KR102526901B1 - Healthcare system - Google Patents

Abstract

According to an embodiment of the present invention, a method performed in one or more processors of a healthcare device is provided. The method may include acquiring an analysis target image including feces, which is an analysis target; analyzing the shape of the analysis target in the analysis target image; determining the type of the analyte based on the shape and stool classification chart; and inferring the user's health condition based on the type.

Description

Healthcare system {HEALTHCARE SYSTEM}

The present invention relates to a healthcare system.

BACKGROUND OF THE INVENTION [0002] In general, the food we ingest goes through a digestive process passing through the esophagus, stomach, small intestine, large intestine, etc., and the remaining material is discharged as feces through the anus. At this time, by examining the shape, color, size and amount of feces, the individual's health condition can be known. For example, if the color of the stool is maroon or red, gastrointestinal bleeding or hemorrhoids, intestinal disease, etc. may be suspected, and if the stool is green and diarrhea, gastric bacterial infection may be suspected. Also, hard and small grainy stools can be suspicious of a lack of fiber to flush waste out of the digestive system.

However, conventionally, in order to diagnose an individual's health condition using stool, each person has to take a stool sample of himself (or a patient) and submit it to a medical institution for examination, so the stool sample is transported to a medical institution. There is a possibility that the stool sample may be contaminated during the process or examination process, and there are problems such as a case where the stool sample is reversed while storing it in a medical institution. In addition, since it takes a certain amount of time for test results to come out after submitting a stool sample to a medical institution, there is a problem in that it is difficult to immediately diagnose an individual's health condition.

KR 10-2019-0030822 A KR 10-1595434 B1 KR 10-2018-0015858 A

The present invention is to solve the above problems, and an object of the present invention is to provide a health care system capable of analyzing a user's health condition.

The technical problems of the present invention are not limited to the technical problems mentioned above, and other technical problems not mentioned will be clearly understood by those skilled in the art from the description below.

According to an embodiment of the present invention, a method performed in one or more processors of a healthcare device is provided. The method may include acquiring an analysis target image including feces, which is an analysis target; analyzing the shape of the analysis target in the analysis target image; determining the type of the analyte based on the shape and stool classification chart; and inferring the user's health condition based on the type.

In addition, the analysis target image may be obtained from a camera mounted on the toilet.

In addition, the step of analyzing the shape of the analysis target may include performing primary edge detection to extract a region of interest in which the analysis target is located in the analysis target image; and performing secondary edge detection on the region of interest to analyze the shape of the analysis object.

In addition, the first edge detection may be performed based on a canny edge detection algorithm, and the second edge detection may be performed based on a local binary pattern (LBP).

In addition, the step of performing the secondary edge detection is performed by analyzing the boundary of the chunk to recognize the analysis target in units of chunks, and the chunk is a part of the stool, and at least one of the Chunks may cluster to form the analyte.

In addition, the analysis of the shape may include information on at least one of the size of chunks included in the analysis object, the number of chunks, the degree of cohesion of chunks, the shape of chunks, or whether chunks have been extracted.

In addition, the stool classification chart may be a Bristol stool chart.

According to an embodiment of the present invention, a computer program stored in a medium to perform a method performed by one or more processors of a healthcare device is provided.

A healthcare device is provided according to an embodiment of the present invention. The apparatus includes a memory for storing computer executable instructions; And a processor for executing instructions stored in the memory, wherein the processor acquires an analysis target image including stool, an analysis target, analyzes the shape of the analysis target in the analysis target image, and analyzes the shape and Based on the stool classification chart, the type of the object to be analyzed may be determined, and the user's health condition may be inferred based on the type.

According to the present invention, it is possible to provide a method for analyzing a user's health condition.

In order to more fully understand the drawings cited in the detailed description of the present invention, a brief description of each drawing is provided.
1 is a block diagram of a healthcare system according to an embodiment of the present invention.
2 is a block diagram of a healthcare device constituting a healthcare system according to an embodiment of the present invention.
3 is a schematic diagram showing an artificial neural network according to an embodiment of the present invention.
4 is a diagram illustrating a Bristol stool chart according to an embodiment of the present invention.

Hereinafter, exemplary embodiments according to the present invention will be described in detail with reference to the contents described in the accompanying drawings. In addition, a method of configuring and using the embodiment will be described in detail with reference to the contents described in the accompanying drawings. Like reference numerals or numerals in each drawing indicate parts or components that perform substantially the same function. For convenience described below, directions of up, down, left, and right are based on the drawings, and the scope of the present invention is not necessarily limited to the corresponding directions.

