KR102597000B1 - Device for asking and diagnosing a patient for surgery, and a method for collecting patient status information through the device - Google Patents

Abstract

A paper weight device is provided. The medical examination device includes a question unit that outputs a question that induces a voice answer from the patient; a collection unit that collects the voice response; It may include a management unit that manages the voice response by matching it with the patient's identification information.

Description

Device for asking and diagnosing a patient for surgery, and a method for collecting patient status information through the device}

The present invention relates to an apparatus and method for interviewing a surgical patient and collecting information on the patient's condition.

When a patient is hospitalized for surgery, doctors make rounds every day to check the patient's condition.

Because a large number of patients must be checked, the time allotted for interviewing one patient during rounds decreases, and this may reduce the quality of the interview.

Korean Patent Publication No. 1584308 shows a smart electronic medical examination system that can collect medical examination data through a mobile terminal and use it for customized treatment for each patient. However, since a preset questionnaire is provided to the patient terminal, there is a limitation in that only short-answer answers can be obtained from the patient.

Korean Patent Publication No. 1584308

The present invention is intended to provide a device that automatically performs a medical examination on patients admitted to a hospital for surgery, etc., and a method of collecting information on the patient's condition through the same.

The medical examination device of the present invention includes a question unit that outputs a question that induces a voice answer from the patient; a collection unit that collects the voice response; a conversion unit that converts the voice response into text; an extraction unit that extracts status information of the patient from the voice response or the text; a recording unit for recording the voice response; It may include a management unit that matches and manages at least one of the text, the status information, and the recording file of the voice response to the patient's identification information.

The collection method of the present invention includes transmitting a list of interview subjects from a server to a questionnaire robot; moving the medical examination robot to target a specific patient included in the list; The medical examination robot identifying identification information of the specific patient and transmitting the identified identification information to the server; Searching the server for whether existing questionnaire data exists targeting the identification information of the specific patient; transmitting a first question to the medical examination robot if the existing medical examination data does not exist in the server, and transmitting a second question to the medical examination robot if the existing medical examination data exists; Converting the first question or the second question into voice and outputting it in the medical examination robot; Collecting voice answers from the specific patient from the medical examination robot and transmitting the voice answers to the server; Determining whether the voice answer is correct at the server and transmitting the answer recognition result including the judgment result and the next question to the questionnaire robot; storing the answer recognition result in the medical examination robot and outputting the next question targeting the specific patient; When the medical examination of the specific patient by the medical examination robot is completed, transmitting the answer recognition results and recorded data of the voice answer stored in the medical examination robot to the server; It may include databaseizing the answer recognition result and the voice answer in the server.

According to the medical examination device and collection method of the present invention, a medical examination dialogue system that automatically collects the patient's condition after surgery can be provided. Since the medical examinations that were previously performed by medical staff are automated, the burden of medical examinations on medical staff can be reduced.

According to the present invention, data on patient progress within the hospital can be efficiently managed by creating a database of pre- and post-surgical progress.

The medical examination device and collection method of the present invention may not make any judgment about the condition based on the patient's answers. The questionnaire device and collection method of the present invention can only perform the function of accurately obtaining answers from the patient and delivering the obtained answers as is to the medical staff.

1 is a schematic diagram showing the paperweight device of the present invention.
Figure 2 is a schematic diagram showing the actions of the question order variable algorithm.
Figure 3 is a flowchart showing the method of collecting patient condition information according to the present invention.
4 is a diagram illustrating a computing device according to an embodiment of the present invention.

Below, with reference to the attached drawings, embodiments of the present invention will be described in detail so that those skilled in the art can easily implement the present invention. However, the present invention may be implemented in many different forms and is not limited to the embodiments described herein. In order to clearly explain the present invention in the drawings, parts that are not related to the description are omitted, and similar parts are given similar reference numerals throughout the specification.

In this specification, duplicate descriptions of the same components are omitted.

Also, in this specification, when a component is mentioned as being 'connected' or 'connected' to another component, it may be directly connected or connected to the other component, but may be connected to the other component in the middle. It should be understood that may exist. On the other hand, in this specification, when it is mentioned that a component is 'directly connected' or 'directly connected' to another component, it should be understood that there are no other components in between.

Additionally, the terms used in this specification are merely used to describe specific embodiments and are not intended to limit the present invention.

Also, in this specification, singular expressions may include plural expressions, unless the context clearly dictates otherwise.

