KR102624931B1 - Machine learning-based urinary disorder diagnosis and diaper dermatitis diagnosis system - Google Patents

Abstract

The machine learning-based urinary disorder diagnosis and diaper dermatitis diagnosis system of the present invention includes a smart diaper that is filled between the user's legs, collects urine, and generates a second electrical signal when urinating, and transmits a first electrical signal to the smart diaper. A terminal that receives the second electrical signal from the smart diaper, generates a urine generation signal based on the transmitted second electrical signal, and transmits the generated urine generation signal and the transmitted second electrical signal to the server. , and urination-related data from at least one terminal - the urination-related data includes urination start resistance value (Rs), urination end resistance value (Re), duration of one urination (Rh), and total number of urinations during the day (Utd). , including the total number of urinations per day (Und) from 9 PM to 7 AM and the total urination time per week (W), and determine urination disorder or diaper dermatitis based on the urination-related data. may include.

Description

Machine learning-based urinary disorder diagnosis and diaper dermatitis diagnosis system}

The present invention relates to a system for diagnosing urinary disorders and diaper dermatitis through urination pattern analysis using machine learning.

Urination control disorders, including urinary incontinence, are common diseases that affect more than half of the elderly living in long-term care facilities. Failure to change diapers in time immediately after urination exposes the patient to the risk of dermatitis or urinary tract infection.

To prevent this, the patient's diaper must be checked and changed frequently, which interferes with sufficient rest and sound sleep for the patient and guardian, lowering the quality of life. In addition, in acute medical institutions, inconveniences arise such as having to weigh diapers every time to measure the patient's urine output.

Republic of Korea Patent Publication No. 1020060068819 (published on June 21, 2006); Diapers that remind you to urinate

The present invention has the advantage of calculating the probability of occurrence of urinary disorders and diaper dermatitis and inducing appropriate management according to the probability of occurrence of urinary disorders and diaper dermatitis.

The machine learning-based urinary disorder diagnosis and diaper dermatitis diagnosis system of the present invention includes a smart diaper that is filled between the user's legs, collects urine, and generates a second electrical signal when urinating, and transmits a first electrical signal to the smart diaper. A terminal that receives the second electrical signal from the smart diaper, generates a urine generation signal based on the transmitted second electrical signal, and transmits the generated urine generation signal and the transmitted second electrical signal to the server. , and urination-related data from at least one terminal - the urination-related data includes urination start resistance value (Rs), urination end resistance value (Re), duration of one urination (Rh), and total number of urinations during the day (Utd). , including the total number of urinations per day (Und) from 9 PM to 7 AM and the total urination time per week (W), and determine urination disorder or diaper dermatitis based on the urination-related data. may include.

Additionally, the server can calculate the probability of occurrence of urinary disorders or diaper dermatitis using a machine learning algorithm.

Additionally, the server can calculate the probability of occurrence of urinary incontinence using the equation (1) below.

(Re-Rs)/Rh = Cr1 Equation (1)

(Here, Cr1 is the probability of urinary incontinence occurring, Re is the resistance value at the end of urination, Rs is the resistance value at the start of urination, and Rh is the duration of one urination)

Additionally, the server can calculate the probability of enuresis using the equation (2) below.

Und/Utd = Cr2 equation (2)

Here, Cr2 is the probability of enuresis, Und is the total number of urinations during the day from 9 PM to 7 AM, and Utd is the total number of urinations during the day.)

Additionally, the server can calculate the probability of oliguria occurrence using equation (3) below.

Utd/Utad = Cr3 equation (3)

(Here, Cr3 is the probability of occurrence of oliguria, Utd is the total number of urinations per day of the user, and Utd is the average total number of urinations per day for all users)

Additionally, the server can calculate the probability of oliguria occurrence using equation (4) below.

It/W = Cr4 Equation (4)

(Here, Cr4 is the probability of developing diaper dermatitis, It is the total incontinence time accumulated during the day, and W is the total urination time per week)

The present invention has the advantage of inducing appropriate management of urinary disorders by calculating the probability of occurrence of urinary disorders in elderly people living in long-term care facilities through machine learning.

In addition, the present invention has the advantage of improving the quality of life of elderly people with a high probability of developing dermatitis by increasing the frequency of diaper changes by calculating the probability of developing diaper dermatitis based on machine learning.

Additionally, the present invention has the advantage of lowering the burden of frequently checking and replacing the patient's diaper.

