Technological Area:

This invention is about an artificial intelligence method developed to enable the automatic determination of the level of comfort and pain of babies.

Current Situation of The Technique:

Comfort is a process of meeting the individual’s physical, psychological, social and environmental needs. Every age group needs a comfortable life. Providing comfort especially for newborn babies to ensure their growth is not negatively impacted, is much more important. Newborn babies are in a quiet, dark, warm environment inside the mother’s womb with a high comfort level and away from stressors (stress causing circumstances). But some newborns will be taken to newborn intensive care unit (NICU) after birth for various reasons. NICU is an environment with many external stressors (stress causing circumstances) and much more noisy considered to the life inside the mother’s womb. Routine nursing and invasive operations made in these units are causing comfort to lessen even more for babies which are not developed enough yet. And comfort levels being interfered prolongs the recovery time and hospital stay of babies. For this reason, applications intended to ensure comfort for newborns in NICUs, are getting more and more important in recent years. Important tasks fall upon nurses especially who spend the most time with babies in this matter. Nurses take part in determining the patient and family’s comfort needs and increasing, assessing and maintaining their comfort levels. Increasing the comfort level of a patient with proper nursing initiatives at NICUs, is an essential part of professional nursing care. Among these practices are giving glucose or sucrose orally to reduce pain and stress, kangaroo care, massage, non-nutritive sucking, positioning, swaddling etc. Following the practices made, assessing the baby’s comfort level is quite important. Comfort scales are utilized for this. One of the mostly used scales to assess the comfort level is the “Newborn Comfort Behavior Scale (NCBS)”. This Likert type scale is used to assess the sedation and comfort needs of the newborn babies. Comfort Scale (CS) was first created by Ambuel and et al. (1992) to assess the distress of the newborns who get mechanical ventilator (ventilatory support device) at NICU. Afterwards it was revised by Van Dijk et al. (1992), examined for its’ validity reliability as the COMFORTneo scale without physiological parameters only to scale the behavior in newborn babies and designated that its Cronbach alpha coefficient was between 0.84-0.88. Turkish validity reliability of the COMFORTneo scale was made by Kahraman et al. in 2014. The Cronbach alpha coefficient of the scale was found to be 0.85 prior to operation and 0.92 after the operation. The scale is used in babies born between the 26th-42nd weeks. Newborn Comfort Behavior Scale (NCBS) is consisted of seven parameters such as alertness, sedation/agitation, respiratory response, crying, body movements, face strain and muscle tone. In newborns having mechanical ventilator support “Respiratory Response”, in newborns not having mechanical ventilator “Crying” is taken into account. If the total point of the scale is below 14, it will be pointed out that the baby has a pain or distress, is comfortless and need initiations which provide comfort. Despite all its’ benefits, it’s seen that NCBS are not routinely used in clinics. One of the main reasons of this is considered the nurses seeing this as a waste of time and unnecessary work load. At this point it’s important to develop an application which will make the use of NCBS practical in terms of ensuring the follow-up of the comfort of newborn babies at regular intervals. But there have been no studies found that will ease the use of this kind of scales at newborn units.

