An information tracking and recording system for pediatric in-hospital nursing

Abstract

The present invention relates to the field of medical data processing technology, and in particular to an information tracking and recording system for pediatric in-hospital nursing, the system comprising: an acquisition module for acquiring sound curve information; a determination module for determining the possibility that a target user has not made a sound in the current sound analysis window of a collector; a determination module for determining a sound analysis window to be determined; a determination module for determining whether a sound signal is a noise signal; a fitting module for fitting the noise signals of each collector to obtain a noise signal curve for a period of no sound; a correction module for correcting the average sound curve information of the target user at each moment to obtain a corrected sound curve; and an identification module for inputting the corrected sound curve into an identification model to identify the physiological needs of the target user. In this way, the present invention improves the noise reduction effect of the child's sound and accurately and effectively identifies the child's physiological needs.

Description

Technical Field

[0001] The invention relates to the technical field of medical data processing, and in particular to an information tracking and recording system for pediatric in-hospital nursing. Background Art

[0002] In the field of pediatric nursing, its core task is to conduct in-depth research on the growth and development of children, and to do a good job in the prevention and treatment of children's diseases and health care. Children in infancy and toddler period cannot actively and clearly express their intentions because their language ability is not fully developed. Therefore, caregivers need to devote a lot of energy to pay close attention to the physiological condition of the children in real time. Traditionally, it is generally believed that the crying of children can directly reflect their needs, and caregivers mainly judge the needs of children by identifying crying. For example, crying with strong regularity and continuity, gradually increasing from low to high, and accompanied by sucking reflex sounds, most likely means that the child is hungry. Crying that occurs suddenly, is low and intermittent, often indicates that the child's diaper may be wet or dirty. Sudden and sharp crying indicates that the child may be in pain or discomfort. However, relying solely on the voice of the child (especially crying) to judge his physiological needs places extremely high demands on medical staff. They must always be highly vigilant, otherwise it is very easy to have problems such as untimely care, which significantly increases the difficulty of manual care.

[0003] In some scenarios, real-time monitoring and collection of crying information of sick children to identify their physiological needs has become an effective way to reduce the difficulty of nursing. When collecting the voices of sick children, traditional solutions usually use methods such as frequency decomposition to reduce the noise of the children's voices recorded by the collector. Although this general noise reduction method can obtain the crying sounds of sick children to a certain extent, due to the large number of equipment and instruments in the ICU, the complex sound sources, and the inclusion of sounds of multiple frequencies, the noise reduction effect of the children's voices is unsatisfactory, and the physiological needs of the children cannot be accurately and effectively identified. Summary of the invention

[0004] In order to solve the technical problem that the noise reduction effect of children's voices is poor and the physiological needs of children cannot be accurately and effectively identified, the purpose of the present invention is to provide an information tracking and recording system for pediatric in-hospital nursing. The technical solutions adopted are as follows:

[0005] In a first aspect, an embodiment of the present invention provides an information tracking and recording system for in-hospital nursing care in pediatrics, comprising: an acquisition module for acquiring sound curve information of a target user at each moment collected by each collector in a pediatric hospital; a determination module for determining the possibility that the target user has not made a sound in the current sound analysis window of the collector according to the sound information of the sound curve information in the sound analysis window; the determination module is also used to determine, when the possibility is greater than a first threshold, a to-be-determined sound analysis window in which the target user has not made a sound in the current sound analysis window; the determination module is also used to determine, based on the signal overlap of each to-be-determined sound analysis window, whether the sound signal of each sampling point value in the to-be-determined sound analysis window in which the target user has not made a sound is a noise signal; a fitting module is used to fit the noise signals of each collector in the case of a noise signal to obtain a noise signal curve for a period of no sound; a correction module is used to correct the average sound curve information of the target user at each moment based on the noise signal curve to obtain a corrected sound curve; and an identification module is used to input the corrected sound curve into a recognition model to identify the physiological needs of the target user, so that the nursing staff can provide care for the target user based on the physiological needs.

