Nutrition risk assessment and monitoring method and system for cancer palliative treatment patient

Abstract

The invention relates to the technical field of medical health monitoring and evaluation, and discloses a method and system for evaluating and monitoring the nutrition risk of a cancer palliative treatment patient, and the method comprises the steps: obtaining a daily diet image of the cancer patient through an image device, recognizing the type of food in the daily diet image, recognizing the weight of food in the daily diet image, and calculating the nutrition risk of the cancer palliative treatment patient; analyzing diner actions in the daily diet image; the daily diet condition of the cancer patient is analyzed according to the food category, the food weight, the diner action and the target food, the nutrition change item of the initial nutrition condition is determined according to the daily diet condition, and the nutrition risk condition of the cancer patient is identified according to the nutrition change item; determining target monitoring equipment of the cancer patient from non-image equipment; and constructing a nutrition risk sequence of the nutrition risk level, analyzing the nutrition risk trend of the cancer patient by using the nutrition risk sequence, and determining a nutrition risk monitoring result of the cancer patient. The convenience of nutrition risk assessment of the patient can be improved.

Description

Technical Field

[0001] The present invention relates to a method and system for nutritional risk assessment and monitoring of cancer palliative care patients, belonging to the technical field of medical health monitoring and assessment. Background Art

[0002] At present, cancer palliative care patients in the prior art often face the problem of malnutrition, which directly affects the quality of life and survival period of patients. Traditional nutritional assessment methods rely on regular hospital visits and laboratory tests. These methods are not only costly but also unable to provide real-time data. With the development of wearable technology, devices such as smart bracelets provide new possibilities for real-time health monitoring. However, existing monitoring means are mostly regular inspections and cannot achieve real-time monitoring, resulting in insufficient timeliness of nutritional risk assessment and intervention.

[0003] Currently, some technologies collect patients' physiological parameters, nutritional indicators, and activity data through various biosensors, analyze the collected data using advanced algorithms to evaluate patients' nutritional risks, automatically generate personalized nutritional intervention plans according to the evaluation results, monitor patients' nutritional status in real time, and feedback the evaluation results and intervention plans to medical staff and patients. However, the above solutions have defects. Patients need to wear multiple biosensors on their bodies. If the biosensors are large in size, it will cause the problem of complex device wearing for patients. Therefore, the simplicity of patients' nutritional risk assessment is insufficient. Summary of the Invention

[0004] The present invention provides a method and system for nutritional risk assessment and monitoring of cancer palliative care patients, and its main purpose is to improve the simplicity of patients' nutritional risk assessment.

[0005] To achieve the above object, a method for nutritional risk assessment and monitoring of cancer palliative care patients provided by the present invention includes:

[0006] Detect the initial nutritional status of cancer patients, and determine the nutritional monitoring device of the cancer patients by using the initial nutritional status, wherein the nutritional monitoring device includes an image device and a non-image device;

[0007] Use the image device to obtain the daily diet images of the cancer patients, identify the food categories in the daily diet images, identify the food weights in the daily diet images, and analyze the actions of the diners in the daily diet images, and record the target foods corresponding to the actions of the diners;

[0008] Analyze the daily diet status of cancer patients through the food category, the food weight, the actions of the diners, and the target food, determine the nutritional change items of the initial nutritional status using the daily diet status, and identify the nutritional risk status of the cancer patients using the nutritional change items;

[0009] Determine the target monitoring device of the cancer patient from the non-image devices according to the nutritional change items and the nutritional risk status, monitor the nutritional status of the cancer patient during the period through the target monitoring device, and analyze the nutritional risk level corresponding to the nutritional status during the period;

[0010] Construct a nutritional risk sequence of the nutritional risk level, analyze the nutritional risk trend of the cancer patient using the nutritional risk sequence, and determine the nutritional risk monitoring result of the cancer patient based on the nutritional risk trend.

[0011] Optionally, the identification of the food weight in the daily diet image includes:

[0012] Divide the side view and the top view in the daily diet image;

[0013] Identify the height of the whole plate of food in the daily diet image from the side view;

[0014] Identify the bottom area of the whole plate of food in the daily diet image from the top view;

[0015] Calculate the volume of the whole plate of food in the daily diet image using the height of the whole plate of food and the bottom area of the whole plate of food;

[0016] Obtain the food category in the daily diet image;

[0017] Query the number of times of the food and the food density of the food category;

[0018] Calculate the food weight in the daily diet image according to the volume of the whole plate of food, the number of times of the food, and the food density.