In addition, in this specification, terms such as 'include' or 'have' are only intended to designate the presence of features, numbers, steps, operations, components, parts, or combinations thereof described in the specification, and one or more It should be understood that this does not exclude in advance the presence or addition of other features, numbers, steps, operations, components, parts, or combinations thereof.

Also, in this specification, the term 'and/or' includes a combination of a plurality of listed items or any of the plurality of listed items. In this specification, 'A or B' may include 'A', 'B', or 'both A and B'.

Additionally, in this specification, detailed descriptions of well-known functions and configurations that may obscure the gist of the present invention will be omitted.

For patients after surgery, doctors may make rounds daily or hourly to check the progress of the surgery. At this time, most of the rounds for patients after surgery can be comprised of a simple process of checking for abnormalities in the surgical area.

It is clear that if this simple process is automated, the burden on medical staff will be greatly reduced. The medical examination device and collection method of the present invention can automate rounds performed on patients after surgery and general medical examinations on general patients. The paperweight device of the present invention may be implemented in the form of a robot. The robot can collect information about the patients' condition and deliver the collected condition information to the medical staff. It may be sufficient for medical staff to directly conduct questionnaires only on patients who have unusual conditions among the status information received from the robot. Therefore, according to the medical examination device and collection method of the present invention, the medical staff's time spent on rounds or medical examinations can be significantly reduced.

In fields that deal with human life and health, it may still be premature for robots to determine conditions on behalf of medical staff due to reliability and risk in actual medical settings. Therefore, the medical examination device and collection method of the present invention can focus solely on accurately collecting condition information from the patient and delivering it to the medical staff as is, without making any judgment about the condition.

1 is a schematic diagram showing the paperweight device of the present invention.

The medical examination device shown in FIG. 1 may include a medical examination robot 10 and a server 30.

The medical examination robot 10 can ask questions to the patient and collect the patient's voice answers to the questions. In order to ask questions by voice and collect answers, the medical examination robot 10 can be spatially placed within a few meters of the patient being questioned.

As an example, the medical examination robot 10 may be formed to have a moving means that travels along the ground and directly visits the patient. Alternatively, the medical examination robot 10 may be formed in the form of a wearable device operated by the patient. Alternatively, the medical examination robot 10 may be installed in a hospital room used by a patient or may be installed in the form of a speaker and microphone installed around the bed.

The server 30 extracts questions to be asked for a specific patient using the identification information and delivers them to the interview robot 10, records the voice answers collected by the interview robot 10 in the database 290, and delivers them to the medical staff. there is.

The paperweight device of the present invention internally includes a questioning unit 110, a collection unit 120, a conversion unit 220, an extraction unit 230, a recording unit 140, a management unit 130, a moving unit 150, It may include a determination unit 240 and a processing unit 210. At this time, some elements may be installed in the paperweight robot 10, and other elements may be installed in the server 30. The installation location of some elements may vary depending on various environments.

As an example, the questioning unit 110 and the collection unit 120 that function at a position facing the patient may be provided in the medical examination robot 10. The processing unit 210 that performs a database conversion process of recording voice answers obtained from a plurality of patients in the database 290 is preferably provided in the server 30 that can communicate with a plurality of medical examination robots 10. On the other hand, the conversion unit 220, extraction unit 230, recording unit 140, management unit 130, moving unit 150, and judgment unit 240, which can function even in positions not facing the patient, are used for medical examination. It may be provided in the robot 10 or in the server 30.

In Figure 1, the questionnaire unit 110, the collection unit 120, the recording unit 140, the moving unit 150, and the management unit 130 are provided in the questionnaire robot 10. In FIG. 1, a processing unit 210, a conversion unit 220, an extraction unit 230, and a determination unit 240 are provided in the server 30. At this time, it is reminded once again that the conversion unit 220, the extraction unit 230, etc. may be provided in the paperweight robot 10.

The question unit 110 may output a question that induces a voice answer from the patient. The question unit 110 may include a speaker that audibly outputs various questions or a display that visually outputs various questions.

The question unit 110 is preferably placed within a few meters from the user so that the patient can directly recognize the question content.

The collection unit 120 may collect the patient's voice response. The collection unit 120 may include a microphone that receives voice responses.

The recording unit 140 may record the patient's voice responses collected through the collection unit 120 to create a recording file.