[0006] Optionally, the determination module is also used to determine the sound energy of the sound analysis window of the sound curve information of the collector based on the sound information in the sound analysis window; and to determine the possibility that the target user has not made a sound in the current sound analysis window of the collector by utilizing the sound energy of the target user in the current sound analysis window and the sound energy of the sound analysis window adjacent to the current sound analysis window.

[0007] Optionally, the determination module is further used to superimpose the sound information of the sound curve information in the sound analysis window to obtain the sound energy of the sound analysis window of the sound curve information.

[0008] Optionally, the determination module is also used to calculate the absolute value of the first difference between the sound energy of the current sound analysis window and the sound energy of each adjacent sound analysis window before the current sound analysis window, and superimpose the absolute values ​​of each first difference to obtain a first superimposed value; determine a first number in which the first difference between the sound energy of the current sound analysis window and the sound energy of each adjacent sound analysis window before the current sound analysis window is a positive number, and a second number of each adjacent sound analysis window before the current sound analysis window; calculate a second difference between the second number and a predetermined value; determine a first ratio between the first number and the second difference; determine a first product between the first superimposed value and the first ratio as the degree of increase of the sound energy of each adjacent sound analysis window before the current sound analysis window; normalize the first difference to obtain a first normalized value, and calculate a second product between the first normalized value and the degree of increase; normalize the second product to obtain a possibility.

[0009] Optionally, the determination module is also used to determine the signal overlapping part of each sound analysis window to be determined, calculate the variance of the time interval of each sampling point value in the signal overlapping part; determine the probability that the sound signal of each sampling point value in the sound analysis window to be determined when the target user does not make a sound is a noise signal based on the variance; when the probability is greater than a second threshold, determine that the sound signal of the sampling point value is a noise signal.

[0010] Optionally, the determination module is further used to calculate a first sum between a predetermined value and a variance; and determine the reciprocal of the first sum as the probability.

[0011] Optionally, the fitting module is also used to, in the case of noise signals, superimpose the noise signals of each collector into the same sequence in chronological order, and use the least squares method to fit the noise signals in the sequence to obtain the noise signal curve of the silent period.

[0012] Optionally, the correction module is also used to calculate the average value of the sound curve information collected by each collector to obtain the average sound curve information; extend and expand the noise signal curve until it has the same length as the sound curve information to obtain a corrected noise signal curve; and subtract the average sound curve information from the corrected noise signal curve to obtain a corrected sound curve.

[0013] Optionally, the recognition module is also used to input the modified sound curve into the VGGish recognition model to identify the physiological needs of the target user; send the physiological needs to the target terminal so that the caregiver can provide care for the target user based on the physiological needs, and record the care information of the target user at the target terminal.

[0014] In a second aspect, an embodiment of the present invention provides another information tracking and recording system for pediatric in-hospital nursing, comprising: a processor and a memory; wherein the memory is used to store a computer program that can be run on the processor; the processor is used to execute the program stored in the memory to implement the following steps: obtaining sound curve information of the target user collected by each collector in the pediatric hospital at each moment; determining the possibility that the target user has not made a sound in the current sound analysis window of the collector based on the sound information of the sound curve information in the sound analysis window; when the possibility is greater than a first threshold, determining a to-be-determined sound analysis window in which the target user has not made a sound in the current sound analysis window; determining whether the sound signal of each sampling point value in the to-be-determined sound analysis window in which the target user has not made a sound is a noise signal based on the signal overlap of each to-be-determined sound analysis window; in the case of a noise signal, fitting the noise signal of each collector using the least squares method to obtain a noise signal curve for the non-sound period; based on the noise signal curve, correcting the average sound curve information of the target user at each moment to obtain a corrected sound curve; inputting the corrected sound curve into the recognition model to identify the physiological needs of the target user, so that the nursing staff can provide care for the target user based on the physiological needs.

