A system for the early diagnosis of pressure ulcers.

Abstract

The present invention relates to a system for the early diagnosis of pressure ulcers (100). A system for the early diagnosis of pressure ulcers (100) includes at least one mattress / bed pad (110). The mattress / bed pad (110) contains pressure sensors to measure the pressure exerted on the mattress / bed pad (110) by areas of the patient's body; temperature sensors to measure the temperature generated on the mattress / bed pad (110) by areas of the patient's body; and humidity sensors to measure the humidity.A system for early diagnosis of pressure ulcers (100) also includes at least one camera (120) to capture an image of the patient's face; at least one microphone (130) to capture the patient's vocal expressions; and at least one processing unit configured to run a pressure ulcer diagnosis algorithm adapted to detect whether the patient has a pressure ulcer from a dataset containing data from pressure sensors, temperature sensors and humidity sensors, the camera (120) and the microphone (130).

Description

1 TARIFF A SYSTEM FOR EARLY DIAGNOSIS OF PRESSURE ULCERS Technical Field to Which the Invention Belongs The invention in question enables the early diagnosis of pressure ulcers in bedridden patients. It is related to a system that provides. 5 Background of the Invention Pressure ulcers, also known as pressure sores, are commonly seen in bedridden patients. body areas with bony prominences that are subjected to prolonged and uninterrupted pressure. Pressure sores are localized tissue damages that occur in the skin and / or underlying tissues as a result of prolonged exposure. Early detection and prevention of the wound increases the effectiveness of the treatment process and 10 This is critically important in preventing complications. Therefore, health... professionals or caregivers should regularly check the skin condition of individuals at risk. observe any possible color or temperature changes and, if necessary, inform the patient. They need to be repositioned in bed. However, this process is particularly difficult in a hospital setting. It increases the workload of healthcare workers who are dealing with a large number of patients, raising the risk of burnout by 15%. It increases the risk of injury and negatively affects patient comfort. It also compromises skin integrity. When the wound deteriorates, the risk of infection increases, and the wound area comes into contact with urine or feces. The possibility of developing pressure sores can also lead to serious complications. Pressure sores are not only caused by strenuous activity. not only the treatment processes but also the length of hospital stay, the patient and This leads to a decrease in the quality of life of their relatives and an increase in healthcare costs. It opens. A solution to the aforementioned problems is to enable early diagnosis of pressure ulcers. The system requires technical expertise. Purposes of the Invention The primary aim of the present invention is to create a system for the early diagnosis of pressure ulcers. to accomplish. Another aim of the invention is to reduce the pressure exerted by the patient's body on the bed / bed pad (110) to 30 pressure sensors to measure areas of the patient's body in the bed / bed 2 Temperature sensors and humidity sensors are used to measure the temperature and humidity created in the pad (110). from at least one bed / bed pad equipped with sensors, an image of the patient's face to capture, at least one camera, and at least one camera to capture the patient's vocal expressions. By analyzing data from a microphone, it can be determined whether or not the patient has a pressure ulcer. The goal is to create an AI-powered system that detects this. 5 Another aim of the invention is to capture at least one image of the patient's face. Data from the camera and at least one microphone to capture the patient's vocal expressions. with a measurement taken from the bed / bed pad to determine whether the patient is feeling pain. By analyzing the data and this detection data, it is determined whether or not the patient has a pressure ulcer. 10 The goal is to create an artificial intelligence-powered system. Detailed Description of the Invention An application realized to achieve the purposes of this invention is shown in the attached figures. as shown, and the details of the invention have been determined by considering the entire description. 15 These figures should be evaluated. Figure 1. Suitable for early diagnosis of pressure ulcers in a sample application of the invention. It is a schematic view of the system. The corresponding reference numbers used in the figures are given below. 