Figure 1 shows the smart diaper of the present invention and the terminal separated.
Figure 2 shows a terminal mounted on the smart diaper of the present invention.
Figure 3 is a block diagram of the urinary disorder diagnosis and diaper dermatitis diagnosis system through urination pattern analysis using machine learning of the present invention.
Figure 4 is a block diagram of the terminal of the present invention.
Figure 5 is a block diagram of the server of the present invention.
Figure 6 shows the probability of occurrence of urinary incontinence according to equation (1) of the present invention.
Figure 7 shows the probability of enuresis according to equation (2) of the present invention.
Figure 8 shows the probability of oliguria occurring according to equation (3) of the present invention.
Figure 9 shows the probability of occurrence of diaper dermatitis (IAD) according to equation (4) of the present invention.
Figure 10 shows the body of the present invention.

The system for diagnosing urinary disorders and diaper dermatitis through urination pattern analysis using machine learning according to an embodiment of the present invention includes a smart diaper 100, a terminal 200, and a server 300.

<Smart Diaper (100)>

The smart diaper 100 according to an embodiment of the present invention may include a diaper portion 110, a pattern portion 120, and a first connection terminal 130.

1. Diaper section (110)

The diaper unit 110 is a device that is filled between the user's legs and collects urine and may include a body 111 and a waist belt 112.

The body 111 is disposed to face the user's excretory organ that excretes urine. The body 111 has a length extending from the user's hips through between the legs to the stomach area.

The waist belt 112 extends from the side of one end of the body 111 and has a length extending from the user's hips through the user's waist to the user's stomach.

The body 111 includes a surface layer 111a where urine first reaches, a diffusion layer 111b that diffuses urine, an absorption layer that absorbs and stores urine, and a waterproof layer 111e that prevents urine from leaking to the outside. .

The surface layer 111a is formed of a hydrophilic fiber material and receives the wearer's urine and transfers it to the diffusion layer 111b.

The diffusion layer (111b) is made of a hydrophobic material to diffuse the urine delivered from the water conduction layer and transfer it to the absorption layer.

The absorption layer may include an upper layer 111c made of a hydrophilic material and a lower layer 111d made of a cellulose material. The upper layer 111c absorbs a certain amount of urine. The lower layer (111d) traps the remaining urine that cannot be absorbed by the upper layer (111c) and prevents reverse leakage of urine. Here, the reverse gland refers to the reverse flow of urine from the absorbent layer to the conductive layer when pressure is applied to the urinary incontinence panties due to the wearer's posture change, cough, etc.

The waterproof layer 111e is a portion that meets the inside of the panty section and may be a backsheet made of a waterproof material to prevent urine absorbed in the urine absorbent layer from being discharged to the outside. The backsheet may have fine holes that allow moisture in a vapor state to be discharged without discharging liquid moisture.

2. Pattern part (120)

The pattern portion 120 may be formed on the inner surface of the waterproof layer 111e or the outer surface of the waterproof layer 111e.

The pattern portion 120 may be formed by opposing the first pattern 121 and the second pattern 122 at regular intervals. The first pattern 121 and the second pattern 122 may be in the form of parallel lines or may include a plurality of branches extending from the parallel lines.

At the end of the pattern portion 120, there are a plurality of devices that transmit electrical signals from the terminal 200 to the pattern portion 120 and transmit electrical signals generated in the pattern portion 120 to the terminal 200. One connection terminal 130 may be formed.

3. First connection terminal (130)

The first connection terminal 130 may be made of a metal material to receive and transmit electrical signals. In addition, it is formed in an even number of 2, 4, 6, etc. and is connected to the first pattern 121 and the second pattern 122, respectively. Two first connection terminals 130 are formed, one first connection terminal 130 is connected to the first pattern 121, and the other first connection terminal 130 is connected to the second pattern 122. connected.

Additionally, the first connection terminal 130 may be detachably connected to a second connection terminal 210 formed in the terminal 200, which will be described later.

A protrusion is formed on the first connection terminal 130, and a recess is formed on the second connection terminal 210 at a position corresponding to the protrusion, so that the protrusion can be aligned and connected to the recess. Additionally, when the protrusion is separated from the concave portion, the second connection terminal 210 is separated from the first connection terminal 130. The number of the first connection terminals 130 and the second connection terminals 210 is the same, and they are formed at corresponding positions so that the protrusions and concave parts are aligned.

Conversely, a protrusion is formed in the second connection terminal 210, and a recess is formed in the first connection terminal 130 at a position corresponding to the protrusion, so that the protrusion can be aligned and connected to the recess. Additionally, when the protrusion is separated from the concave portion, the second connection terminal 210 is separated from the first connection terminal 130.

Alternatively, the first connection terminal 130 and the second connection terminal 210 are formed of a magnetic metal material, so that the first connection terminal 130 and the second connection terminal 210 are connected to each other by magnetic force. It may be connected.