[0019] Optionally, the analysis of the actions of the diners in the daily diet image includes:

[0020] Obtain the mouth target, hand target, tableware target, and food target in the daily diet image;

[0021] Query the tableware moment when the hand target touches the tableware target;

[0022] Query the plate moment when the tableware target touches the food target;

[0023] Query the mouth moment when the tableware target touches the mouth target;

[0024] Within the time period from the tableware moment to the dinner plate moment and from the dinner plate moment to the mouth moment, the diner corresponding to the mouth target and the dining action are taken as the diner actions in the daily diet image.

[0025] Optionally, analyzing the daily diet status of the cancer patient through the food category, the food weight, the diner action, and the target food includes:

[0026] Calculating the food consumption amount of the food category according to the food weight

[0027] Counting the dining frequency of the diner action with respect to the target food;

[0028] Allocating the weight weight of the diner corresponding to the diner action according to the dining frequency;

[0029] Calculating the dining weight of the cancer patient according to the weight weight;

[0030] Taking the dining weight of the cancer patient with respect to the food category as the daily diet status.

[0031] Optionally, using the daily diet status to determine the nutritional change items of the initial nutritional status includes:

[0032] Obtaining the food category and the dining weight in the daily diet status;

[0033] Calculating the first nutritional change item of the first nutritional component in the initial nutritional status based on the food category and the dining weight;

[0034] Calculating the second nutritional change item of the second nutritional component in the initial nutritional status;

[0035] Taking the first nutritional change item and the second nutritional change item as the nutritional change items of the initial nutritional status.

[0036] Optionally, using the nutritional change items to identify the nutritional risk status of the cancer patient includes:

[0037] Calculating the hidden layer vector of the nutritional change items;

[0038] Decoding and reconstructing the hidden layer vector to obtain a decoded vector;

[0039] Querying the nutritional risk status corresponding to the decoded vector.

[0040] Optionally, determining the target monitoring device of the cancer patient from the non-image device according to the nutritional change items and the nutritional risk status includes:

[0041] Determine whether there is a risk in the nutritional risk status;

[0042] When there is a risk in the nutritional risk status, extract the abnormal nutritional items in the nutritional change items;

[0043] Identify the target monitoring device corresponding to the abnormal nutritional item from the non-image devices.

[0044] Optionally, constructing the nutritional risk sequence of the nutritional risk level includes:

[0045] Query the level analysis moment of the nutritional risk level;

[0046] Arrange the nutritional risk levels in ascending order of the level analysis moment to obtain a nutritional risk sequence.

[0047] Optionally, using the nutritional risk sequence to analyze the nutritional risk trend of the cancer patient includes:

[0048] Determine the trend analysis model of the nutritional risk sequence;

[0049] Use the bidirectional long short-term memory network in the trend analysis model to analyze the first sequence feature of the nutritional risk sequence;

[0050] Use the long short-term memory network in the trend analysis model to analyze the second sequence feature of the first sequence feature;

[0051] Use the time attention mechanism in the trend analysis model to analyze the attention weight of the second sequence feature;

[0052] Use the weighted sum layer in the trend analysis model to perform weighted sum between the second sequence feature and the attention weight to obtain a weighted sum result;

[0053] Analyze the nutritional risk trend of the cancer patient through the weighted sum result.

[0054] To solve the above problems, the present invention also provides a nutritional risk assessment and monitoring system for cancer palliative care patients, and the system includes:

[0055] A device determination module, configured to detect the initial nutritional status of a cancer patient, and use the initial nutritional status to determine the nutritional monitoring device of the cancer patient, where the nutritional monitoring device includes an image device and a non-image device;

[0056] A food recording module, which is used to obtain the daily diet images of the cancer patient by using the image device, identify the food categories in the daily diet images, identify the food weights in the daily diet images, analyze the actions of the diners in the daily diet images, and record the target foods corresponding to the actions of the diners;

[0057] A risk identification module, which is used to analyze the daily diet status of the cancer patient through the food categories, the food weights, the actions of the diners, and the target foods, determine the nutritional change items of the initial nutritional status by using the daily diet status, and identify the nutritional risk status of the cancer patient by using the nutritional change items;

[0058] A level analysis module, which is used to determine the target monitoring device of the cancer patient from the non-image devices according to the nutritional change items and the nutritional risk status, monitor the nutritional status of the cancer patient during a period through the target monitoring device, and analyze the nutritional risk level corresponding to the nutritional status during the period;

[0059] A risk monitoring module, which is used to construct a nutritional risk sequence of the nutritional risk level, analyze the nutritional risk trend of the cancer patient by using the nutritional risk sequence, and determine the nutritional risk monitoring result of the cancer patient based on the nutritional risk trend.