Terms including ordinal numbers, such as first and second, may be used to describe various elements, but elements are not limited by the terms. Terms are only used to distinguish one component from another. For example, a first element may be termed a second element, and similarly, a second element may be termed a first element, without departing from the scope of the present invention. The terms and/or include a combination of a plurality of related items or any one of a plurality of related items.

Terms used in this specification are used to describe embodiments, and are not intended to limit and/or limit the present invention. Singular expressions include plural expressions unless the context clearly dictates otherwise. In this specification, the terms include or have are intended to designate that a feature, number, step, operation, component, part, or combination thereof described in the specification exists, but one or more other features, numbers, or steps However, it should be understood that it does not preclude the possibility of existence or addition of operations, components, parts, or combinations thereof.

Throughout the specification, when a part is said to be connected to another part, this includes not only a case where it is directly connected but also a case where it is indirectly connected with another element interposed therebetween. In addition, when a part includes a certain component, this means that it may further include other components without excluding other components unless otherwise stated.

In addition, terms such as "...unit" and "module" described in the specification mean a unit that processes at least one function or operation, which may be implemented as hardware or software or a combination of hardware and software. .

In this specification, network functions, artificial neural networks, and neural networks may be used interchangeably.

1 is a block diagram of a healthcare system according to an embodiment of the present invention.

Referring to FIG. 1 , a healthcare system 1000 may include a measurement device 100 , a healthcare device 200 and a user terminal 300 .

The measuring device 100 may generate an analysis target image by photographing the analysis target. The healthcare device 200 may receive the image to be analyzed, analyze the user's health condition using the image, and transmit the result to the user terminal 300 .

In an embodiment, an analysis object for obtaining information on a user's health condition may include secretions derived from the user's body, and the secretions may be feces, for example. However, it is not limited thereto.

In an embodiment, the measuring device 100 may be provided at a location suitable for photographing the analysis target. For example, when the analysis target is a toilet, the measuring device 100 may be a part of the toilet or a bidet, and an analysis target image may be captured from a camera in the measuring device 100 .

In an embodiment, the healthcare device 200 may include a cloud server for collecting various data from the measurement device 100, but is not limited thereto, and may be a user end computer device. In addition, the healthcare device 200 may collect data from the measurement device 100 or may collect data from the measurement device 100 through the user terminal 300 .

The healthcare system 1000 shown in FIG. 1 is exemplary, and various configurations may be applied according to an embodiment to which the present invention is applied.

2 is a block diagram of a healthcare device constituting a healthcare system according to an embodiment of the present invention.

The healthcare device 200 may obtain an analysis target image, analyze the shape of the analysis target therefrom, and apply the result to an analysis model to infer the health state of the user. Here, the object to be analyzed may be feces, and the image to be analyzed may be taken from a camera mounted on the toilet bowl. However, it is not limited thereto.

In an embodiment, the healthcare device 200 may perform predetermined image processing on the analysis target image before analyzing the shape of the analysis target in the analysis target image. Image processing is applying changes to an analysis target image to be suitable for shape analysis, and may include, for example, color conversion, noise removal, image quality adjustment, etc., but is not limited thereto.

In an embodiment, the healthcare device 200 may analyze the shape of the object to be analyzed in the image to be analyzed. In this case, the analysis target may be stool, and the analysis target image may be an image of stool obtained from a flush toilet. Specifically, the healthcare device 200 may analyze the shape by obtaining external information of the analysis target from the analysis target image. The appearance information may include, for example, the size of feces, the number of chunks, the shape of chunks, the size of chunks, and the degree of cohesion of chunks. However, it is not limited thereto, and various information for identifying and classifying feces may be included.

In an embodiment, the healthcare device 200 may accompany edge detection to analysis of the shape of the object to be analyzed. For example, since feces, which is an analysis target, is located under the water surface, light reflection by the water surface and color of the water surface affect the analysis target image. In particular, the feces are not fixed in position in the toilet, but may be moved or aligned. Also, for example, when feces are composed of a plurality of chunks, the chunks are clustered in one place, making it difficult to grasp the overall shape of each chunk or feces. Therefore, the healthcare device 200 performs one or more stages of edge detection to make the boundaries of objects included in the analysis target image clearer and clearer and/or remove/minimize unnecessary objects for analysis, thereby increasing analysis reliability. there is. For example, edge detection may be performed based on a canny edge detection algorithm, a local binary pattern (LBP), or the like. In particular, in the case of LBP, it is suitable because the boundary and texture between chunks can be grasped more clearly.