The first connection terminal 130 and the second connection terminal 210 are made of a metal material capable of exchanging and receiving electrical signals.

<Terminal (200)>

The terminal 200 may be mounted on the outer surface of the waist belt 112 of the diaper.

The terminal 200 may include a second connection terminal 210, a battery 220, a determination unit 230, and a first communication unit 240.

The terminal 200 may be mounted below or above the wearer's navel. Preferably, it can be mounted below the navel to prevent discomfort during wearing.

1. Second connection terminal (210)

As described above, the second connection terminal 210 is formed to be detachable from the first connection terminal 130. The second connection terminals 210 are formed in the same number as the first connection terminals 130, transmit the first electrical signal to the first connection terminal 130, and transmit the first electrical signal to the first connection terminal 130. Receives a second electrical signal from.

2. Battery (220)

The terminal 200 may have a built-in battery 220 that provides driving power for urinary incontinence management. The battery 220 can be either disposable or rechargeable.

3. Judgment unit (230)

The determination unit 230 may generate an electrical signal that supplies an electrical signal to the pattern unit 120 and transmit it to the battery 220. When receiving the electricity generation signal, the battery 220 may generate a preset first electric signal and transmit it to the pattern portion 120.

The first electrical signal is transmitted to the first connection terminal 130 through the second connection terminal 210. The first electrical signal transmitted to the first connection terminal 130 is transmitted to the first pattern 121 portion 120.

When the urine collection part is not stained with urine, the resistance value between the first pattern 121 part 120 and the second pattern 122 part 120 is large and the dielectric constant is small, so that the first pattern 121 ) The first electrical signal transmitted to the unit 120 cannot be transmitted to the second pattern 122 unit 120.

Conversely, when urine is stained on the urine collection part, the resistance value between the first pattern 121 part 120 and the second pattern 122 part 120 decreases and the dielectric constant increases, and the first pattern 121 part 120 decreases and the dielectric constant increases. The first electrical signal transmitted to the unit 120 (121) is transmitted to the second pattern unit 120 (122).

When the second pattern 122 unit 120 receives the first electrical signal, it outputs a second electrical signal and transmits it to the first connection terminal 130. The second electrical signal transmitted to the first connection terminal 130 may be transmitted to the determination unit 230 through the second connection terminal 210.

The determination unit 230 may determine that urine has occurred when the second electrical signal is received.

Additionally, the determination unit 230 may generate a urine generation signal based on the second electrical signal and transmit the urine generation signal and the second electrical signal to the communication unit.

4, Ministry of Communications

When the communication unit receives the urine generation signal from the determination unit 230, the communication unit may transmit the urine generation signal and the second electrical signal to the portable terminal or the server 300.

<Server (300)>

The server 300 according to an embodiment of the present invention includes an input unit 310, a big data unit 320, and a calculation unit 330.

The input unit 310 receives urination start resistance value (Rs), urination end resistance value (Re), duration of one urination (Rh), total number of urinations during the day (Utd), and urination time at 9 PM from a plurality of terminals. The total number of urinations (Und) and the total number of urination times per week (W) from start to 7 a.m. are received, and the received values are transmitted to the big data unit 320.

The big data unit 320 receives the daily urination start resistance value (Rs), urination end resistance value (Re), urination time (Rh), total number of urinations per day (Utd), and total number of urinations per day (Utd) on a daily basis from the input unit 310. Enter the number of urinations (Und) and urination time per week.

Here, the urination start resistance value (Rs), the urination end resistance value (Re), the duration of one urination (Rh), and the total number of urinations during the day (Utd), stored in the big data unit 320, at 9 PM during the day. The total number of urinations (Und) and total urination time (W) per week from start to 7 am are referred to as urination-related data.

The calculation unit 330 extracts statistical characteristics and urination disorder patterns from the urination-related data through mathematical equations to be described later. Each feature extracted in this way is provided as data for a machine learning algorithm to derive the probability of urinary disorder.

(1) Statistical characteristics related to urinary incontinence

Among urination-related data, daily urination start resistance value (Rs), urination end resistance value (Re), and duration of one urination (Rh) are used. Statistical characteristics representing urinary incontinence use equation (1).

(Re-Rs)/Rh = Cr1 Equation (1)

Here, Cr1 is the probability of occurrence of urinary incontinence, Re is the resistance value at the end of urination, Rs is the resistance value at the start of urination, and Rh is the duration of one urination.

Figure 6 shows the probability of occurrence of urinary incontinence according to equation (1).