The conversion unit 220 can convert the patient's voice response into text.

The extraction unit 230 may extract patient status information from a voice response or text.

The patient's condition information may be as follows, including the start time of the interview 'Date: 2020-11-16 11:58:55'. The patient's condition information is stored in the database 290 by the processing unit 210, and the patient's identification information, index, etc. may be added to enable search and extraction.

-----Patient's condition information-----

Date: 2020-11-16 11:58:55

Position: ['Knee']

Quality: ['throbbing']

R_factor2: ['When walking', 'When sitting']

Severity: 4

Timing2: 1 year

Mood: Normal

Anxiety: Yes

Sleep1: 1

Sleep2: bad

-----------------------

The patient's condition information may include check items that the medical staff checks through questionnaires during rounds.

Questions in the questionnaire unit 110 may be generated to induce a voice response from the patient to fill out the corresponding check items.

As an example, the first question, which includes questions for pre-operative patients about to undergo surgery, may include the following 10 questions.

1. Where is the most painful area right now? If there are multiple areas, please list them in order of pain. (Pain area: Question to fill in ‘Position’ of status information)

2. What kind of pain do you feel? Please express it as numb or throbbing. (Pain pattern: Questions to fill in the ‘Quality’ of status information)

3. When is the pain the most severe? (When pain intensifies)

4. What position causes the most pain? (Pain intensification posture: Questions to fill in ‘R_factor2’ of status information)

5. Please rate how severe the pain is between 0 and 10. (Pain intensity: question to fill in ‘Severity’ of status information)

6. How many days did the pain get worse? (Pain duration: question to fill in ‘Timing2’ of status information)

7. How are you feeling now? Please say good, average, or bad. (Current mood; question to fill in ‘Mood’ of status information)

8. Are you currently worried or anxious? Please answer yes or no. (Anxiety: Questions to fill in ‘Anxiety’ in status information)

9. How many hours did you sleep? (Sleep time: Question to fill ‘Sleep1’ in status information)

10. Did you sleep well without waking up in the middle of the night? (Sleep quality: Questions to fill in ‘Sleep2’ of status information)

As another example, the second questionnaire, which includes questions for patients after surgery, may include the following 10 questions.

1. Do you feel pain at the surgical site? (Pain present or not)

2. What kind of pain do you feel at the surgery site? Please express it as numb or throbbing. (Pain pattern)

3. Please rate how severe the pain is in the surgical area between 0 and 10. (pain intensity)

4. How many hours a day does the pain at the surgical site last? (Pain Duration)

5. Do you feel pain in the places that were hurting before the surgery? (Pain present or not)

6. What kind of pain do you feel? Please express it as numb or throbbing. (Pain pattern)

7. Please rate from 0 to 10 how severe the pain is in the area that hurt before surgery. (pain intensity)

8. How many hours a day does the pain in the previously painful area last? (Pain Duration)

9. How are you feeling now? Please say good, average, or bad. (current mood)

10. Are you currently worried or anxious? Please answer yes or no. (Anxiety)

11. How many hours did you sleep? (sleep time)

12. Did you sleep well without waking up in the middle of the night? (quality of sleep)

The management unit 130 may additionally match and manage at least one of text and status information to the patient's identification information. As an example, the management unit 130 may store or transmit matching identification information, voice responses, text, status information, recorded files, etc. to the server 30 .

The patient's identification information includes a UID (User Identifier) that is a combination of at least one of numbers, letters, and special symbols, and may be formed regardless of the patient's personal information.

The interrogation unit 110 and the collection unit 120 may be placed in a position facing the patient by the moving unit 150.

The moving unit 150 can physically or communicatively move the interrogation unit 110 and the collection unit 120.

As an example, when a body on which the interrogation unit 110 and the collection unit 120 are mounted, the moving unit 150 is provided, a driving means such as a motor that moves the body, a wheel that rotates or walks by the driving means, and an infinite It may include means of movement such as orbits and articulated legs. In this case, the interrogation unit 110 and the storage unit can physically travel through hallways, hospital rooms, etc. using the moving unit 150 to visit a specific patient in a bed in a specific hospital room.

As another example, when the interrogation unit 110 and the collection unit 120 are installed in a hospital room, a bed, or a wall around the bed, the moving unit 150 includes a plurality of interrogation units 110 installed in each hospital room or each bed and a collection unit ( 120) A control signal that operates the interrogation unit 110 and the collection unit 120 facing the bed of a specific patient can be transmitted.