In an embodiment, the healthcare device 200 may extract an analysis target from an analysis target image through edge detection. That is, a region of interest, which is a predetermined region including the analysis object, may be extracted from the analysis object image. For such edge detection, an edge detection algorithm that has relatively little computational capacity or is easy to distinguish between feces and areas other than feces may be used. For example, a Kenny edge detection algorithm may be applied. Once the region of interest has been extracted, a more in-depth analysis can be performed on it.

In an embodiment, the healthcare device 200 may perform first edge detection and second edge detection. The primary edge detection is for extracting a region of interest where the analysis target is located in the analysis target image, and the secondary edge detection is applied to the region of interest to determine the shape of the analysis target. By extracting a portion to be effectively analyzed from the analysis target image through the extraction of the region of interest, the amount of computation for analyzing the analysis target can be reduced. In addition, for the secondary edge detection, a more effective algorithm may be applied to grasp the shape of the object to be analyzed. For example, secondary edge detection may be used to read the texture of an analysis object. For example, the first edge detection may be based on the Kenny edge detection algorithm, and the second edge detection may be based on LBP.

In an embodiment, the healthcare device 200 may analyze a chunk boundary of an analysis object to recognize the analysis object in units of chunks. A chunk is a part of an analysis target (ie, feces), and at least one chunk may form an analysis target by clustering. At this time, each chunk may have a certain shape based on a person's health condition, intestinal lactic acid bacteria condition, moisture condition, and the like. Boundary analysis for chunk recognition may be performed through, for example, edge detection such as LBP. In this case, the edge detection may be secondary edge detection, but is not limited thereto.

In an embodiment, the healthcare device 200 may determine the type of the analyte based on the shape of the analyte. Such classification may be based on a predetermined stool classification chart. The feces classification chart includes the form of feces and identification information and characteristics thereof. For example, the feces classification chart may be a Bristol feces chart. In the case of the Bristol stool chart, the healthcare device 200 may classify the analysis object to which one of the first to seventh types of the Bristol stool chart belongs to the analysis object based on the shape of the analysis object. In the following, examples of the present invention will be described based on the Bristol stool chart for convenience of explanation, but are not limited thereto.

In an embodiment, the healthcare device 200 may allocate an analysis object to a classification candidate group based on the type of the analysis object before finally classifying the type of the analysis object. For example, the first classification candidate group may include classifications for types that can be classified relatively easily, and the second classification candidate group may include classifications for types that are difficult to classify easily.

For example, if they belong to the first type, the fifth type, the sixth type, and the seventh type of the Bristol stool chart, they may be classified as the first classification candidate group. In addition, if it does not belong to the first classification candidate group or if it belongs to the second to fourth types, it may be classified as the second classification candidate group.

If they belong to the first classification candidate group (first type, fifth type, sixth type, and seventh type), chunks are relatively clearly distinguished and recognized, and the healthcare device 200 can easily classify them. On the other hand, if they belong to the second classification candidate group (the second to the fourth type), the second type, the third type, and the fourth type are classified according to the texture of the large lump, so they have different characteristics from the first classification candidate group. , additional analysis of the texture within the mass may be required.

In an embodiment, the healthcare device 200 determines the shape of the object to be analyzed through edge detection, and determines whether the object to be analyzed is a whole lump, composed of a plurality of lumps, or a case where the lump cannot be read. can Based on this, it is possible to determine whether feces, which are analysis objects, belong to the first classification candidate group or the second classification candidate group. That is, when one large lump is set as the region of interest, the healthcare device 200 may estimate that the object to be analyzed belongs to the second classification candidate group. In addition, when a partial pattern of the first classification candidate group is searched for (eg, a region of interest is most of an image), the healthcare device 200 may estimate that the analysis object belongs to the first classification candidate group. .

In an embodiment, the healthcare device 200 determines whether at least some of the analysis objects belong to the first type, the fifth type, the sixth type, and the seventh type when the analysis object belongs to the first classification candidate group, and However, if it belongs to each type, it is possible to classify the analyte into that type.