100. System 110. Bed / mattress pad 111. Bed surface 25 Camera 120 130. Microphone Stand 140 141. Wheel The subject of the application is a system for the early diagnosis of pressure ulcers (100) at least one bed / bed Includes pad (110). On the bed / bed pad (110): - the pressure exerted by areas of a patient's body on the bed / bed pad (110) pressure sensors for measurement; 3 - the temperature created by the patient's body parts on the bed / bed pad (110) Temperature sensors for measuring temperature and humidity sensors for measuring humidity; It is located there. A system for early diagnosis of pressure ulcers (100) also: 5 - a small camera to capture the image of the patient's face (120); - at least one microphone to capture the patient's vocal expressions (130); - from pressure sensors, temperature and humidity sensors, camera (120) and from the data set containing data received from the microphone (130) whether the patient has a pressure ulcer to determine if it is not present, it runs a pressure ulcer diagnostic algorithm adapted for this purpose. 10 It includes at least one processing unit configured for this purpose. The subject of the application is a system for the early diagnosis of pressure ulcers (100) bedridden Includes a bed or mattress pad (110) on which the patient will lie. The bed / mattress pad (110) is for the patient To measure the pressure exerted by the parts of the body on the bed / mattress pad (110), the pressure is 15. sensors that form the areas of the patient's body on the bed / bed pad (110) It includes temperature sensors to measure temperature and humidity sensors to measure humidity. The aforementioned sensors are distributed throughout the bed / mattress pad (110) and the bed / mattress The sensors are located inside the mattress / bed pad (110). The sensors are located along the mattress / bed pad (110). 20 selected options, preferably distributed at selected intervals in a grid pattern. They may be densely distributed in high-risk areas and sparsely distributed in low-risk areas. The term "sensors" used in the application includes pressure sensors, temperature sensors, and It has been used to refer to humidity sensors. Sensors, preferably a group of sensors. assembled in a way that each sensor group is preferably in a grid form. It may be distributed within the bed / bed pad (110) in such a way as to form. Example 25 of the invention In one application, each sensor group consists of 4 pressure sensors and 2 temperature sensors. It includes a sensor and a humidity sensor. Sensors are positioned in a distributed manner throughout the bed / mattress pad (110) and each The position of the sensor on the bed / bed pad (110) is predetermined. Thus 30 measurement data of the sensors and the position of the sensors on the bed / bed pad (110) According to this, the patient's sleeping position can be determined. Based on the determined sleeping position... According to this, the data measured by the sensors indicates which areas of the patient's body are being applied. This is determined. For example, the shoulder, hip, and of a patient lying face down... The kneecap and body parts apply more pressure to the bed / bed pad (110) 35 4 This is expected. Therefore, on the bed / mattress pad with a shoulder, hip, and kneecap pattern. (110) located sensors have more than sensors located in other regions When high pressure, temperature, and humidity values ​​are measured, the patient is lying face down. It can be perceived that the patient is in the position of the bed / bed pad (110). The pressure, temperature, and humidity values ​​applied by each area of ​​the body in contact are 5 This will be measured. It will show how long each area of ​​the patient's body has been affected and by what means. that it contacts the bed / mattress pad (110) at the pressure, temperature and humidity values ​​given. Measurement data is critical for determining whether a pressure ulcer has formed in the affected area. It provides data. The application concerns a system for the early diagnosis of pressure ulcers (100) and also the patient a small camera to capture the image of the face (120); the patient's vocal expressions It includes at least one microphone (130) to capture the patient's face via the camera (120). Changes in facial muscles and facial expressions were observed by capturing their images. It can be perceived. The patient's vocal expressions are detected via the microphone (130). can be detected. Here, vocal expressions especially include groaning, grunting, and sighing. These are expressions such as blowing and puffing. In the patient's facial muscles and facial expressions... The perception of changes and the capture of vocal expressions are particularly important for a patient experiencing discomfort. It is a sign that you are feeling pain. The discomfort or pain mentioned is a symptom of a pressure sore. It provides critical data for determining whether or not it has occurred. 