a chronic kidney disease patient prompt message; When the hemoglobin data, perfusion index data and predicted anemia risk level meet the secondary warning conditions, a secondary warning is triggered to generate a prompt message for the chronic kidney disease patient's medical and / or family members; When the hemoglobin data, perfusion index data, and predicted anemia risk level meet the third-level warning conditions, a third-level warning is triggered to automatically locate and activate the emergency channel; Among them, the first-level warning condition is that the hemoglobin at the current time is less than the first preset threshold and greater than the second preset threshold; the first-level warning condition is that the predicted anemia risk level is greater than or equal to the moderate anemia risk level, and when the predicted anemia risk level is equal to the moderate anemia risk level, the predicted probability of the predicted anemia risk level is greater than the preset probability threshold; the third-level warning condition is that the hemoglobin at the current time is less than the second preset threshold or the perfusion index at the current time is less than the preset perfusion index threshold.

2. The home non-invasive anemia monitoring and early warning system for chronic kidney disease according to claim 1 is characterized in that: Before determining the predicted anemia risk level of the chronic kidney disease patient based on the physiological data, the dialysis data, and the physical sign data of the chronic kidney disease patient, the method further includes: Acquiring altitude data of the chronic kidney disease patient; The hemoglobin and the perfusion index are corrected based on the altitude data.

3. The home non-invasive anemia monitoring and early warning system for chronic kidney disease according to claim 2 is characterized in that: After correcting the hemoglobin and the perfusion index according to the altitude data, the method further includes: comparing the perfusion index with a preset perfusion index threshold; If the perfusion index is less than the perfusion index threshold, the perfusion index is manually confirmed. If the confirmation is passed, the perfusion index is retained; if the confirmation is not passed, the perfusion index is discarded.

4. The home non-invasive anemia monitoring and early warning system for chronic kidney disease according to claim 1 is characterized in that: The processing device is also used to generate a hemoglobin-ultrafiltration volume correlation heat map based on physiological data, generate an anemia risk trend curve based on the predicted anemia risk level and several historical predicted anemia risk levels, and / or generate a correlation map between dialysis efficacy and anemia risk based on the predicted anemia risk level to demonstrate the impact of dialysis on hemoglobin levels.

5. The home non-invasive anemia monitoring and early warning system for chronic kidney disease according to claim 1 is characterized in that: The home non-invasive anemia monitoring and early warning system in the chronic kidney disease scenario also includes: The cloud device is used to synchronize physiological data, dialysis data and predicted anemia risk level, and push the predicted anemia risk level to medical staff and / or family members.

6. A non-invasive home anemia monitoring and early warning method for chronic kidney disease, characterized in that: The non-invasive home anemia monitoring and early warning system for chronic kidney disease according to any one of claims 1 to 5 is used, and the non-invasive home anemia monitoring and early warning method for chronic kidney disease specifically comprises: Collecting physiological data of a chronic kidney disease patient, wherein the physiological data includes hemoglobin data and perfusion index data; Perform dialysis on chronic kidney disease patients and record dialysis data; The predicted anemia risk level of the chronic kidney patient is determined based on the physiological data, the dialysis data and the vital sign data of the chronic kidney patient, and anemia warning is performed based on the predicted anemia risk level and the physiological data.

7. The method for non-invasive home anemia monitoring and early warning in the context of chronic kidney disease according to claim 6, characterized in that: The method further comprises: The physiological data, dialysis data, and predicted anemia risk level are synchronized to a cloud device, so that the predicted anemia risk level is pushed to a medical care end and / or a family member end via the cloud device.

Images