[0060] Compared with the problems described in the background art, in the embodiment of the present invention, the daily diet images of the cancer patient are obtained by using the image device, so as to analyze the food content eaten by the patient from the daily diet images in the follow-up. Further, in the embodiment of the present invention, the nutritional risk status of the cancer patient is identified by using the nutritional change items, so as to preliminarily analyze whether the eaten foods improve the malnutrition status of the patient based on the food content and food categories. If the malnutrition status of the patient is improved, there is no need to use other non-image devices except the image device to monitor the patient's body, which can greatly reduce the usage rate of other non-image devices, devices that are difficult to wear and cannot be worn, etc., thereby improving the simplicity of the nutritional risk assessment of the patient. Description of the Drawings

[0061] Figure 1 It is a schematic flowchart of a method for nutritional risk assessment and monitoring of cancer palliative care patients provided by an embodiment of the present invention;

[0062] Figure 2 It is a schematic diagram of a module for implementing the method for nutritional risk assessment and monitoring of cancer palliative care patients provided by an embodiment of the present invention.

[0063] The realization, functional features and advantages of the object of the present invention will be further described with reference to the embodiments and the accompanying drawings. Detailed Embodiment

[0064] It should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0065] An embodiment of the present application provides a method for evaluating and monitoring the nutritional risk of cancer palliative care patients. The execution subject of the method for evaluating and monitoring the nutritional risk of cancer palliative care patients includes, but is not limited to, at least one of electronic devices such as a server, a terminal, etc. that can be configured to execute the method provided by the embodiment of the present application. In other words, the method for evaluating and monitoring the nutritional risk of cancer palliative care patients can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to: a single server, a server cluster, a cloud server, or a cloud server cluster, etc.

[0066] Example 1:

[0067] Referring to Figure 1 As shown, it is a schematic flowchart of a method for evaluating and monitoring the nutritional risk of cancer palliative care patients provided by an embodiment of the present invention. In this embodiment, the method for evaluating and monitoring the nutritional risk of cancer palliative care patients includes:

[0068] S1. Detect the initial nutritional status of a cancer patient, and use the initial nutritional status to determine the nutritional monitoring device of the cancer patient, where the nutritional monitoring device includes an image device and a non-image device.

[0069] In an embodiment of the present invention, the initial nutritional status includes the names and weights of nutritional components of the cancer patient. The name of the nutritional component refers to the name of a nutritional element that can determine whether a patient is malnourished. The image device in the nutritional monitoring device refers to a device that collects image information in front of the patient, and the non-image device refers to a device that monitors the numerical changes of the nutritional component corresponding to the name of the nutritional component, such as a biosensor.

[0070] S2. Use the image device to obtain the daily diet image of the cancer patient, identify the food categories in the daily diet image, identify the food weights in the daily diet image, and analyze the actions of the diners in the daily diet image, and record the target foods corresponding to the actions of the diners.

[0071] In an embodiment of the present invention, the daily diet image refers to an image taken when the patient is eating, including the patient, the patient's family members, each dish on the table in the home scene, chopsticks, etc.

[0072] Optionally, the process of identifying food categories in the daily diet image refers to the process of using an object detection model to detect food types, such as food categories like tomatoes, eggs, rice, etc., and the object detection model is, for example, the YOLO algorithm model.