In an embodiment, the healthcare device 200 may classify whether the analysis object belongs to which type of the second classification candidate group by analyzing the texture of the analysis object belonging to the second classification candidate group through secondary edge detection. . That is, when the type of object to be analyzed belongs to the second classification candidate group, the healthcare device 200 may perform edge detection to classify the object into one of the second type, the third type, and the fourth type, corresponding to each type. It is also possible to output the analysis result in the form of probability.

In an embodiment, the healthcare device 200 may infer the user's health condition by analyzing the analysis object. Depending on the embodiment, such inference may be accompanied by the above-described classification of the analysis target. Specifically, the healthcare device 200 analyzes the shape of the analysis object, classifies the type of analysis object based on this, and then infers the user's health condition based on the characteristics of the type, the user's physical condition, and the like. .

In an embodiment, the healthcare device 200 may use an analysis model in analyzing the shape of an analysis object, classifying the type of analysis object, and/or inferring a health condition. In this case, the analysis model may be based on an analysis model learned through deep learning based on a neural network. However, it is not limited thereto.

In an embodiment, the analysis model may be a model trained to take an analysis target image as an input and output a type of stool classification chart (eg, Bristol stool chart). The analysis model may include, for example, an artificial neural network model capable of image processing. The analysis model may be a model learned to apply an analysis method differently according to classification candidates. The analysis model may be a model learned to process input data with different submodels based on classification candidates. In addition, the analysis model may be a model learned to output a type of feces classification chart by inputting information about classification candidates together with analysis data.

In an embodiment, the healthcare device 200 may process the analysis target image differently based on the classification candidate group in processing the analysis model. For example, the healthcare device 200 may input an analysis target image and a classification candidate group together into an analysis model, and the healthcare device 200 may process the analysis target image based on the classification candidate group as a sub-model of the analysis model. can also be selected.

More specifically, the analysis model may be trained to analyze the entire region of the analysis object when the classification candidate group corresponds to, for example, the second classification candidate group of the Bristol Stool Chart. That is, when the classification candidate group corresponds to the second classification candidate group, the analysis model may be trained to read the entire region of the analysis object and draw a final conclusion.

In addition, the analysis model determines whether the classification candidate group corresponds to, for example, the first classification candidate group), whether the shape of stool included in some of the areas of the analysis target is classified into any type of the type of Bristol stool chart. can be learned That is, the analysis model makes a corresponding conclusion when the classification candidate group corresponds to the first classification candidate group and when at least some of the regions of the analysis target are classified into one of the first type, fifth to seventh type of the Bristol Stool Chart. It can be learned to make final conclusions.

In an embodiment, the healthcare device 200 may determine the type of feces included in the analysis target image based on the edge detection result, and feces based on the edge detection result and the classification result of the analysis model based on the edge detection result. type can be determined. The healthcare device 200 first tentatively classifies the type of feces based on the edge detection result, and secondarily reclassifies the type of feces using a classification model accordingly, thereby re-verifying the classification result of the primary classification You can increase the accuracy by doing this.

In an embodiment, the healthcare device 200 may generate inferred information related to the user's health condition based on information related to the type on the stool classification chart and information related to the color of stool. The healthcare device 200 may generate information related to the user's nutritional status, intestinal health status, and the like, based on analysis information on the user's feces.

In an embodiment, the healthcare device 200 may firstly generate information related to the type of feces and secondarily generate information related to the color of the feces. In order to determine the color of feces, a Hue Saturation Value (HSV) model may be used, but the present embodiment is not limited thereto. The healthcare device 200 may determine the user's intestinal health status based on information related to the type and color of stool. Depending on the color of the stool, in the case of red bloody stool, the user may have enteritis, hemorrhoids, fissures, anal lacerations, congenital abnormalities, or intestinal obstruction. If the color of stool is gray, the user may have a liver or biliary obstruction disease. If the color of the stool is dark yellow, the user may be in a state in which bile pigments are excreted. If the color of the feces is tan cleft, the user may be in a normal state. If the color of the stool is green, the user may be in a state of indigestion, excessive intake of green vegetables, or unabsorbed iron excretion in the body. If the color of the stool is black, it may be a condition such as esophageal/gastric/duodenal bleeding or inflammatory bowel disease.