20 The subject of the application is a system for the early diagnosis of pressure ulcers (100) and at least one procedure. It includes a processing unit. Here, the processing unit consists of pressure sensors, temperature sensors, and humidity sensors. data set containing data from sensors, camera (120) and microphone (130) pressure ulcer scanner 25 trained to detect whether a patient has a pressure ulcer. It is configured to run the diagnostic algorithm. A system for the early diagnosis of pressure ulcers in an application of the invention (100), pressure from its sensors, temperature sensors, humidity sensors, camera (120) and data received from the microphone (130) at selected time intervals as input and selected 30 at time intervals, the patient's body was labeled to determine whether or not there was a pressure ulcer. an artificial intelligence trained with a training dataset that includes regional information as output It includes a pressure ulcer diagnosis algorithm with a model. The pressure ulcer diagnosis used The algorithm has been trained with at least one artificial intelligence model. The dataset used is selected. from pressure sensors that measure in intervals, temperature sensors, humidity 35 from its sensors, the camera (120) which captures images at selected intervals and selected data received from the microphone (130) which captures vocal expressions in intervals and whether there is a bass ulcer It contains information about the labeled areas of the patient's body that are not present. Here, very much... For analyzing a multi-component dataset, Multi-Input CNN-LSTM (Multi-Input Convolutional Neural Network – Long Short-Term Memory) artificial intelligence model 5 This model uses different input channels for video and audio data. Each channel is created using a CNN (Convolutional Neural Network) based system. It performs feature extraction. Then, to model temporal dependencies, LSTM (Long Terrestrial Modeling) is used. Short-term memory layers are used. With the help of the Multi-Input CNN-LSTM model. The movements in the images captured by the camera (120) and the 10 captured by the microphone (130) The study analyzes how patterns formed by vocal expressions change over time. Alternatively, a combination of classical machine learning methods An ensemble voting artificial intelligence model is used. In the ensemble voting model... Random Forest, SVM (Support Vector Machine) and Voting Classifier artificial intelligence models By running them together, the output of each model is combined, and a decision is made as to whether or not there is a pressure ulcer. 15 It is determined by majority vote. A system for early diagnosis of pressure ulcers in an application of the invention (100), from data received from the camera (120) and / or microphone (130) at selected time intervals The outputs of a pain detection algorithm adapted to identify when a patient is experiencing pain are 20 a pressure ulcer created through an artificial intelligence model trained with a dataset containing It includes a diagnostic algorithm. A system for early diagnosis of pressure ulcers in an application of the invention (100), camera (120) changes in pixels in images captured over time 25 By magnifying the image, the patient can determine if they are experiencing pain based on the muscle movements in their face. trained with a dataset containing outputs of a pain detection algorithm adapted for its detection. created through an artificial intelligence model, in other words, incorporating it in its infrastructure, It includes a pressure ulcer diagnosis algorithm. For the pain detection algorithm, EVM (Eulerian) is preferably used. Video Magnification) algorithm can be used. 30 captured by camera (120) in EVM. Small movements and color changes in the images are magnified to make them visible. This ensures that the changes that occur in the pixels over time are brought into play. It is analyzed and amplified within the frequency range. The microscopic features invisible to the naked eye on the patient's face... Movements, involuntary muscle spasms, or twitches are clearly revealed with EVM. This allows for the precise detection of fine motor movements that may indicate pain. 35 6 A system for early diagnosis of pressure ulcers in an application of the invention (100), the patient Determining whether a patient is experiencing pain based on changes in frequency and intensity of vocal expressions. an artificial intelligence system trained with data containing outputs of a pain detection algorithm adapted for this purpose. It includes a pressure ulcer diagnosis algorithm created through an intelligence model. These 5 In practice, the pain detection algorithm is preferably also trained using an artificial intelligence model. Here, we will primarily use an adapted MFCC (Mel-Frequency Cepstral Coefficients) approach. Spectral curvature, energy and frequency of vocal expressions taken from the microphone (130) using Basic acoustic properties, such as those of its components, are extracted. MFCC analyzes how the human ear responds to frequencies. mimicking the response and the basic acoustic characteristics of sound signals in vocal expressions 10 It represents a compact form. The MFCC provides a short-term power output of the audio signal. by creating its spectrum, especially sounds such as groaning, grunting, sighing, blowing, and puffing. The patient's vocal expressions indicating that they are experiencing pain are characteristic. Spectral curvature. (spectral centroid), represents the center of the sound's "brightness" or spectral energy distribution. It is used to assess the presence of high-frequency components. This feature is 15 thanks to the high-frequency content of pain-related sounds and other neutral (non-pain-related) sounds It becomes possible to separate the sounds. Energy measurement determines the intensity of the sound signal. and especially enables the capture of sound levels that increase during moments of pain. Frequency components Spectral components indicate the frequencies at which sound is concentrated and the tone of the sound. It details its characteristics. Thanks to these inferences, each sound sample becomes a multi-dimensional 20 It is represented in the feature space. The inference of features in the aforementioned vocal expressions... Then, preferably CNN (Convolutional Analysis) for classifying examples in vocal expressions. Artificial intelligence models based on Neural Networks (NNETs) and / or GRUs (Gated Recurrent Units). It is used. In CNN-based models, visual patterns are particularly obtained through spectrograms. By capturing and using a two-dimensional transformation of vocal expression data, frequency and time 25 The relationships between the pieces of information are modeled. Painful sounds form a specific sequence. GRU-based In the model, the temporal differences in the vocal expression are necessary to detect the sequence in question. It resolves dependencies. In this application, the CNN-based model and the GRU-based model are used together. Used alone or in a hybrid form, tonal structure (intonation and pitch) in vocal expressions, Frequency distribution and dynamics (intensity changes) are analyzed. Thus, the 30 vocal expressions are analyzed. Pain-specific patterns are detected with high accuracy, allowing the patient to determine whether or not they are experiencing pain. It detects with a high degree of accuracy that one cannot perceive. A system for the early diagnosis of pressure ulcers in an application of the invention (100), environment Pain detection algorithm adapted for filtering noise with spectral gating 35 7 a print job created through an artificial intelligence model trained with data including its outputs It includes a wound diagnosis algorithm. This device is frequently encountered, especially in hospital settings. beeps, monitor sounds, alarm sounds, conversations, and other environmental noises, vocal expressions This can lead to serious interference in the analysis of the audio signal. Such unwanted noises can cause significant problems in the analysis of the audio signal. By suppressing components below certain thresholds in the frequency spectrum, only specific 5 spectral gating, which highlights key sounds in certain frequency ranges. is minimized. A system for early diagnosis of pressure ulcers in an application of the invention (100), in the patient It should generate a warning signal when a pressure ulcer is detected, and the warning signal should be 10 a processing unit configured to send data to a communication device containing a visual interface This application of the invention includes a visual interface for the healthcare professional who will be caring for the patient. This allows us to inform the patient that they have a pressure ulcer. A system for early diagnosis of pressure ulcers in an application of the invention (100), a mobile 15 The invention includes a communication device. This application of the invention enables healthcare professionals to care for the patient. The staff member can remotely monitor the patient's health status regarding the pressure ulcer. In a variation of the invention's applications involving a communication device, the patient's body It includes a warning signal that provides information about which area(s) have a pressure ulcer. 20 In this application of the invention, the warning signal is directed to which part of the patient's body? He has information that there is a pressure ulcer in the region(s). Thus, the communication device The label indicates which area(s) of the patient's body have a pressure ulcer. or it can be shown visually. A system for early diagnosis of pressure ulcers in an application of the invention (100), a bed a viscoelastic sponge located between the surface (111), the bed surface (111) and the sensors Includes mattress / mattress pad (110). Mattress / mattress pad (110) naturally includes ergonomics and It should contain materials suitable to meet medical requirements. On the one hand... While the aforementioned requirements are met, on the other hand, the sensors must be reliable. 