[0015] The present invention has the following beneficial effects: first, an acquisition module acquires the sound curve information of the target user collected by each collector in a pediatric hospital at each moment; then, a determination module determines the possibility that the target user has not made a sound in the current sound analysis window of the collector according to the sound information of the sound curve information in the sound analysis window; when the possibility is greater than a first threshold, a to-be-determined sound analysis window in which the target user has not made a sound in the current sound analysis window is determined; and according to the signal overlapping part of each to-be-determined sound analysis window, it is determined whether the sound signal of each sampling point value in the to-be-determined sound analysis window in which the target user has not made a sound is a noise signal; secondly, a fitting module fits the noise signal of each collector in the case of a noise signal to obtain a noise signal curve of the non-sound period; then, a correction module corrects the average sound curve information of the target user at each moment based on the noise signal curve to obtain a corrected sound curve; finally, the recognition module inputs the corrected sound curve into the recognition model to identify the physiological needs of the target user, so that the nursing staff can provide care for the target user based on the physiological needs.

[0016] In this way, the embodiment of the present invention can determine the target user's unvoiced sound analysis window based on the target user's sound information, and based on the noise signal in the target user's unvoiced sound analysis window. And fit the noise signals of each collector to obtain the noise signal curve of the unvoiced period, use the noise signal curve to correct the average sound curve information of the target user at each time, obtain the corrected sound curve, and finally use the recognition model to identify the target user's physiological needs based on the corrected sound curve. In this way, the target user's sound signal is denoised by the target user's sound characteristics, which improves the noise reduction effect of the child's sound and accurately and effectively identifies the child's physiological needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 A schematic diagram of the structure of an information tracking and recording system for pediatric in-hospital nursing provided by one embodiment of the present invention;

[0019] Figure 2 A schematic diagram of energy of a target user's voice provided by an embodiment of the present invention;

[0020] Figure 3 A schematic diagram of the structure of an information tracking and recording system for pediatric in-hospital nursing provided by another embodiment of the present invention. DETAILED DESCRIPTION

[0021] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the information tracking and recording system for pediatric in-hospital nursing proposed by the present invention, its specific implementation, structure, features and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0022] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0023] The specific scheme of the information tracking and recording system for pediatric in-hospital nursing provided by the present invention is described in detail below in conjunction with the accompanying drawings.

[0024] Embodiment 1:

[0025] See also Figure 1 , which shows a schematic diagram of the structure of an information tracking and recording system for pediatric in-hospital nursing provided by an embodiment of the present invention, such as Figure 1 As shown, the information tracking and recording system 100 for pediatric in-hospital nursing includes: an acquisition module 101, which is used to acquire the sound curve information of the target user collected by each collector in the pediatric hospital at each moment; a determination module 102, which is used to determine the possibility that the target user has not spoken in the current sound analysis window of the collector according to the sound information in the sound analysis window of the sound curve information; the determination module 102 is also used to determine the to-be-determined sound analysis window in which the target user has not spoken in the current sound analysis window when the possibility is greater than a first threshold; the determination module 102 is also used to determine the to-be-determined sound analysis window in which the target user has not spoken according to the signal of each to-be-determined sound analysis window The overlapping part determines whether the sound signal of each sampling point value in the to-be-determined sound analysis window of the collector when the target user does not make a sound is a noise signal; the fitting module 103 is used to fit the noise signal of each collector in the case of a noise signal to obtain the noise signal curve of the silent period; the correction module 104 is used to correct the average sound curve information of the target user at each moment based on the noise signal curve to obtain a corrected sound curve; the recognition module 105 is used to input the corrected sound curve into the recognition model to identify the physiological needs of the target user, so that the nursing staff can provide care for the target user based on the physiological needs.

[0026] Specifically, in the embodiment of the present invention, collectors can be installed at the four corners of the child care bed, generally not less than 2, respectively located at the head and tail of the bed, and the operating frequency of each collector is generally not less than 10kHz, and the collection frequency is the same. The collected sound data is transmitted to the in-hospital server via wired or wireless means, and the in-hospital server performs the operation of the subsequent modules in the embodiment of the present invention to identify the physiological needs and other information of the target user. Among them, the target user can be children, infants, and adults.