[0073] In an embodiment of the present invention, the identifying the food weight in the daily diet image includes: dividing the side view and top view in the daily diet image; identifying the height of the whole plate of food in the daily diet image from the side view; identifying the bottom area of the whole plate of food in the daily diet image from the top view; calculating the volume of the whole plate of food in the daily diet image by using the height of the whole plate of food and the bottom area of the whole plate of food; obtaining the food categories in the daily diet image; querying the number of times of the food categories and the food density of the food categories; and calculating the food weight in the daily diet image according to the volume of the whole plate of food, the number of times of the food, and the food density by using the following formula:

[0074]

[0075] where x i represents the food weight of the i-th food category, V represents the volume of the whole plate of food, N0 represents the total number of times of all food categories, N i represents the number of times of the i-th food category, and μ i represents the food density of the i-th food category.

[0076] Among them, the food weight refers to the weight of each food category eaten by the person in the daily diet image. For example, the weight of the tomatoes eaten by a certain person in the daily diet image. The side view refers to the top view of each plate of dish, and the top view refers to the side view of each plate of dish. The height of the whole plate of food refers to the height between the top of the dish in the plate and the bottom of the plate in the side view. The bottom area of the whole plate of food refers to the top view area occupied by each plate of dish in the top view. The volume of the whole plate of food refers to the product of the height of the whole plate of food and the bottom area of the whole plate of food. The number of times of the food refers to the number of times each food category is touched by chopsticks. When the chopsticks touch the food category, it is defaulted that the chopsticks pick up the food category and eat this food category.

[0077] In an embodiment of the present invention, the analyzing the dining person's actions in the daily diet image includes: obtaining the mouth target, hand target, tableware target, and food target in the daily diet image; querying the tableware moment when the hand target touches the tableware target; querying the plate moment when the tableware target touches the food target; querying the mouth moment when the tableware target touches the mouth target; and taking the dining person corresponding to the mouth target and the dining action as the dining person's actions in the daily diet image within the tableware moment to the plate moment and the plate moment to the mouth moment.

[0078] Among them, the actions of the diner include the action information of the user from picking up chopsticks to eating within a period of time. The dining action only includes the actions from picking up chopsticks to eating within a period of time and does not include the diner. The target food refers to the food category consumed by the user. For example, when the tableware target contacts the mouth target at the mouth moment, the tableware target is, for example, chopsticks, and the tomato clamped on the chopsticks.

[0079] S3. Analyze the daily diet status of the cancer patient through the food category, the food weight, the actions of the diner, and the target food. Use the daily diet status to determine the nutritional change items of the initial nutritional status, and use the nutritional change items to identify the nutritional risk status of the cancer patient.

[0080] In an embodiment of the present invention, the analysis of the daily diet status of the cancer patient through the food category, the food weight, the actions of the diner, and the target food includes: According to the food weight, use the following formula to calculate the food consumption amount of the food category:

[0081]

[0082] Among them, Δx i represents the food consumption amount, represents the food weight of the i-th food category at the end of the meal, represents the food weight of the i-th food category at the start of the meal, and i represents the serial number of the food category;

[0083] Count the dining frequency of the actions of the diner regarding the target food; According to the dining frequency, use the following formula to allocate the weight of the diner corresponding to the actions of the diner:

[0084]

[0085] Among them, α ij represents the weight of the j-th diner, β ij represents the dining frequency of the j-th diner for the i-th food category, β im represents the total dining frequency of m diners for the i-th food category, j represents the serial number of the diner, and i represents the serial number of the food category;

[0086] According to the weight, use the following formula to calculate the dining weight of the cancer patient:

[0087] Δx ij =α ij Δx i , j = δ

[0088] Among them, Δxij represents the dining weight, δ represents the serial number of the cancer patient, and α ij represents the weight weight of the j-th diner, and Δx i represents the food consumption amount, j represents the serial number of the diner, and i represents the serial number of the food category;

[0089] Take the dining weight of the cancer patient for the food category as the daily diet condition.

[0090] Among them, the dining frequency refers to the number of times the diner uses the tableware to contact the target food, and the dining frequency can be determined by the plate moment when the aforementioned tableware targets to contact the food target. The daily diet condition refers to the dining weight of each food category eaten by each diner.