In the case of the physically moving moving unit 150, the interrogation unit 110 and the collection unit 120 may operate only for a specific patient at a certain point in time.

On the other hand, when the interrogation unit 110 and the collection unit 120 are provided for each bed, a plurality of interrogation units 110 and collection units 120 arranged in the same hospital room can operate simultaneously in terms of hardware. However, in the same hospital room, multiple question units 110 and multiple collection units 120 operate simultaneously, and when multiple patients individually provide voice answers, the sounds are mixed and propagated, making it difficult for both questions and answers to be performed normally. there is. The moving unit 150 transmits a signal that controls only one of the interrogation units 110 and collection units 120 to operate at a certain point among the plurality of interrogation units 110 and collection units 120 arranged in the same hospital room. It can be transmitted to each question unit 110 and each collection unit 120 through a communication network.

When the list of interview subjects is obtained, the moving unit 150 can move the questioning unit 110 and the collection unit 120 to target a bed in a hospital room where a specific patient included in the list is hospitalized. For example, the moving unit 150 may travel through corridors and hospital rooms from its initial location and arrive at the bed of a specific patient together with the interrogation unit 110 and the collection unit 120. Alternatively, the moving unit 150 operates (on) the interrogation unit 110 and the collection unit 120 installed on the bed in which the specific patient is hospitalized or the wall around the bed, and operates the remaining interrogation unit 120 in the hospital room where the specific patient is hospitalized. 110 and the collection unit 120 can be turned off. Through this, the same effect can be achieved as if the normally operating interrogation unit 110 and collection unit 120 were physically moved to a specific patient.

If the interrogation unit 110 and the collection unit 120 satisfy the condition of physically or communicationally facing the bed of a specific patient, the moving unit 150 directly identifies the patient's current status from the patient located in the bed independently of the list. Information can be obtained. The questionnaire unit 110 may output different questions depending on whether existing questionnaire data targeting the current identification information obtained through the mobile unit 150 exists.

In medical cases, serious accidents can occur if the diagnosis of one patient and the diagnosis of another patient are reversed. To prevent this, the mobile unit 150 can confirm once again who the patient in a specific patient's bed is, even if it is assumed to be the specific patient.

The first patient can go to the bathroom while a second patient in another hospital room is visiting the first patient. In this state, the mobile unit 150 that visits the bed may make an error in conducting a medical examination of the first patient while targeting the second patient. To prevent this, the mobile unit 150 can once again check the current identification information of the patient (second patient) currently located in the bed. Identification information can be confirmed by scanning the patient identification code worn on the patient's arm.

If the mobile unit 150 visits the first patient recorded in the list, but a second patient is recognized instead of the first patient through the current identification information, the mobile unit 150 determines whether the second patient is included in the list. You can check . If the second patient is included in the list, since the questionnaire for the second patient must be performed anyway, the questioning unit 110 and the collection unit 120 may perform the questionnaire for the second patient.

If the current identification information of the first patient is obtained from the first patient's bed, the questionnaire unit 110 can output different questions depending on whether existing questionnaire data exists for the first patient according to the list. there is.

Existing questionnaire data may include status information obtained through the performance of the first question described above or obtained through the performance of the second question. In this case, if existing questionnaire data for a specific patient exists, the question unit 110 may ask the second question. If there is no existing questionnaire data for a specific patient, the questioning unit 110 may ask the first question.

The judgment unit 240 may determine whether the patient's voice answer to a specific question intended to be a first type of answer is correct. In this specification, a correct voice answer may mean that the voice answer exists in the set answer list for the question. If the voice answer does not exist in the set answer list, the determination unit 240 may determine that the voice answer is incorrect.

As an example, the determination unit 240 may extract a feature word from a plurality of words included in the voice response through analysis of the auditory voice response or analysis of text converted by the conversion unit 220. Feature words may include various words that can be answers to status information. For example, ‘knee’, ‘throbbing’, ‘when walking’, ‘1 year’, ‘normal’, etc. may be used as feature words.

The determination unit 240 may determine whether a feature word exists in the set answer list for the question. If a feature word does not exist in the set answer list, the question unit 110 may output a similar question intended as a first type of answer instead of moving on to the next question.