[0027] Furthermore, in the nursing room in the hospital, the noise source is usually the nursing equipment distributed outside the bed. The noise generated by these nursing equipment over a period of time is generally stable and continuous. On the contrary, the sound generated by the target user itself is generally intermittent due to factors such as its lung capacity and physical condition. For example, when the target user is an infant, when the infant feels hungry, the crying sound is generally spaced 0.5s to 1s apart. During this intermittent time, the sound generated by the infant itself is low. Because the degree of sound attenuation in the air is generally inversely proportional to the square of the distance, when the sound emitted by the infant itself is small, the source of the difference in its sound is more likely to be caused by the noise generated by the nursing equipment. Usually during a period of unattended time, since the noise source in the nursing room is relatively stable, the noise signal can be evaluated by the sound difference of each collector in the interval where the infant's sound is small, and then the noise signal of this period is used to correct the fragment containing more infant sound information, so as to achieve the purpose of obtaining the real sound of the infant.

[0028] Furthermore, the collector is generally located at the four corners of the nursing bed of the target user, and there are generally a large number of physiological monitoring devices around the nursing bed, such as electrocardiogram monitors, respiratory monitors, infusion pumps, sphygmomanometers, etc. The noises generated by each of them are different, but they are all relatively stable. Therefore, for the collector, various noises are superimposed on each other to generate a relatively stable trend in a short time. Compared with the stable and continuous noise of the equipment, the amplitude of the sound emitted by the target user is large and intermittent and unstable. If the intensity of the target user's voice is reflected by the energy value of the sound, the higher the energy value of the sound, the closer the distance to the sound source. At the same time, the energy source of the sound is noise and the sound of the target user, and the sound of the target user is not stable compared to the noise, so the energy shows a fluctuating trend. When the sound energy of the collector is constantly increasing, indicating that the sound of the target user is gradually increasing, the subsequent energy suddenly decays at the moment when the target user is breathing, etc., which means that the target user does not make a sound at this moment, and the sound obtained by the collector comes from the equipment noise. Therefore, the embodiment of the present invention first calculates the window energy at each moment, and then obtains the possibility that the target user does not make a sound at different times according to the trend of energy change.

[0029] Furthermore, the embodiment of the present invention obtains the sound signal of the target user of each collector at each time, and the collection period in the embodiment of the present invention is generally selected to be no less than 5s. The sound signal of the target user at each time is fitted to obtain sound curve information.

[0030] Furthermore, as an optional embodiment of the present invention, the determination module 102 is also used to determine the sound energy of the sound analysis window of the sound curve information of the collector based on the sound information in the sound analysis window of the sound curve information; and to determine the possibility that the target user has not made a sound in the current sound analysis window of the collector by utilizing the sound energy of the target user in the current sound analysis window and the sound energy of the sound analysis window adjacent to the current sound analysis window.

[0031] Specifically, the embodiment of the present invention sets the sound analysis window size to 0.05s. For the selected sound curve information, the sound energy of each collector in each sound analysis window is calculated. When determining the sound energy, the determination module 102 is also used to superimpose the sound information of the sound curve information in the sound analysis window to obtain the sound energy of the sound analysis window of the sound curve information.

[0032] Specifically, the embodiment of the present invention specifically uses the following formula to calculate the sound energy of the sound analysis window:

[0033]

[0034] In the above formula, Represents the sound energy of the ith sound analysis window of the sound curve information of the Kth collector. m represents the index of the frame, and N represents the number of sub-frames of the sound curve information in the ith sound analysis window. L represents the length of the sampling points of the frame length of 5ms. W represents the frame shift, which is half the frame length. Indicates the data size of the nth sampling point in the i-th sound analysis window. Its value is the voltage value converted by the collector, in V.