On of the most important circumstances that reduces the comfort level of newborn babies and affect their recovery process, is pain. In adult patient groups that can communicate verbally and make themselves understand with ease, the existence of pain can be easily determined. But for those who can’t communicate verbally, especially for newborn babies, determining the pain is difficult. Newborn babies are exposed to many painful operations as of their birth such as heel lance, opening vascular access, vaccination, mechanical ventilator (ventilator support device) application. Today there are pain scales available in order to overcome the difficulty of assessing pain in this age group and they’re requested to be used by many authorities with the Turkish Society of Neonatology coming at the first place. But these scales requiring more than one parameter to be involved, are not seen to be used in clinics largely. Therefore, many babies are not able to be noticed and recovered by the personnel even though they have pain. And this causes several short-time and long-time unfavorable things in the baby. Some of these are; the dragging of the recovery time of the baby, emerging comorbidities, mental retardation at further ages, psycho-social disorders and neuro motor disorders. To prevent these harms caused by pain to the baby, existing pain scales need to be arranged for easier use and so their level of use must be increased. There are many scales developed to assess the pain of newborn babies. Some of these scales measure the acute, some measure the chronic pain and while some of them measure the pain of the babies who born normally, the others measure the pain of those who bom prematurely. This causes confusion due to the requirement of using a different scale for each baby. On the other hand, a scale which can be used both for normal and premature born babies, measures both the acute and the chronic pain and can be used in babies connected to a respirator, the Newborn Pain, Agitation and Sedation Scale (N- PASS), enables the convenience of using a single scale in clinic. Turkish Society of Neonatology also recommends the use of N-PASS. N-PASS was developed in 2003 and adapted to Turkish in 2011 by Acikgoz et al. This scale is used in babies between 0-100 days old. Scale has two separate parts measuring the sedation and the pain level of the baby. Scale has 5 sub parameters. These are; crying and disturbance, behavior-status, facial expression, hands and feet and body strain and signs of life. Determining the patient’s pain score with artificial intelligence instead of the nurse assessing each parameter one by one, will provide great convenience in practice. Today “Studies on Artificial Intelligence” become more and more frequent in health area. But these studies are conducted mostly by physician and engineer cooperation and intended for diagnosis. In national and international literature searches, there have been no comprehensive studies found on artificial intelligence that involves also the nurses or makes it easy for them to provide care. Yet nurses are the healthcare personnel who spend the most time with the patient to provide 24 hours of non-stop service. With this artificial intelligence developed it’s planned to reduce the work load of clinics, rapidly determine the pain and comfort levels of newborn babies, quickly respond to those with pain or low comfort level and reduce the harms to the baby. This study will be important and a pioneering one by means of being the first study on artificial intelligence for nursing. In the literature search made, there have been no studies found made on this matter.

Applications found in the literature, intended to determine the pain or comfort level, are given below.

In the application issue numbered CN106778657A, the “Newborn pain expression classification method based on CNN (Convolution Neural Network)” is described. The method subject to invention includes the following steps: first, the newborn pain expression images are obtained and classified by healthcare professionals as sedated, crying, mild pain and intense pain and a newborn pain expression image library is created, secondly, the patterns in the newborn pain expression image library are perceived as data inputs by the CNN which includes a data layer, three convolutional layers, two full-connection layers and a classification layer, the repeated training is conducted with a back prop algorithm on CNN and the training is made to be optimized with global parameters to reduce the network output loss value.

In the application issue numbered US2018160905A1, the “Signal receiving method and system” is described. Current invention provides a signal receiving method and system that includes signal pick-up, analysis, calculation and output functions. System can be configured to get vital signs: it obtains and outputs the vital signs by picking up two or more physiological signals, calculating the relation between two or more physiological signals and calculating the vital signs based on relation info.

In the application issue numbered CN106778657A described in brief above, only the newborn baby’s crying situation is assessed. And in the application issue numbered US2018160905A1, it’s remarked that physiological parameters are assessed. These applications involve only the behavioral parameters and a limited number of physiological parameters. And this does not provide an adequate determination of the comfort and pain situations of babies.

In conclusion there’s a need for a new technology that can overcome the disadvantages mentioned above, process heart rate, respiration rate, blood pressure, oxygen saturation data and sound and image data obtained from the camera and the microphone and rapidly and simply determine the comfort and pain levels of newborn babies, easy to use, saves time and can be used also through a mobile device.

The Invention Definition:

This invention is an artificial intelligence method that enables assessment of pain and comfort level of newborn babies and its’ characteristic is being a new technology that processes heart rate, respiration rate, blood pressure, oxygen saturation data and sound and image data obtained from the camera and the microphone and rapidly and simply determine the comfort and pain levels of newborn babies, easy to use, saves time and can be used also through a mobile device.

In order to realize all purposes mentioned above and offered by the detailed description below, the invention; takes the Newborn Comfort Behavior Scale (NCBS) and the Newborn Pain/ Agitation, Sedation Scale (N-PASS) used at newborn units in hospitals and makes them to be used more practically by healthcare personnel to promote its’ use for all newborn babies.

One of the most important ways to realize the purposes mentioned above is the “Studies on Artificial Intelligence”. Instead of the nurses assessing each parameter one by one, directly determining the patient’s comfort and pain level with artificial intelligence technology, provides a great convenience in practicing. The invention enables both lessening the workload of clinics and determining the pain and comfort level more rapidly and simply. In this context, the invention aims to assess the comfort and pain levels of newborns with artificial intelligence technique.