[0091] In an embodiment of the present invention, the nutritional change item for determining the initial nutritional condition using the daily diet condition includes: obtaining the food category and dining weight in the daily diet condition; based on the food category and the dining weight, using the following formula to calculate the first nutritional change item of the first nutrient in the initial nutritional condition:

[0092] y k1 = y k0 - y kΔt + ρ ik Δx ij , j = δ

[0093] Among them, y k1 represents the first nutritional change item, y k0 represents the initial weight of the k-th first nutrient, and y kΔt represents the weight of the k-th first nutrient decomposed within the time period of Δt, and ρ ik represents the ratio between the weight of the nutrient and the weight of the food category, and Δx ij represents the dining weight, δ represents the serial number of the cancer patient, j represents the serial number of the diner, and i represents the serial number of the food category;

[0094] Use the following formula to calculate the second nutritional change item of the second nutrient in the initial nutritional condition:

[0095] z k'1 = z k'0 - z k'Δt

[0096] Among them, z k′1 represents the second nutritional change item, and z k′0 represents the initial weight of the k'-th second nutrient, and z k′Δt represents the weight of the k'-th second nutrient decomposed within the time period of Δt;

[0097] Take the first nutritional change item and the second nutritional change item as the nutritional change items of the initial nutritional status.

[0098] Among them, the first nutritional component refers to the components contained in the food eaten by the patient, and the second nutritional component refers to the components not contained in the detected food during the time period of identifying the food category. For example, if the food eaten during the lunch dining period does not contain the second nutritional component, then since the patient has not supplemented the second nutritional component, the second nutritional component will be gradually digested in the patient's body. The speed at which the patient digests the nutritional component, that is, y kΔt , z k′Δt It also needs to be obtained based on historical data statistics. For example, during a historical period, it is statistically determined how long it takes for the patient to digest a certain weight of a certain nutritional component, so as to obtain y kΔt , z k′Δt , similarly, ρ ik represents the proportion of the food eaten by the user that is converted into the corresponding first nutritional component. During a historical period, it is statistically determined how much weight of food the patient eats and how much weight of the first nutritional component can be converted.

[0099] In an embodiment of the present invention, the use of the nutritional change item to identify the nutritional risk status of the cancer patient includes: calculating the hidden layer vector of the nutritional change item using the following formula:

[0100] g = σ e (W(y k1 , z [[ID=2,5]] k′1 ) + b)

[0101] Among them, g represents the hidden layer vector, σ e represents the activation function of the encoder, W represents the weight of the encoder, b represents the bias of the encoder, y k1 represents the first nutritional change item, and z k′1 represents the second nutritional change item;

[0102] Decode and reconstruct the hidden layer vector using the following formula to obtain a decoded vector:

[0103] h = σ d (W′g + b)

[0104] Among them, h represents the decoded vector, σ d represents the activation function of the decoder, W represents the weight of the decoder, b represents the bias of the decoder, and g represents the hidden layer vector;

[0105] Query the nutritional risk status corresponding to the decoded vector.

[0106] Among them, the decoding vector refers to the probability value of the risk level, and the risk level is the data at the output end set when training the encoder and decoder. The risk level refers to the level of evaluating whether a patient is malnourished, such as healthy, good, low risk, high risk, etc.

[0107] S4. According to the nutritional change items and the nutritional risk status, determine the target monitoring device of the cancer patient from the non-image devices, monitor the nutritional status of the cancer patient during the period through the target monitoring device, and analyze the nutritional risk level corresponding to the nutritional status during the period.

[0108] In an embodiment of the present invention, the determining the target monitoring device of the cancer patient from the non-image devices according to the nutritional change items and the nutritional risk status includes: judging whether there is a risk in the nutritional risk status; when there is a risk in the nutritional risk status, extracting the abnormal nutritional items in the nutritional change items; identifying the target monitoring device corresponding to the abnormal nutritional items from the non-image devices.

[0109] Among them, the abnormal nutritional item refers to the nutritional component whose value exceeds the normal range and the value of the nutritional component. It should be noted that when the value of the abnormal nutritional item is relatively low and the impact of the category of the abnormal nutritional item on malnutrition is relatively low, even if there are abnormal nutritional items in the nutritional change items, the risk level can be healthy or good. The target monitoring device refers to the device specifically for monitoring the abnormal nutritional item. For example, if the abnormal nutritional item is substance A, the target monitoring device is the device specifically for monitoring substance A. Further, the nutritional status during the period refers to the numerical change of the abnormal nutritional item monitored by the target monitoring device in the subsequent time period, and the subsequent time period refers to the time period after the moment when the target monitoring device corresponding to the abnormal nutritional item is identified. The nutritional risk level is similar to the aforementioned nutritional risk status. Similarly, the process of analyzing the nutritional risk level corresponding to the nutritional status during the period is similar to the principle of using the nutritional change items to identify the nutritional risk status of the cancer patient, and will not be elaborated here.