[0035] Furthermore, the higher the sound energy is, the closer the sound source is to the collector. Because the sound of the target user is superimposed on the equipment noise, the sound energy calculated above includes the sound energy and noise energy of the target user. When the sound energy of the target user is relatively small, the noise energy accounts for a higher proportion. When the target user speaks, in most cases the sound tends to increase gradually, that is, the energy of the sound gradually increases from the beginning of the sound until the sound is spoken again after ventilation, and there is no sound during the ventilation interval. Then in a sound analysis window, if the sound energy of multiple previous sound analysis windows is gradually increasing, the lower the sound energy of the current sound analysis window is than the sound energy of the earliest sound analysis window, the higher the possibility that the target user has not spoken in the current sound analysis window. For example, if Figure 2 As shown, Figure 2 A schematic diagram of energy of a target user's voice provided by an embodiment of the present invention, Figure 2In different time intervals, there are unvoiced breathing stages and voicing stages. As time goes by, the sound energy in the unvoiced breathing stage is relatively stable, and the sound energy in the voicing stage has an upward trend. That is, when the target user speaks, in most cases the sound tends to gradually increase, that is, the energy of the sound gradually increases from the beginning of the sound until the sound is spoken again after breathing, and there is no sound in the unvoiced breathing interval.

[0036] Further, as an optional embodiment of the present invention, the determination module 102 is also used to calculate the absolute value of the first difference between the sound energy of the current sound analysis window and the sound energy of each adjacent sound analysis window before the current sound analysis window, and superimpose the absolute values ​​of each first difference to obtain a first superimposed value; determine a first number of first differences between the sound energy of the current sound analysis window and the sound energy of each adjacent sound analysis window before the current sound analysis window as a positive number, and a second number of each adjacent sound analysis window before the current sound analysis window; calculate a second difference between the second number and a predetermined value; determine a first ratio between the first number and the second difference; determine a first product between the first superimposed value and the first ratio as the degree of increase of the sound energy of each adjacent sound analysis window before the current sound analysis window; normalize the first difference to obtain a first normalized value, and calculate a second product between the first normalized value and the degree of increase; normalize the second product to obtain a possibility.

[0037] Specifically, the predetermined value in the embodiment of the present invention can be 1. The embodiment of the present invention specifically uses the following formula to calculate the increment degree:

[0038]

[0039] In the above formula, represents the continuous sound of the target user collected by the Kth collector before the i-th sound analysis window The degree of increase in sound energy in each sound analysis window. Represents the sound energy of the j-th sound analysis window of the sound curve information of the K-th collector. Represents the sound energy of the j-1th sound analysis window of the sound curve information of the Kth collector. represents the second number of sound enhancement trend windows selected before the sound analysis window i. It can be determined according to actual conditions. In the embodiment of the present invention, the value is 10. The first quantity represents a first difference between the sound energy of the current sound analysis window and the sound energies of each sound analysis window adjacent to the current sound analysis window, which is a positive number.

[0040] Furthermore, the embodiment of the present invention specifically uses the following formula to calculate the probability:

[0041]

[0042] In the above formula, It represents the possibility that the target user of the i-th sound analysis window calculated by the K-th collector has not spoken. The second number of the sound enhancement trend windows selected before the i-th sound analysis window is generally selected to be 10. represents the continuous sound of the target user collected by the Kth collector before the i-th sound analysis window The degree of increase of the sound energy in each sound analysis window. The greater the degree of increase, the more obvious the trend of the target user's voice increase. Represents the sound energy of the i-th sound analysis window of the sound curve information of the K-th collector. The first one represents the sound curve information of the Kth collector. The sound energy of a sound analysis window. represents the exponential function, which is used to Perform normalization. Represents the normalization function, which is used to Perform normalization.

[0043] Furthermore, the first threshold can be determined based on actual conditions. In the embodiment of the present invention, the value is 0.7. When the possibility is greater than 0.7, it is considered that the target user has not made any sound in the sound analysis window, and most of the sound energy is noise energy. The signal in the sound analysis window represents a noise signal that the target user has not made any sound.

[0044] Furthermore, the sound analysis windows of the target user who has not made a sound are calculated and obtained through the above embodiments of the present invention. The signals of these sound analysis windows are mainly caused by noise, and may also contain other sounds of the target user, such as movement sound, friction sound, etc. This information also has an auxiliary effect on the subsequent sound recognition of the target user, and belongs to useful information, so the sound signal of this stage cannot be completely regarded as the device noise signal. If the friction sound related to the movement of the target user appears during the period, the signal at the time of movement will be enhanced, but the movement noise will not be repeated, so the signal sampling points at the time of movement have a small repetition amount and a large interval in the selected target user's non-sound period signal. Because the noise of the device has a stable feature, the data size of a sampling point will experience reappearance, so by performing an interval periodic analysis on the sampling point data of the sound signal of the target user who has not made a sound, the sampling points of the same data size are in the signal interval of the target user who has not made a sound, if the frequency interval is more stable, it indicates that the signal value generated by the device noise is more likely. Furthermore, the embodiment of the present invention evaluates the final noise signal by calculating the possibility that the data of different sampling points are noise signal values.