In an embodiment, the healthcare device 200 may receive user biometric information from the measurement device 100 and additionally generate user health state information based on the user biometric information. For example, the toilet may collect various biometric information of the user from the user's skin in contact with the toilet seat. For example, the toilet may include a biocurrent collection module in the toilet seat, and the toilet collects the user's biocurrent through this to allow the healthcare device 200 to determine whether there is a health risk to the user. . In addition, the healthcare device 200 may generate integrated health condition information including the user's body composition and heartbeat-related information based on the user's biometric information collected from the toilet.

In an embodiment, the healthcare device 200 may deliver a notification to the user when the user's health condition is suspected. Also, the healthcare device 200 may determine the user's health abnormality based on the user's basic biometric information stored in the healthcare device 200 . For example, the healthcare device 200 may diagnose the user's disease based on the user's basic biometric information (eg, health examination data) and stool classification results, and notify the user of the diagnosis. That is, when the green stool is classified as the second type, the health care device 200 combines this information and the user's health examination data to excrete unabsorbed iron in the body when the user's condition is not a simple excessive intake of green vegetables. It can also be judged as a state. The above description is only an example and the present embodiment is not limited thereto.

As shown in FIG. 2 , the healthcare device 200 may include a processor 210 , a memory 230 , and a network unit 250 .

The processor 210 may control overall operations of the healthcare device 200 . For example, the processor 210 may overall control components within the healthcare device 200 by executing a program stored in the healthcare device 200 . For example, the processor 210 may be configured to process at least one instruction by performing basic arithmetic, logic, and input/output operations for the operation of the healthcare device 200 . The instructions described above may be provided to processor 210 from memory 230 . That is, processor 210 may be configured to execute instructions according to program codes stored in a recording device such as memory 230 . Alternatively, the instruction may be received by the healthcare device 200 through the network unit 250 and provided to the processor 210 . In addition, the processor 210 may include one or more cores, including a central processing unit (CPU), a general purpose graphics processing unit (GPGPU), and a tensor processing unit (TPU). ), etc., and a processor 210 for data analysis and deep learning. The processor 210 may read the computer program stored in the memory 230 and perform data processing for machine learning according to an embodiment. According to an embodiment, the processor 210 may perform an operation for learning the neural network. The processor 210 is used for neural network learning, such as processing input data for learning in deep learning (DL), extracting features from input data, calculating errors, and updating neural network weights using backpropagation. calculations can be performed. At least one of the CPU, GPGPU, and TPU of the processor 210 may process learning of the network function. For example, the CPU and GPGPU can process learning of network functions and data classification using network functions. Also, in an embodiment, the processors 210 of the plurality of healthcare devices 200 may be used together to process learning of the network function and data classification using the network function. In addition, the computer program executed in the healthcare device 200 according to the embodiment may be a CPU, GPGPU or TPU executable program.

In an embodiment, the memory 230 may store any type of information generated or determined by the processor 210 and any type of information received by the network unit 250 .

In an embodiment, the memory 230 is a flash memory type, a hard disk type, a multimedia card micro type, or a card type memory 230 (eg For example, SD or XD memory (230), etc.), RAM (Random Access Memory, RAM), SRAM (Static Random Access Memory), ROM (Read-Only Memory, ROM), EEPROM (Electrically Erasable Programmable Read-Only Memory), It may include at least one type of storage medium among a programmable read-only memory (PROM), a magnetic memory 230, a magnetic disk, and an optical disk. The healthcare device 200 may operate in relation to a web storage performing a storage function of the memory 230 on the Internet. The above description of the memory 230 is only an example, and the present embodiment is not limited thereto.

In an embodiment, the network unit 250 according to the Public Switched Telephone Network (PSTN), x Digital Subscriber Line (xDSL), Rate Adaptive DSL (RADSL), Multi Rate DSL (MDSL), Very High Speed (VDSL) DSL), UADSL (Universal Asymmetric DSL), HDSL (High Bit Rate DSL), and various wired communication systems such as a local area network (LAN) may be used.

In an embodiment, the network unit 250 may include Code Division Multi Access (CDMA), Time Division Multi Access (TDMA), Frequency Division Multi Access (FDMA), Orthogonal Frequency Division Multi Access (OFDMA), and Single Carrier-FDMA (SC-FDMA). FDMA) and other systems may be used.

In an embodiment, the network unit 250 may be configured regardless of its communication type, such as wired and wireless, and is composed of various communication networks such as a personal area network (PAN) and a wide area network (WAN). It can be. In addition, the network may be the known World Wide Web (WWW), or may use a wireless transmission technology used for short-range communication, such as Infrared Data Association (IrDA) or Bluetooth.