For example, the question unit 110 asks, 'Where is the most painful area currently?' to obtain a feature word indicating the pain area. If there are multiple areas, you can ask questions like, ‘Please tell me in the order that it hurts most.’

In response to this, the patient may not respond with an answer such as 'knee,' but may respond that the cat is in pain.

In this case, the question section 110 asks, unlike the existing question, 'Which part is hurting the most right now?' in order to induce the [pain area] 'knee'. Alternatively, you can print similar questions such as ‘Where is the most uncomfortable place right now?’

The question unit 110 may output similar questions such as ‘How does the pain feel?’ and ‘How does it hurt?’ in order to induce an answer appropriate to [pain pattern].

The question section 110 may output similar questions such as 'When is the pain the most severe?' and 'When does it hurt the most during the day?' in order to induce an appropriate answer for the [period of pain intensification]. .

In order to induce an answer appropriate to [pain intensity], the question section 110 asks questions such as ‘If you were to express the pain between 0 and 10, how would you rate it?’, ‘Can you please express how much it hurts with a number between 0 and 10?’ Similar questions can be printed together.

In order to derive an answer appropriate for [pain duration], the question section 110 may output similar questions such as ‘How long has the pain been getting worse?’ and ‘How long has the pain recently gotten worse?’ .

When the first answer is obtained through the first question to obtain the first type of feature word, the collection unit 120 can detect not only the first type of feature word but also the second type of feature word from the first answer. there is.

As an example, the question unit 110 asks, 'Where is the most painful area currently?' to obtain a characteristic word indicating the pain area. If there are multiple areas, you can ask questions like, ‘Please tell me in the order that it hurts most.’

At this time, instead of answering ‘my knee hurts,’ the patient may answer ‘my knee hurts when I walk.’ The collection unit 120 may detect ‘knee’ corresponding to the first type of feature word in ‘knee hurts when walking’. Not stopping here, the collection unit 120 may additionally detect 'when walking' corresponding to the second type of feature word in the first answer.

When the second type of feature word is detected in the first answer, the question unit 110 asks the question immediately following the first question regardless of the preset question order to obtain the second type of feature word. You can decide and print it.

The answer to the first question, ‘Knees hurt when walking,’ also includes the answer to the fourth question.

In this state, in order, the second question, ‘What kind of pain do you feel? Please express it as numb or throbbing.’ and the third question, ‘What kind of pain do you feel? After performing ‘Please express it as numb or throbbing’, the fourth question ‘When is the pain the most severe?’ can be asked. However, in this case, the patient is likely to have negative feelings that he or she is being made fun of through puns. As a way to suppress the patient's negative emotions regarding the medical examination, the questioning unit 110 of this embodiment can perform the fourth question immediately after the first question when the answer to the fourth question is obtained through the first question.

In other words, if the patient answers, ‘My knees hurt when I walk,’ the question unit 110 will skip the second and third questions and ask the fourth question, ‘What position does it hurt the most?’ You can. In this case, although repeated answers are requested, the patient may perceive it as a way to confirm his or her answers and may not have negative feelings. Afterwards, the second and third questions that are pushed back in the question order may be performed.

In order for the patient to recognize his or her answer as a way to confirm his or her answer, the question unit 110 checks the second type of feature word included in the first answer instead of the fourth question corresponding to the preset second question. The modified second question of the dimension can be output.

For example, the question unit 110 may ask the fourth question corresponding to the second question immediately after the first question, and in this case, the fourth question may be modified to be a confirmation question.

In other words, if the patient answers, 'My knees hurt when I walk,' the question unit 110 uses the second type of feature word, 'What position does it hurt the most?', which corresponds to the fourth question. You can ask confirmatory questions such as ‘Does it hurt the most when you walk?’, which includes ‘when you walk’. According to this embodiment, the user's discomfort due to incorrect questioning method can be minimized.

Meanwhile, 'What kind of pain do you feel? To the question ‘Please express it as numb or throbbing,’ the patient may answer ‘It tingles.’

If the patient answers ‘It hurts,’ it is realistically difficult to conduct a questionnaire to classify the patient’s condition. This is because, experimentally, the expression ‘tingling’ is commonly (intersectingly) used in somatic pain, neuropathic pain, and visceral pain.