30 close to the bed surface (111) where the patient’s body will come into contact in order to take measurements They must be positioned in such a way that these two opposing conditions can be met simultaneously. It will be placed between the sensors and the patient's body, supporting the patient's body. Choosing materials that ensure ergonomics is of utmost importance. (mentioned) In order to ensure the best possible conditions for this application of the invention, a viscoelastic 35 8 Sponge was used. Viscoelastic sponge between the mattress surface (111) and the sensors. It is positioned. The viscoelastic foam has a structure that responds to body temperature and pressure. It has the aforementioned structure of viscoelastic foam, which allows sensors to physically engage the patient. It prevents discomfort. In addition, the viscoelastic foam reacts to body temperature, regulating body heat. By quickly transmitting the information to the sensors, it supports the measurement sensitivity of the temperature sensors. Visko 5 The elastic foam molds to the patient's sleeping position, thus improving pressure distribution. It facilitates the acquisition of data. Viscoelastic sponge in one application of the invention. It is essentially 7cm thick. A system for early diagnosis of pressure ulcers in an application of the invention (100), 10 mattress surface which is antibacterial, liquid-impermeable and / or flame-retardant fabric (111) It includes. Suitable for medical use thanks to the antibacterial properties of the mattress surface (111). a mattress / mattress pad (110) is made. The mattress surface (111) is liquid impermeable Its structure prevents sensors and electronic components from coming into contact with liquids. This is prevented. Thanks to the non-flammable structure of the bed surface (111), any 15 that may occur in the sensors is prevented. Patient safety is guaranteed even in the event of any malfunction. A system for early diagnosis of pressure ulcers in an application of the invention (100), I2C It includes temperature and humidity sensors that support the communication protocol. This application of the invention preferably uses temperature sensors of type AHT15 and / or SHT31 and 20 It includes humidity sensors. With this application of the invention, each temperature in a distributed manner... The sensor and humidity sensor can be connected to each other. The purpose here is I2C communication. each through a network of sensors connected to each other via a protocol This allows measurements to be taken from both the temperature sensor and the humidity sensor. Thus, each 25 no need for it. Thus, the advantages of low cost, high reliability and long service life are achieved. is obtained. A system for early diagnosis of pressure ulcers in an application of the invention (100), force It includes pressure sensors exhibiting sensitive resistance variability. This application of the invention is 30 It preferably includes FSR-406 and / or RP-S40-ST type pressure sensors. A system for early diagnosis of pressure ulcers in an application of the invention (100), pressure sensors, temperature sensors distributed throughout the bed / mattress pad (110) and humidity sensors are electronically connected and analog signals are received from these sensors 35 9 signals are converted to digital, pressure sensors, temperature sensors and humidity sensors. sensors enable the transfer of measurement data to the processing unit, connecting to each other. It contains electronically connected voltage divider circuit cells. Voltage divider circuit cells preferably to form a grid shape along the bed / mattress pad (110) They are distributed. The distribution forms a grid, with 5 voltage divider cells. It is used in the sense of being distributed in a way that forms columns and pillars. Voltage Analog data from pressure sensors is converted into digital data via divider circuit cells. It is converted and transferred to the processing unit via the I2C communication protocol. In a sample application of the invention, each voltage divider circuit cell has 4 pressure sensors. It includes sensors, and 2 each of temperature and humidity sensors. 10 A system for early diagnosis of pressure ulcers in an application of the invention (100), a stand (140) on which the camera (120) and / or microphone (130) is positioned Includes. Stand (140) camera (120) and / or microphone (130) depending on the patient. To facilitate its positioning, the stand (140) should be moved on a floor. 15 It may include wheels (141) that facilitate.