The configuration of the healthcare device 200 shown in FIG. 2 is only a simplified example. In the embodiment, the healthcare device 200 may include other components for performing a computing environment of the healthcare device 200, and only some of the disclosed components may constitute the healthcare device 200.

3 is a schematic diagram showing an artificial neural network according to an embodiment of the present invention.

A computational model, a neural network, a network function, and a neural network may be used interchangeably. A neural network may consist of a set of interconnected computational units, which may generally be referred to as nodes. These nodes may also be referred to as neurons. A neural network includes one or more nodes. Nodes (or neurons) constituting neural networks may be interconnected by one or more links.

In a neural network, one or more nodes connected through a link may form a relative relationship of an input node and an output node. The concept of an input node and an output node is relative, and any node in an output node relationship with one node may have an input node relationship with another node, and vice versa. As described above, an input node to output node relationship may be created around a link. More than one output node can be connected to one input node through a link, and vice versa.

In a relationship between an input node and an output node connected through one link, the value of data of the output node may be determined based on data input to the input node. Here, a link interconnecting an input node and an output node may have a weight. The weight may be variable, and may be changed by a user or an algorithm in order to perform a function desired by the neural network. For example, when one or more input nodes are interconnected by respective links to one output node, the output node is set to a link corresponding to values input to input nodes connected to the output node and respective input nodes. An output node value may be determined based on the weight.

As described above, in the neural network, one or more nodes are interconnected through one or more links to form an input node and output node relationship in the neural network. Characteristics of the neural network may be determined according to the number of nodes and links in the neural network, an association between the nodes and links, and a weight value assigned to each link. For example, when there are two neural networks having the same number of nodes and links and different weight values of the links, the two neural networks may be recognized as different from each other.

A neural network may be composed of a set of one or more nodes. A subset of nodes constituting a neural network may constitute a layer. Some of the nodes constituting the neural network may form one layer based on distances from the first input node. For example, a set of nodes having a distance of n from the first input node may constitute n layers. The distance from the first input node may be defined by the minimum number of links that must be passed through to reach the corresponding node from the first input node. However, the definition of such a layer is arbitrary for explanation, and the order of a layer in a neural network may be defined in a method different from the above. For example, a layer of nodes may be defined by a distance from a final output node.

An initial input node may refer to one or more nodes to which data is directly input without going through a link in relation to other nodes among nodes in the neural network. Alternatively, in a relationship between nodes based on a link in a neural network, it may mean nodes that do not have other input nodes connected by a link. Similarly, the final output node may refer to one or more nodes that do not have an output node in relation to other nodes among nodes in the neural network. Also, the hidden node may refer to nodes constituting the neural network other than the first input node and the last output node.

The neural network according to the embodiment may have the same number of nodes in the input layer as the number of nodes in the output layer, and may be a neural network in which the number of nodes decreases and then increases as the number of nodes progresses from the input layer to the hidden layer. In addition, the neural network according to another embodiment of the present embodiment may be a neural network in which the number of nodes of the input layer may be less than the number of nodes of the output layer, and the number of nodes decreases as the number of nodes increases from the input layer to the hidden layer. . In addition, the neural network according to another embodiment of the present embodiment may be a neural network in which the number of nodes of the input layer may be greater than the number of nodes of the output layer, and the number of nodes increases as the number of nodes increases from the input layer to the hidden layer. there is. A neural network according to another embodiment of the present embodiment may be a neural network in the form of a combination of the aforementioned neural networks.

A deep neural network (DNN) may refer to a neural network including a plurality of hidden layers in addition to an input layer and an output layer. Deep neural networks can reveal latent structures in data. In other words, it can identify the latent structure of a photo, text, video, sound, or music (e.g., what objects are in the photo, what the content and emotion of the text are, what the content and emotion of the audio are, etc.). . Deep neural networks include convolutional neural networks (CNNs), recurrent neural networks (RNNs), auto encoders, generative adversarial networks (GANs), and restricted boltzmann machines (RBMs). machine), a deep belief network (DBN), a Q network, a U network, a Siamese network, a Generative Adversarial Network (GAN), and the like. The description of the deep neural network described above is only an example, and the present embodiment is not limited thereto.