Thanks to the invention, nurses are not spending an extra time and effort to fill up NCBS and N-PASS. So, comfort and pain levels of babies can be determined without delay and in time practices can be made to increase the comfort level and decrease the pain level of babies with low comfort levels and experiencing a pain. In addition, with the artificial intelligence program developed regarding the invention, studies can also be conducted in different clinics and in different medical interventions.

The invention is important and is pioneering both by lessening the workloads of clinics and harm to babies with rapid responses due to fast determination of comfort and pain levels and by being the first study on artificial intelligence for nursing.

All the advantages of the method subject to invention will be understood more clearly with the help of the scheme given below and the detailed description written by referring to this scheme, for this reason assessment should be made considering this scheme and the detailed description. Description of Figures:

The invention will be described by referring the figures enclosed, this way the characteristics of the invention will be understood and assessed more clearly, but this is not intended to limit the invention with these specific arrangements. Vice versa it’s intended to cover all alternatives, changes and equivalents which can be involved under the area the invention is defined in by the claims enclosed. It must be understood that the shown details are given only for the description of the preferred arrangements of the current invention and presented to provide the most utilizable and clear definition of both the formation of the methods and the rules and conceptual characteristics of the invention.

Figure 1: Schematic view of the operation of invention. Sub steps of the Figure 1 helping to get this invention be understood, are numbered as stated in the attached figure and given below by their names.

Description of References: 1. Data Module

2. Artificial Intelligence Module

2.1. Comfort Assessment Module

2.2. Pain Assessment Module The Description of Invention:

The invention include a data module (1) where to obtain data sets such as image, sound, heart rate, oxygen saturation, blood pressure, respiration rate, an artificial intelligence module (2) where obtained records are processed with artificial intelligence algorithms, a comfort assessment module (2.1) where the comfort level of artificial intelligence results is showed and a pain assessment module (2.2) where the pain level of artificial intelligence results is showed.

The invention; has a data module (1) where parameters such as alertness, sedation/agitation, respiratory response, crying, body movements, facial strain, muscle tone, heart rate, oxygen saturation, blood pressure, respiration rate are obtained and recorded.

The invention; has an artificial intelligence module (2) where the respiratory response in newborn babies having mechanical ventilator (ventilation device) support) and crying sounds in those not having mechanical ventilator (ventilation device) support), are assessed.

The invention; has an artificial intelligence module (2) where data obtained from the facial images and body movements of newborn babies, are recorded into the data module (1) and pain and comfort levels are determined. The invention; has an artificial intelligence module (2) where the newborn baby’s pain and comfort levels are determined with the data obtained from the camera and the microphone and recorded into the data module (1).

The invention has an artificial intelligence module (2) where pain and comfort levels are determined by using the heart rate, oxygen saturation, blood pressure, respiration rate data.

The invention; has an artificial intelligence module (2) which designates the reference points on transferred images with image processing algorithms and conducts the sectioning according to the designated reference points, conducts the training with one of the system’s deep learning techniques, convolution neural network (CNN), as a result of the obtained sections and determines the pain and comfort levels. The invention; has an artificial intelligence module (2) that classifies sound records with one of the deep learning techniques, long short-term memory (LSTM) networks.

The invention; has an artificial intelligence module (2) where the newborn baby’s pain and comfort status result is obtained with the estimation results of the convolution neural networks (CNN) and the long short-term memory (LSTM) networks and machine learning algorithms of heart rate, oxygen saturation, blood pressure, respiration rate values.

The invention; has an artificial intelligence module (2) that uses N-PASS scale to designate the pain level, which is the Agitation and Sedation Scale of the data in the data module.

The invention; has a comfort assessment module (2.1) where the assessment result of the comfort level is specified.

The invention; has a pain assessment module (2.1) where the assessment result of the pain level is specified.