(2) Statistical characteristics related to enuresis

Among urination-related data, the total number of urinations per day (Utd) and the total number of urinations per day from 9 PM to 7 AM (Und) are used. Statistical characteristics representing enuresis use equation (2).

Und/Utd = Cr2 equation (2)

Here, Cr2 is the probability of enuresis, Und is the total number of urinations during the day from 9 PM to 7 AM, and Utd is the total number of urinations during the day.

Figure 7 shows the probability of enuresis according to equation (2).

(3) Statistical characteristics related to oliguria

Among urination-related data, the total number of urinations per day (Utd) is used. The big data unit 320 receives the total number of urinations per day (Utd) of a plurality of users from a plurality of terminals. The calculation unit 330 receives the total number of urinations per day (Utd) of a plurality of users and calculates the average total number of urinations per day (Utad) of all users based on this.

Based on the calculated average total number of urinations per day (Utad) of all users and the total number of urinations per day (Utd) of all users calculated above, statistical characteristics related to oliguria are derived using Equation (3).

Utd/Utad = Cr3 equation (3)

Here, Cr3 is the probability of occurrence of oliguria, Utd is the total number of urinations per day of the user, and Utd is the average total number of urinations per day for all users.

Figure 8 shows the probability of oliguria occurring according to equation (3).

(4) Statistical characteristics related to diaper dermatitis (IAD)

The calculation unit 330 calculates the duration of incontinence on a daily basis based on the urination start resistance value (Rs), the urination end resistance value (Re), and the duration of one urination (Rh). If the difference between the urination start resistance value (Rs) and the urination end resistance value (Re) is within a predetermined range, it is judged to be incontinence, and the urination start time and urination end time are calculated to calculate the incontinence time per session. The total incontinence time calculated in this way is accumulated over a day and the total incontinence time per session is set as It. Additionally, the calculation unit 330 can calculate the total urination time calculated on a weekly basis. Let W be the total urination time per week calculated in this way.

Based on the total urination time per week (W) and the total incontinence time accumulated during the day (It), statistical characteristics related to diaper dermatitis (IAD) are derived using Equation (4).

It/W = Cr4 Equation (4)

Figure 9 shows the probability of occurrence of diaper dermatitis (IAD) according to equation (4).

Here, Cr4 is the probability of developing diaper dermatitis, It is the total incontinence time accumulated during the day, and W is the total urination time per week.

The server 300 can express the probability of occurrence of urinary disorders and diaper dermatitis together. For example, the server 300 can be expressed as (incontinence 60%, IAD occurrence probability 45%), (oliguria 70%, IAD occurrence probability 57%), etc.

The calculated probability of occurrence of urinary disorder and diaper dermatitis can be displayed through a display connected to the server 300.

Claims (2)

A smart diaper that is filled between the user's legs, catches urine, and generates a second electrical signal when urinating;
Transmitting a first electrical signal to the smart diaper, receiving the second electrical signal from the smart diaper, generating a urine generation signal based on the transmitted second electrical signal, and transmitting the generated urine generation signal and the A terminal that transmits the second electrical signal to the server; and
Urine-related data from at least one terminal - The urination-related data includes urination start resistance value (Rs), urination end resistance value (Re), duration of one urination (Rh), total number of urinations per day (Utd), and per day. Includes a server that receives the total number of urinations (Und) from 9 PM to 7 AM and the total urination time per week (W) and determines urination disorder or diaper dermatitis based on the urination-related data. However,
The server calculates the probability of occurrence of the urinary disorder or diaper dermatitis using a machine learning algorithm,
The server is
Calculate the probability of urinary incontinence using the equation (1) below,
Calculate the probability of enuresis using equation (2) below,
Calculate the probability of oliguria using the equation (3) below,
A machine learning-based urinary disorder diagnosis and diaper dermatitis diagnosis system characterized by calculating the probability of oliguria using the equation (4) below.
(Re-Rs)/Rh = Cr1 Equation (1)
(Here, Cr1 is the probability of urinary incontinence occurring, Re is the resistance value at the end of urination, Rs is the resistance value at the start of urination, and Rh is the duration of one urination)
Und/Utd = Cr2 equation (2)
Here, Cr2 is the probability of enuresis, Und is the total number of urinations during the day from 9 PM to 7 AM, and Utd is the total number of urinations during the day.)
Utd/Utad = Cr3 equation (3)
(Here, Cr3 is the probability of occurrence of oliguria, Utd is the total number of urinations per day for the user, and Utad is the average total number of urinations per day for all users)
It/W = Cr4 Equation (4)
(Here, Cr4 is the probability of developing diaper dermatitis, It is the total incontinence time accumulated during the day, and W is the total urination time per week)
delete