In this case, even if multiple similar questions with different expressions are asked, the patient may still answer 'tingles', which is an intersectional feature word. In order to avoid or exclude the intersection feature word, the question unit 110 may generate an avoidance question that includes the intersection feature word and also includes an exclusive word that excludes the intersection feature word. Avoidance questions include the intersectional feature word 'sting', such as 'please express it differently, excluding stinging', and may also include exclusive words such as 'excluding'. If you ask avoidance questions that include both intersectional feature words and exclusive words, the patient may express pain using an expression other than 'it hurts.'

The medical examination device can be operated until the daily medical examination list is completed. If there is no previous questionnaire data for a specific patient, the medical examination device can set a medical examination script for the patient before surgery.

The medical examination device can set up a medical examination script for patients after surgery if there is previous (existing) medical examination data for a specific patient.

The preoperative patient interview script will include questions about the area of pain associated with the pain, the pattern of the pain, when the pain intensifies, the position in which the pain intensifies, the intensity of the pain, the duration of the pain, the current mood related to stress, anxiety, sleep duration, and sleep quality. You can.

The post-operative patient interview script collects information on the presence or absence of pain related to pain, pain pattern, pain intensity, and pain duration by dividing them into the surgical area and the area that was painful before surgery. Current mood related to stress, anxiety, sleep time, and sleep duration. Questions may be included to gather information about the quality of the product.

The paperweight device of the present invention applies a reinforcement learning algorithm for natural conversation, so it does not always ask questions in the same order, but can variably change the order of questions depending on the situation. For example, if a specific answer includes the answer to another question, that other question may be displayed as the next question for natural conversation.

Figure 2 is a schematic diagram showing the actions of the question order variable algorithm.

The question order variable algorithm can be expressed as Equation 1 below.

[Equation 1]

The Q-Learning algorithm can be used to change the question order.

The action indicated by 'a' in Equation 1 may consist of each question, first greeting, and last greeting.

Actions include 'Greeting', 'Position', 'Quality', 'R_factor1', 'R_factor2', 'Severity', 'Timing', 'Mood', 'Anxiety', 'Sleep1', 'Sleep2', 'Bye', etc. may include. Each action can contain at least three sentences, including similar questions with the same meaning. After an action is selected, one question among several questions for that action may be randomly asked.

The state represented by 's' in Equation 1 can indicate whether each question (action) has been performed as an array of 0 and 1.

For example, the state is [0,0,0,0,0,0,0,0,0,0,0,0] to [1,1,1,1,1,1,1, It can be composed of 2 ¹² = 4096 of [1,1,1,1,1].

As an example, [0,0,0,0,0,0,0,0,0,0,0,0] may indicate a state in which no questions are asked. [1,0,0,0,0,0,0,0,0,0,0,1] can indicate a state in which only 'Greeting' and 'Bye' are performed.

The reward represented by 'r' in Equation 1 is 'Greeting' first, performing the 'Bye' question after all answers are received, pain questions first, and then stress-related questions, etc. It can be designed to give a reward value when the question sequences that meet the criteria are executed.

The question order variable algorithm can be designed so that rewards and penalties are given. If the penalty value satisfies the set value, the turn can be forcibly ended.

For example, when Greeting appears first, when Position questions are asked after Greeting, when Quality questions are asked after Greeting and Position questions, when the remaining Pain-related words are asked after Greeting and Position, and when the Stress questions are asked after completing Pain questions. When performing this, if the Mood question is performed before the Anxiety question, a reward may be granted if Bye is printed after all answers are received. If a reward is granted, the priority level to be adopted as the next question to be printed may increase.

For example, when a Bye question is asked when no other answers have been received, when a Position answer is not received and the action is not Greeting or Position, when a Quality answer is not received, the action is not one of Greeting, Position or Quality. In this case, if a stress-related question is asked before the pain question has been completed, if the greeting question does not appear first, a penalty may be given if a question that has already been answered is asked again. If a penalty is granted, the priority level to be adopted as the next output question may be lowered. The questionnaire device of the present invention can determine whether the patient's answer is an appropriate answer to the question after performing the question and obtaining the patient's answer. . At this time, the medical examination device determines the patient's condition information by matching the data of the vocabulary database 290 and the patient's answers, and if the answer is not stored in the vocabulary database 290 or if the answer falls into a different category, the patient Assuming that you did not accurately recognize the intent of the question, you can help to understand the question again by asking similar questions.

When the correct answer is obtained, the medical examination device updates the status information, and when answers to all questions are obtained in this way, the medical examination result corresponding to the collected status information can be stored and transmitted to the server 30.