Detailed Description of Invention:

The system subject to invention, basically consists of the data module (1), the artificial intelligence module (2), the comfort assessment module (2.1) and the pain assessment module (2.2). The invention begins to obtain the baby’s image and sound records through the data module (1) in order to determine the baby’s comfort and pain levels with camera after the data collecting form is filled up by the parent in control assessment. After the obtained records are transferred to the computer, the baby’s sound record is separated in computer environment later to be assessed in the artificial intelligence module (2). The result of the comfort assessment is indicated in the comfort assessment module (2.1) and the result of the pain assessment is indicated in the pain assessment module (2.2). To create the comfort assessment module (2.1), parameters such as alertness, sedation/agitation, respiratory response, crying, body movements, facial strain, muscle tone are paid attention to be visible in order to determine the comfort levels of the baby while taking the image of the baby. Newborn Comfort Behavior Scale (NCBS) is consisted of seven parameters such as; alertness, sedation/agitation, respiratory response, crying, body movements, facial strain, muscle tone. In babies having mechanical ventilator support “Respiratory Response”, in babies not having mechanical ventilator “Crying” is taken into account. If the scale total score is below 14 it’s remarked out that the baby is comfortless and need comfort providing interventions. The scale including these parameters is given below in Table-1. In the method subject to invention, nurses directly access the patient’s comfort behavior score with artificial intelligence technology without assessing each parameter one by one.

Table-1 To create the pain assessment module (2.2), parameters such as crying and disturbance, behavior-status, facial expression, hand and feet and body strain and vital signs are paid attention to be visible in order to determine the comfort levels of the baby while taking the image of the baby. Because the scale has 5 sub parameters. These are; crying and disturbance, behavior-status, facial expression, hand and feet and body strain and vital signs. The scale has two separate sections measuring the sedation level and the pain level of the baby. In the method subject to invention, the pain assessment is made in the section measuring the pain level with the 5th vital sign. In every situation assessing the vital signs, pain is also assessed. The pain is scored from 0 to +2 for each behavioral and physiological criterion and summed up later. This scoring system is shown in Table-2. Because premature babies’ behavioral response capacities are limited, scores are added according to the gestation weeks. If the baby is below 30 gestation weeks +1 point will be added. Total pain score is indicated with positive numbers between 0 to +11. Intervention is advised for known pains/painful stimuli before the score reaches to 3. The purpose of the pain treatment/intervention is maintaining the score at 3 or below.

In the method subject to invention, measuring the sedation level is scoring the assessment of the baby’s response to stimuli for each behavioral and physiological scales in addition to the pain. There’s no need to use sedation assessment in every pain assessment. Each behavioral and physiological scale is scored from 0 to -2 in sedation assessment and indicated as a negative score (from 0 to -10). If there isn’t any sedation in the baby 0 point will be given but this does not mean insufficient reaction. Total score obtained in deep sedation is between -5 and -10 while it’s -2 and -5 in mild sedation. A negative total score without using sedatives/ opioid medications, shows that baby has a pain or in a situation causing stress. In the method subject to invention, the Newborn Agitation and Sedation Scale (N-PASS) in creating pain assessment module (2.2) is given below in Table-2 and in this method, nurses directly access the patient’s pain score with the artificial intelligence technology without assessing each parameter one by one. Table-2

NPASS: Newborn Pain, Agitation and Sedation Scale

Loyola University Health System, Loyola University Chicago, 2009

(Rev 2/10/09) Pat Hummel, MA, APN, NNP, PNP

All rights reserved. It’s prohibited to reproduce any part of this document by electronical or mechanical means without the written consent of the author. Author _ _ _ shall not be held responsible of all kinds of conclusions or damages or failures if any arising from interpretation or execution of this material. With the method subject to invention, pain and comfort assessment is made with an artificial intelligence module (2) instead of a nurse. This way it both lessens the workload of nurses and ensures them to work more productively. Besides it determines the existing case more accurately in a short time by taking many parameters as a basis.

Data sets like image, sound and vital signs mentioned above, are made processable by artificial intelligence module (2). Determining the reference points of the baby’s images to allow these data sets to be processed with image processing algorithms, takes place in the computer environment. With the obtained images by making the sectioning operation according to determined reference points, the system is trained by one of the deep learning techniques, the convolution neural network (CNN). And sound records are classified by again one of the deep learning techniques, long short-term memory (LSTM) networks. With the estimation results of the convolution neural networks (CNN) and long short term memory networks and machine learning algorithms of pulse oximetry values, artificial intelligence module (2) is created where the baby’s comfort status is obtained, and newborn baby’s comfort level is indicated in the comfort assessment module (2.1). With the estimation results of the convolution neural networks (CNN) and long short term memory networks and machine learning algorithms of heart rate, blood pressure, oxygen saturation, respiration rate values, artificial intelligence module (2) is created where the baby’s pain status is obtained, and newborn baby’s pain level is indicated in the pain assessment module (2.2).