[0110] S5. Construct a nutritional risk sequence of the nutritional risk level, analyze the nutritional risk trend of the cancer patient by using the nutritional risk sequence, and determine the nutritional risk monitoring result of the cancer patient based on the nutritional risk trend.

[0111] In an embodiment of the present invention, the constructing the nutritional risk sequence of the nutritional risk level includes: querying the level analysis time of the nutritional risk level; arranging the nutritional risk levels in ascending order of the level analysis time to obtain a nutritional risk sequence.

[0112] Among them, the grading analysis moment refers to the moment when the nutrition risk grade is output by the aforementioned encoder and decoder. There are multiple grading analysis moments, that is, within a period of time, the encoder and decoder will output the corresponding nutrition risk grade based on the nutritional components monitored at each moment during this period.

[0113] In an embodiment of the present invention, the use of the nutrition risk sequence to analyze the nutrition risk trend of the cancer patient includes: determining a trend analysis model of the nutrition risk sequence; using a bidirectional long short-term memory network in the trend analysis model to analyze the first sequence feature of the nutrition risk sequence; using a long short-term memory network in the trend analysis model to analyze the second sequence feature of the first sequence feature; using a time attention mechanism in the trend analysis model to analyze the attention weight of the second sequence feature; using a weighted summation layer in the trend analysis model to perform weighted summation between the second sequence feature and the attention weight to obtain a weighted summation result; and analyzing the nutrition risk trend of the cancer patient through the weighted summation result.

[0114] Among them, the time attention mechanism is a technique for processing time series data in deep learning, which can assign different importance or attention to the information at different time steps.

[0115] Optionally, the process of analyzing the nutrition risk trend of the cancer patient through the weighted summation result refers to calculating the probability value corresponding to the weighted summation result using an activation function, querying the nutrition risk grade corresponding to the maximum probability value, and the trend analysis model will output the nutrition risk grade at each moment with a fixed time interval in the next period. The nutrition risk grades in these consecutive periods are used as the nutrition risk trend. Further, the process of determining the nutrition risk monitoring result of the cancer patient based on the nutrition risk trend is as follows: for example, when the nutrition risk grade at the k + 1 moment is better than that at the k moment, it indicates that the malnutrition state of the patient has improved. At this time, the target monitoring device is turned off, and the aforementioned S2 step is continued, and the improvement of malnutrition is used as the nutrition risk monitoring result. When the nutrition risk grade at the k + 1 moment is worse than that at the k moment, it indicates that the malnutrition state of the patient is gradually getting serious, and it is necessary to continue to use the target monitoring device to monitor the physical condition of the patient, and the poor performance of the nutrition risk grade and the device monitoring result are used as the nutrition risk monitoring result.

[0116] Compared with the problems described in the background art, in the embodiment of the present invention, the daily diet images of the cancer patient are obtained by using the image device, so as to analyze the food content consumed by the patient from the daily diet images in the subsequent process. Further, in the embodiment of the present invention, the nutritional risk status of the cancer patient is identified by using the nutritional change items, so as to preliminarily analyze whether the consumed food improves the malnutrition status of the patient based on the food content and food categories. If the malnutrition status of the patient is improved, there is no need to use other non-image devices except the image device to monitor the patient's body, which can greatly reduce the usage rate of other non-image devices, devices that are difficult to wear and cannot be worn, etc., thereby improving the simplicity of the patient's nutritional risk assessment.

[0117] Embodiment 2:

[0118] As Figure 2 shown, it is a functional module diagram of a nutritional risk assessment and monitoring system for cancer palliative care patients of the present invention.

[0119] The nutritional risk assessment and monitoring system 200 for cancer palliative care patients of the present invention can be installed in an electronic device. According to the functions achieved, the nutritional risk assessment and monitoring system for cancer palliative care patients can include a device determination module 201, a food recording module 202, a risk identification module 203, a level analysis module 204, and a risk monitoring module 205. The modules of the present invention can also be referred to as units, which refer to a series of computer program segments that can be executed by a processor of an electronic device and can complete fixed functions, and are stored in the memory of the electronic device.