In an embodiment, the network function may include an autoencoder. An autoencoder may be a type of artificial neural network for outputting output data similar to input data. An auto-encoder may include at least one hidden layer, and an odd number of hidden layers may be disposed between input and output layers. The number of nodes of each layer may be reduced from the number of nodes of the input layer to an intermediate layer called the bottleneck layer (encoding), and then expanded symmetrically with the reduction from the bottleneck layer to the output layer (symmetrical to the input layer). Autoencoders can perform non-linear dimensionality reduction. The number of input layers and output layers may correspond to dimensions after preprocessing of input data. In the auto-encoder structure, the number of hidden layer nodes included in the encoder may decrease as the distance from the input layer increases. If the number of nodes in the bottleneck layer (the layer with the fewest nodes located between the encoder and decoder) is too small, a sufficient amount of information may not be conveyed, so more than a certain number (e.g., more than half of the input layer, etc.) ) may be maintained.

The neural network may be trained using at least one of supervised learning, unsupervised learning, semi-supervised learning, and reinforcement learning. Learning of the neural network may be a process of applying knowledge for the neural network to perform a specific operation to the neural network.

A neural network can be trained in a way that minimizes output errors. In the learning of the neural network, the learning data is repeatedly input into the neural network, the output of the neural network for the training data and the error of the target are calculated, and the error of the neural network is transferred from the output layer of the neural network to the input layer in the direction of reducing the error. It is a process of updating the weight of each node of the neural network by backpropagating in the same direction. In the case of teacher learning, the learning data in which the correct answer is labeled is used for each learning data (ie, the labeled learning data), and in the case of comparative teacher learning, the correct answer may not be labeled in each learning data. That is, for example, learning data in the case of teacher learning about data classification may be data in which each learning data is labeled with a category. Labeled training data is input to a neural network, and an error may be calculated by comparing an output (category) of the neural network and a label of the training data. As another example, in the case of comparative history learning for data classification, an error may be calculated by comparing input learning data with a neural network output. The calculated error is back-propagated in a reverse direction (ie, from the output layer to the input layer) in the neural network, and the connection weight of each node of each layer of the neural network may be updated according to the back-propagation. The amount of change in the connection weight of each updated node may be determined according to a learning rate. The neural network's computation of input data and backpropagation of errors can constitute a learning cycle (epoch). The learning rate may be applied differently according to the number of iterations of the learning cycle of the neural network. For example, a high learning rate may be used in the early stage of neural network training to increase efficiency by allowing the neural network to quickly obtain a certain level of performance, and a low learning rate may be used in the late stage to increase accuracy.

In neural network learning, generally, training data can be a subset of real data (ie, data to be processed using the trained neural network), and therefore, errors for training data are reduced, but errors for real data are reduced. There may be incremental learning cycles. Overfitting is a phenomenon in which errors for actual data increase due to excessive learning on training data. For example, a phenomenon in which a neural network that has learned a cat by showing a yellow cat does not recognize that it is a cat when it sees a cat other than yellow may be a type of overfitting. Overfitting can act as a cause of increasing the error of machine learning algorithms. Various optimization methods can be used to prevent such overfitting. To prevent overfitting, methods such as increasing the training data, regularization, inactivating some nodes in the network during learning, and using a batch normalization layer should be applied. can

4 is a diagram illustrating a Bristol stool chart according to an embodiment of the present invention.

Referring to FIG. 4, based on the Bristol Stool Chart, the separated chunks, such as the first type of peanuts, may correspond to chunks. In case of the first type, the user may be severely constipated. Also, in the shape of the second type of sausage, lumps may correspond to chunks. That is, in the case of the second type, although it is one chunk as a whole, a plurality of chunks may be included in one chunk. In case of the second type, the user may be mildly constipated. Even in the case of the third type, chunks can be distinguished by surface cracks. That is, although the degree of classification is lower than that of the second type, chunks may be included even in the third type. In case of the third type, the user may be in a normal state. In the case of the fourth type, it may be determined as one chunk and it may be difficult to extract the chunk. In case of the fourth type, the user may be in a normal state. Even in the case of the fifth type, each soft lump may correspond to a chunk. In the case of the fifth type, the user may have a possibility of fiber deficiency. The sixth type may be a case composed of a plurality of smaller chunks. A sixth type may be a colitis condition. The seventh type is difficult to extract chunks as a complete liquid, and can be judged as a whole shape. In the case of the seventh type, the user may have severe colitis, a bad health condition such as occurrence of cancer tissue in the large intestine/rectum, and an imbalance of effects.