Claims

REQUESTS 1. - at least one bed / mattress pad (110); located on the bed / bed pad (110), - the areas of a patient's body applied to the bed / bed pad (110) pressure sensors for measuring pressure; 5 - the areas of the patient's body formed on the bed / bed pad (110) temperature sensors to measure temperature and humidity sensors to measure humidity. sensors; a system for early diagnosis of pressure ulcers (100) and its feature - a small camera to capture the image of the patient's face (120); 10 - at least one microphone to capture the patient's vocal expressions (130); - from pressure sensors, temperature and humidity sensors, camera (120) and from the data set containing data received from the microphone (130) whether the patient has a pressure ulcer an algorithm adapted to diagnose pressure ulcers in order to determine if it is not present It must contain at least one processing unit configured for operation. 15 2. A system for the early diagnosis of pressure ulcers as in Claim 1 (100) and its feature from pressure sensors, temperature sensors, humidity sensors, data received from the camera (120) and microphone (130) at selected time intervals As input, and at selected time intervals, whether or not there is a pressure ulcer 20 a list containing information about the labeled areas of the patient's body as output a press created through an artificial intelligence model trained with a training dataset It includes a wound diagnosis algorithm.

3. As with any of the previous requests, a 25 for early diagnosis of pressure ulcers The system (100) has a feature selected from the camera (120) and / or microphone (130). to determine if the patient is experiencing pain from data collected at different time intervals an artificial intelligence system trained with a dataset containing the outputs of an adapted pain detection algorithm. It includes a pressure ulcer diagnosis algorithm created through an intelligence model.

4. A system for the early diagnosis of pressure ulcers as in Claim 3 (100) and its feature Pixel changes in images captured by camera (120) over time by magnifying the incoming changes according to the muscle movements of the patient's face a pain detection algorithm adapted to detect when a patient is feeling pain 11 through an artificial intelligence model trained with a dataset containing its outputs It includes a developed pressure ulcer diagnosis algorithm.

5. A system for the early diagnosis of pressure ulcers, as in any of claims 3-4. (100) and its characteristic is the changes in frequency and intensity of the patient's vocal expressions. According to the pain detection algorithm, it is adapted to detect when a patient feels pain. through an artificial intelligence model trained with a dataset containing its outputs It is a run of a created pressure ulcer diagnosis algorithm.

6. A system for the early diagnosis of pressure ulcers as in claim 5 (100) and its feature is 10 Pain adapted for filtering ambient noise via spectral gating. an artificial intelligence model trained with a dataset containing the outputs of the detection algorithm It is a run of a pressure ulcer diagnosis algorithm created through this method.

7. As with any of the previous requests, a 15-day period for the early diagnosis of pressure ulcers. The system (100) has the feature that when a pressure ulcer is detected in the patient, It generates a warning signal and communicates the warning signal through a visual interface. It contains a processing unit configured to send data to the device.

8. A system for the early diagnosis of pressure ulcers as in claim 7 (100) and its feature is 20 It includes a communication device, which is a mobile device.

9. A system for the early diagnosis of pressure ulcers, as in claim 7 or claim 8. (100) and its characteristic is that there is pressure on which part / parts of the patient's body It contains a warning signal indicating that there is an injury. 25 10. As with any of the previous requests, for the early diagnosis of pressure ulcers The system (100) has a bearing surface (111), bearing surface (111) a mattress / bed pad containing a viscoelastic foam among the sensors (110) includes. 30 11. A system for the early diagnosis of pressure ulcers (100) as in claim 10, and its feature mattress surface that is antibacterial, liquid-repellent and / or flame-retardant fabric (111) is included. 35 12 12. As with any of the previous requests, for the early diagnosis of pressure ulcers... The system (100) has the feature of supporting the I2C communication protocol and temperature. It includes sensors and humidity sensors.

13. As with any of the previous requests, a 5 for early diagnosis of pressure ulcers. The system (100) has the characteristic of being a pressure that exhibits resistance variability sensitive to force. It includes sensors.

14. As with any of the previous requests, for the early diagnosis of pressure ulcers... The system (100) has the feature that the pressure is distributed throughout the bed / mattress pad (110) 10 sensors, temperature sensors and humidity sensors electronically connected, analog signals received from these sensors are converted to digital, Measurement of pressure sensors, temperature sensors, and humidity sensors. electronically connected devices that enable the transfer of data to the processing unit. It contains voltage divider circuit cells. 15 15. As with any of the previous requests, for the early diagnosis of pressure ulcers... The system (100) has the feature of the camera (120) and / or microphone (130) on it. It includes a stand (140) in which it is positioned.

Images