[0120] In the embodiment of the present invention, the functions of each module / unit are as follows:

[0121] The device determination module 201 is used to detect the initial nutritional status of a cancer patient and determine the nutritional monitoring device for the cancer patient by using the initial nutritional status, wherein the nutritional monitoring device includes an image device and a non-image device;

[0122] The food recording module 202 is used to obtain the daily diet images of the cancer patient by using the image device, identify the food categories in the daily diet images, identify the food weights in the daily diet images, analyze the actions of the dining personnel in the daily diet images, and record the target food corresponding to the actions of the dining personnel;

[0123] The risk identification module 203 is used to analyze the daily diet status of the cancer patient through the food categories, the food weights, the actions of the dining personnel, and the target food, determine the nutritional change items of the initial nutritional status by using the daily diet status, and identify the nutritional risk status of the cancer patient by using the nutritional change items;

[0124] The level analysis module 204 is configured to determine a target monitoring device for the cancer patient from the non-image devices according to the nutrition change item and the nutrition risk status, monitor the in-period nutrition status of the cancer patient through the target monitoring device, and analyze the nutrition risk level corresponding to the in-period nutrition status;

[0125] The risk monitoring module 205 is configured to construct a nutrition risk sequence of the nutrition risk level, analyze the nutrition risk trend of the cancer patient by using the nutrition risk sequence, and determine the nutrition risk monitoring result of the cancer patient based on the nutrition risk trend.

[0126] Specifically, each module in the nutrition risk assessment and monitoring system 200 for cancer palliative care patients in the embodiments of the present invention adopts the same technical means as those in the Figure 1 nutrition risk assessment and monitoring method for cancer palliative care patients described above, and can produce the same technical effects, which will not be elaborated here.

[0127] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and without departing from the spirit or basic characteristics of the present invention, the present invention can be implemented in other specific forms.

[0128] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for nutritional risk assessment and monitoring of cancer palliative care patients, characterized in that The method includes: Detecting the initial nutritional status of a cancer patient, and determining a nutritional monitoring device for the cancer patient by using the initial nutritional status, wherein the nutritional monitoring device includes an imaging device and a non-imaging device; Obtaining daily diet images of the cancer patient by using the imaging device, identifying food categories in the daily diet images, identifying the weights of foods in the daily diet images, analyzing the actions of the diners in the daily diet images, and recording the target foods corresponding to the actions of the diners; Analyzing the daily diet status of the cancer patient based on the food categories, the food weights, the actions of the diners, and the target foods, determining nutritional change items of the initial nutritional status by using the daily diet status, and identifying the nutritional risk status of the cancer patient by using the nutritional change items; Determining a target monitoring device for the cancer patient from the non-imaging devices according to the nutritional change items and the nutritional risk status, monitoring the interim nutritional status of the cancer patient by using the target monitoring device, and analyzing the nutritional risk level corresponding to the interim nutritional status; Constructing a nutritional risk sequence of the nutritional risk level, analyzing the nutritional risk trend of the cancer patient by using the nutritional risk sequence, and determining the nutritional risk monitoring result of the cancer patient based on the nutritional risk trend.

2. The nutritional risk assessment and monitoring method for cancer palliative care patients according to claim 1, characterized in that, The identifying the weights of foods in the daily diet images includes: Dividing a side view and a top view in the daily diet images; Identifying the height of the whole plate of food in the daily diet images from the side view; Identifying the bottom area of the whole plate of food in the daily diet images from the top view; Calculating the volume of the whole plate of food in the daily diet images by using the height of the whole plate of food and the bottom area of the whole plate of food; Obtaining the food categories in the daily diet images; Querying the food frequency and food density of the food categories; Calculating the weights of foods in the daily diet images according to the volume of the whole plate of food, the food frequency, and the food density.

3. The nutritional risk assessment and monitoring method for cancer palliative care patients according to claim 1, characterized in that, The analyzing the actions of the diners in the daily diet images includes: Obtaining a mouth target, a hand target, a tableware target, and a food target in the daily diet images; Querying the tableware moment when the hand target touches the tableware target; Querying the plate moment when the tableware target touches the food target; Querying the mouth moment when the tableware target touches the mouth target; During the period from the tableware moment to the plate moment and from the plate moment to the mouth moment, taking the diner corresponding to the mouth target and the dining actions as the actions of the diners in the daily diet images.