Once the medical examination device has completed the examination of all patients that need to be performed, the results can be delivered to the medical staff in charge. At this time, information about patients who appear to need a doctor's diagnosis, such as the intensity of pain has increased, the duration of pain has increased, or all stress-related information shows poor results, may be additionally delivered to the medical staff.

Figure 3 is a flowchart showing the method of collecting patient condition information according to the present invention.

The collection method of FIG. 3 can be performed by the paperweight device shown in FIG. 1.

The server 30 may transmit a list of medical examination subjects to the medical examination robot 10 (S 501). Through this, the interview robot 10 can obtain a list of interview subjects. The list of interview subjects may include the patient's UID information, hospital room information, and bed information.

The medical examination robot 10 or the server 30 can check whether the medical examination has been completed for all patients included in the list (S 514). When the medical examination of all patients is completed, the medical examination robot 10 or the server 30 can deliver the results of the medical examination to the doctor in charge of each patient (S 515).

The medical examination robot 10 can move to target a specific patient included in the list.

The medical examination robot 10 can determine the identification information of a specific patient and transmit the identified identification information to the server 30.

When the patient UID corresponding to the identification information obtained from the medical examination robot 10 matches the list (S 502), the server 30 can search whether existing medical examination data exists for the identification information of a specific patient. (S 503).

If the existing questionnaire data does not exist, the server 30 may transmit the first question to the questionnaire robot 10 (S 504). If existing questionnaire data exists, the server 30 may transmit the second question to the questionnaire robot 10 (S 505).

The interview robot 10 can convert the questions included in the first question or the second question into voices and output them. The questionnaire robot 10 can ask questions whose order is dynamically adjusted based on reinforcement learning (S 506).

The questionnaire robot 10 can collect voice answers from a specific patient (S 507). The medical examination robot 10 may continue to ask questions about a specific patient until answers are obtained to all questions included in the first or second questions (S 511).

The medical examination robot 10 can transmit a specific patient's voice response to the server 30. The server 30 may determine whether the voice answer is correct and transmit the answer recognition result including the judgment result and the following questions to the questionnaire robot 10 (S 508). If the interview robot 10 determines that the patient's voice answer is incorrect, it can perform similar questions (S 509)

If the patient's voice answer is determined to be correct, the medical examination robot 10 may update or self-store the answer recognition result or the patient's status information extracted from the answer recognition result (S 510, S 512). After saving the results upon answering, the questionnaire robot 10 can output the following questions for a specific patient.

When the interview for a specific patient is completed (S 511), the interview robot 10 may transmit the pre-stored answer recognition results and recorded voice answer data to the server 30 (S 513).

The server 30 can create a database of the answer recognition results and voice answers.

4 is a diagram illustrating a computing device according to an embodiment of the present invention. The computing device TN100 of FIG. 4 may be a device described herein (e.g., a paper weight device, etc.).

In the embodiment of FIG. 4 , the computing device TN100 may include at least one processor TN110, a transceiver device TN120, and a memory TN130. Additionally, the computing device TN100 may further include a storage device TN140, an input interface device TN150, an output interface device TN160, etc. Components included in the computing device TN100 may be connected by a bus TN170 and communicate with each other.

The processor TN110 may execute a program command stored in at least one of the memory TN130 and the storage device TN140. The processor TN110 may refer to a central processing unit (CPU), a graphics processing unit (GPU), or a dedicated processor on which methods according to embodiments of the present invention are performed. Processor TN110 may be configured to implement procedures, functions, and methods described in connection with embodiments of the present invention. The processor TN110 may control each component of the computing device TN100.

Each of the memory TN130 and the storage device TN140 can store various information related to the operation of the processor TN110. Each of the memory TN130 and the storage device TN140 may be comprised of at least one of a volatile storage medium and a non-volatile storage medium. For example, the memory TN130 may be comprised of at least one of read only memory (ROM) and random access memory (RAM).

The transceiving device TN120 can transmit or receive wired signals or wireless signals. The transmitting and receiving device (TN120) can be connected to a network and perform communication.

Meanwhile, the embodiments of the present invention are not only implemented through the apparatus and/or method described so far, but may also be implemented through a program that realizes the function corresponding to the configuration of the embodiment of the present invention or a recording medium on which the program is recorded. This implementation can be easily implemented by anyone skilled in the art from the description of the above-described embodiments.