The devices described above may be implemented as hardware components, software components, and/or a combination of hardware components and software components. For example, devices and components described in the embodiments may include, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable gate array (FPGA) , a programmable logic unit (PLU), microprocessor, or any other device capable of executing and responding to instructions. The processing device may run an operating system (OS) and one or more software applications running on the operating system. A processing device may also access, store, manipulate, process, and generate data in response to execution of software. For convenience of understanding, there are cases in which one processing device is used, but those skilled in the art will understand that the processing device includes a plurality of processing elements and/or a plurality of types of processing elements. It can be seen that it can include. For example, a processing device may include a plurality of processors or a processor and a controller. Other processing configurations are also possible, such as parallel processors.

Software may include a computer program, code, instructions, or a combination of one or more of the foregoing, which configures a processing device to operate as desired or processes independently or collectively. You can command the device. Software and/or data may be any tangible machine, component, physical device, virtual equipment, computer storage medium or device, intended to be interpreted by or provide instructions or data to a processing device. , or may be permanently or temporarily embodied in a transmitted signal wave. Software may be distributed on networked computer systems and stored or executed in a distributed manner. Software and data may be stored on one or more computer readable media.

The method according to the embodiment may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer readable medium. Computer readable media may include program instructions, data files, data structures, etc. alone or in combination. Program commands recorded on the medium may be specially designed and configured for the embodiment or may be known and usable to those skilled in computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks and magnetic tapes, optical media such as CD-ROMs and DVDs, and magnetic media such as floptical disks. - includes hardware devices specially configured to store and execute program instructions, such as magneto-optical media, and ROM, RAM, flash memory, and the like. Examples of program instructions include high-level language codes that can be executed by a computer using an interpreter, as well as machine language codes such as those produced by a compiler. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the embodiments, and vice versa.

As described above, although the embodiments have been described with limited examples and drawings, those skilled in the art can make various modifications and variations from the above description. For example, the described techniques may be performed in an order different from the method described, and/or components of the described system, structure, device, circuit, etc. may be combined or combined in a different form than the method described, or other components may be used. Or even if it is replaced or substituted by equivalents, appropriate results can be achieved. In addition, each embodiment can be operated in combination with each other as needed. For example, an operation of a device may be operated by combining parts of an embodiment and other embodiments. Therefore, other implementations, other embodiments, and equivalents of the claims are within the scope of the following claims.

Claims (9)

A health condition analysis method performed in one or more processors of a healthcare device,
Acquiring an analysis target image containing feces as an analysis target;
analyzing the shape of the analysis target in the analysis target image;
determining the type of the analyte based on the shape and stool classification chart; and
Inferring the user's health condition based on the type;
Analyzing the shape of the object of analysis may include performing primary edge detection to extract a region of interest in which the object of analysis is located in the image of the object of analysis; and performing secondary edge detection on the region of interest to analyze the shape of the analysis target,
The step of performing the secondary edge detection is performed by analyzing the boundary of the chunk to recognize the analysis object in units of chunks, and the chunk is a part of the feces, and at least one of the chunks clustering to form the analyte. According to claim 1,
The method of claim 1, wherein the image to be analyzed is obtained from a camera mounted on a toilet bowl. delete According to claim 1,
The first edge detection is performed based on a canny edge detection algorithm, and the second edge detection is performed based on a local binary pattern (LBP). delete According to claim 1,
The analysis of the shape includes information on at least one of the size of chunks included in the analysis object, the number of chunks, the degree of cohesion of chunks, the shape of chunks, or whether chunks are extracted. According to claim 1,
The method of claim 1, wherein the stool classification chart is a Bristol stool chart. As a healthcare device,
memory for storing computer executable instructions; and
a processor for executing instructions stored in the memory;
The processor acquires an analysis target image including feces as an analysis target, analyzes the shape of the analysis target in the analysis target image, and determines the type of the analysis target based on the shape and the stool classification chart, Inferring the health status of the user based on the type;
The processor performs primary edge detection to extract a region of interest where the analysis target is located in the analysis target image, and performs secondary edge detection on the region of interest to analyze the shape of the analysis target, , Analyzing the shape of the analyte,
The secondary edge detection of the processor is performed by analyzing the boundary of the chunk to recognize the analysis target in units of chunks,
wherein the chunk is part of the feces, and at least one of the chunks clusters to form the analyte. A computer program stored in a medium to perform the method according to any one of claims 1, 2, 4, 6 and 7.

Images