4. The nutritional risk assessment and monitoring method for cancer palliative care patients according to claim 1, wherein The analyzing the daily diet status of the cancer patient based on the food categories, the food weights, the actions of the diners, and the target foods includes: Calculating the food intake of the food categories according to the food weights; Counting the dining frequency of the actions of the diners with respect to the target foods; Allocating a weight for the weight of the diner corresponding to the actions of the diners according to the dining frequency; Calculating the dining weight of the cancer patient according to the weight. Take the dining weight of the cancer patient for the food category as the daily diet condition.

5. The nutritional risk assessment and monitoring method for cancer palliative care patients according to claim 1, wherein, The nutritional change items for determining the initial nutritional condition using the daily diet condition include: Obtain the food category and dining weight in the daily diet condition; Based on the food category and the dining weight, calculate the first nutritional change item of the first nutritional component in the initial nutritional condition; Calculate the second nutritional change item of the second nutritional component in the initial nutritional condition; Take the first nutritional change item and the second nutritional change item as the nutritional change items of the initial nutritional condition.

6. The nutritional risk assessment and monitoring method for cancer palliative care patients according to claim 1, wherein The method for identifying the nutritional risk status of the cancer patient using the nutritional change items includes: Calculate the hidden layer vector of the nutritional change items; Decode and reconstruct the hidden layer vector to obtain a decoded vector; Query the nutritional risk status corresponding to the decoded vector.

7. The nutritional risk assessment and monitoring method for cancer palliative care patients according to claim 1, characterized in that The method for determining the target monitoring device of the cancer patient from the non-image devices according to the nutritional change items and the nutritional risk status includes: Judge whether there is a risk in the nutritional risk status; When there is a risk in the nutritional risk status, extract the abnormal nutritional items in the nutritional change items; Identify the target monitoring device corresponding to the abnormal nutritional items from the non-image devices.

8. The nutritional risk assessment and monitoring method for cancer palliative care patients according to claim 1, characterized in that, The method for constructing the nutritional risk sequence of the nutritional risk level includes: Query the level analysis time of the nutritional risk level; Arrange the nutritional risk levels in ascending order of the level analysis time to obtain a nutritional risk sequence.

9. The nutritional risk assessment and monitoring method for cancer palliative care patients according to claim 1, wherein The method for analyzing the nutritional risk trend of the cancer patient using the nutritional risk sequence includes: Determine the trend analysis model of the nutritional risk sequence; Use the bidirectional long short-term memory network in the trend analysis model to analyze the first sequence feature of the nutritional risk sequence; Use the long short-term memory network in the trend analysis model to analyze the second sequence feature of the first sequence feature; Use the time attention mechanism in the trend analysis model to analyze the attention weight of the second sequence feature; Use the weighted summation layer in the trend analysis model to perform weighted summation between the second sequence feature and the attention weight to obtain a weighted summation result; Analyze the nutritional risk trend of the cancer patient through the weighted summation result.

10. A nutritional risk assessment and monitoring system for cancer palliative care patients, characterized in that, The system includes: An equipment determination module for detecting the initial nutritional condition of a cancer patient and using the initial nutritional condition to determine the nutritional monitoring equipment of the cancer patient, where the nutritional monitoring equipment includes an image device and a non-image device; A food recording module for using the image device to obtain the daily diet image of the cancer patient, identifying the food category in the daily diet image, identifying the food weight in the daily diet image, analyzing the dining person's actions in the daily diet image, and recording the target food corresponding to the dining person's actions; A risk identification module, which is used to analyze the daily diet status of the cancer patient through the food category, the food weight, the actions of the diner, and the target food, determine the nutritional change items of the initial nutritional status by using the daily diet status, and identify the nutritional risk status of the cancer patient by using the nutritional change items; A grade analysis module, which is used to determine the target monitoring device of the cancer patient from the non-image devices according to the nutritional change items and the nutritional risk status, monitor the nutritional status of the cancer patient during a period through the target monitoring device, and analyze the nutritional risk grade corresponding to the nutritional status during the period; A risk monitoring module, which is used to construct a nutritional risk sequence of the nutritional risk grade, analyze the nutritional risk trend of the cancer patient by using the nutritional risk sequence, and determine the nutritional risk monitoring result of the cancer patient based on the nutritional risk trend.

Images