Although the embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto, and various modifications and improvements made by those skilled in the art using the basic concept of the present invention defined in the following claims are also possible. It falls within the scope of invention rights.

10...paperweight robot 30...server
110...Question Department 120...Collection Department
130...Management Department 140...Recording Department
150...moving part 210...processing part
220... conversion unit 230... extraction unit
240...Judgment Department 290...Database

Claims (7)

A question section that outputs questions that induce a vocal response from the patient;
a collection unit that collects the voice response;
A management unit that manages the voice response by matching it with the patient's identification information,
When a first answer is obtained through a first question for obtaining a first type of feature word, the collection unit detects not only the first type of feature word but also a second type of feature word from the first answer,
When the second type of feature word is detected in the first answer, the question unit asks a second question immediately following the first question regardless of the preset question order to obtain the second type of feature word. Decide on the details and print them out,
The questionnaire unit is a paperweight device that performs another question that has been pushed back in the order of questions due to the second question after performing the second question.
According to paragraph 1,
A conversion unit for converting the voice response into text, an extraction unit for extracting status information of the patient from the voice response or the text, and a recording unit for recording the voice response are provided,
The management unit manages by additionally matching at least one of the text and the status information to the identification information,
The identification information includes a UID (User Identifier) that is a combination of at least one of numbers, letters, and special symbols, and is formed independently of the patient's personal information.
A question section that outputs questions that induce a vocal response from the patient;
a collection unit that collects the voice response;
A management unit that manages the voice response by matching it with the patient's identification information,
A moving part is provided that physically moves or communicates the interrogation unit and the collection unit,
When the list of interview subjects is obtained, the moving unit moves the questioning unit and the collection unit to target a bed in a hospital room where a specific patient included in the list is hospitalized,
When the interrogation unit and the collection unit meet the condition of facing the bed, the moving unit obtains the patient's current identification information directly from the patient located on the bed independently of the list,
The questionnaire unit outputs different questions depending on whether existing questionnaire data targeting the current identification information obtained through the mobile unit exists,
The existing questionnaire data includes status information obtained through performing a first question or obtained through performing a second question,
The first question includes questions targeting patients before surgery,
The second question includes questions for patients after surgery,
If the existing questionnaire data exists, the questionnaire unit outputs the second question,
The questionnaire unit outputs the first question if the existing questionnaire data does not exist.
According to paragraph 1,
A judgment unit is provided to determine whether the patient's voice answer to a specific question intended to be a first type of answer is correct,
The determination unit extracts a feature word from a plurality of words included in the voice response,
The determination unit determines whether the feature word exists in the set answer list for the question,
The questionnaire unit outputs a similar question intended as the first type of answer instead of moving to the next question if the feature word does not exist in the set answer list.
delete According to paragraph 1,
The questionnaire unit outputs a modified second question for confirmation containing the second type of characteristic word included in the first answer instead of the preset second question.
In a method of collecting patient status information performed by a medical examination device,
Transmitting a list of medical examination subjects from the server to the medical examination robot;
moving the medical examination robot to target a specific patient included in the list;
The medical examination robot identifying identification information of the specific patient and transmitting the identified identification information to the server;
Searching the server for whether existing questionnaire data exists targeting the identification information of the specific patient;
transmitting a first question to the medical examination robot if the existing medical examination data does not exist in the server, and transmitting a second question to the medical examination robot if the existing medical examination data exists;
Converting the first question or the second question into voice and outputting it in the medical examination robot;
Collecting voice answers from the specific patient from the medical examination robot and transmitting the voice answers to the server;
Determining whether the voice answer is correct at the server and transmitting the answer recognition result including the judgment result and the next question to the questionnaire robot;
storing the answer recognition result in the medical examination robot and outputting the next question targeting the specific patient;
When the interview robot for the specific patient is completed, transmitting the pre-stored answer recognition result and recorded data of the voice answer to the server;
Comprising: databaseizing the answer recognition result and the voice answer in the server,
The existing questionnaire data includes status information obtained through performing a first question or obtained through performing a second question,
The first question includes questions targeting patients before surgery,
The second question includes questions for patients after surgery,
The step of converting the first question or the second question into a voice and outputting it in the questionnaire robot,
If the existing questionnaire data exists, output the second question,
If the existing questionnaire data does not exist, a collection method for outputting the first question.

Images