[0045] Furthermore, as an optional embodiment of the present invention, the determination module 102 is also used to determine the signal overlapping part of each sound analysis window to be determined, and calculate the variance of the time interval of each sampling point value in the signal overlapping part; determine the probability that the sound signal of each sampling point value in the sound analysis window to be determined when the target user does not make a sound is a noise signal based on the variance; when the probability is greater than the second threshold, determine that the sound signal of the sampling point value is a noise signal.

[0046] Specifically, the embodiment of the present invention selects the overlapping part of the signal in the undetermined sound analysis window where each collector has not made a sound as the data source for the following analysis of the embodiment of the present invention. The embodiment of the present invention classifies the sampling point values ​​in the overlapping part of the signal according to the size, that is, the data size with the same sampling point value is regarded as one category, and the time interval between the data is recorded at the same time. For example, the sampling point values ​​in the undetermined sound analysis window are 10, 13, 12, 10, 11, 10, and their corresponding positions are 12, 13, 14, 15, 16, 17. When the first 10 appears, it is located at the 12th position of the sequence, and the position where the last 10 appears is the 17th, so there are 3 data in the class with a sampling point value of 10, and the interval between the 3 data is: 2, 1. Furthermore, each signal value size in the overlapping time interval of different collectors is classified and the possibility is calculated.

[0047] Furthermore, the second threshold value may be set according to actual conditions. In the embodiment of the present invention, the second threshold value is set to 0.9. When the probability is greater than 0.9, the sampling point signal is considered to be a noise signal.

[0048] Further, as an optional embodiment of the present invention, the determination module 102 is further configured to calculate a first sum between a predetermined value and a variance; and determine the reciprocal of the first sum as the probability.

[0049] Specifically, the predetermined value may be 1. The embodiment of the present invention specifically uses the following formula to calculate the probability:

[0050]

[0051] In the above formula, Indicates that K collectors are within the target user's to-be-determined sound analysis window, and the sampling point value is The probability that the sound signal is a noise signal. Indicates that the sampling point value is The variance of the time intervals in the set of sampling points. For example, if the sampling point value is 10 and they are arranged in the order of appearance: 12, 15, 17, then the intervals are: 2, 1, and the interval variance is .

[0052] Furthermore, the fitting module 103 is also used to, in the case of noise signals, superimpose the noise signals of each collector into the same sequence in chronological order, and use the least squares method to fit the noise signals in the sequence to obtain the noise signal curve in the silent period.

[0053] Specifically, in order to avoid signal acquisition anomalies of a single collector, the probability that the sound signals of different collectors are noise signals is calculated separately, and the sampling points with probabilities greater than the second threshold are superimposed into a sequence according to chronological order. Then, the least squares fitting is used to obtain the noise signal curve of the silent period.

[0054] Furthermore, the noise signal curve of the target user in the non-voice stage is obtained by fitting the selected noise signal according to the above embodiment of the present invention, and a signal curve with the same length as the original signal curve is obtained after extension and expansion, and it is used as the noise signal of this stage, and the denoised sound curve is obtained by subtracting the original signal curve from the noise signal curve. Among them, the original signal curve is the average sound curve information obtained by averaging the sound curve information collected by each collector.

[0055] Furthermore, as an optional embodiment of the present invention, the correction module 104 is also used to calculate the average value of the sound curve information collected by each collector to obtain the average sound curve information; extend and expand the noise signal curve until it has the same length as the sound curve information to obtain a corrected noise signal curve; and subtract the average sound curve information from the corrected noise signal curve to obtain a corrected sound curve.

[0056] Specifically, the embodiment of the present invention calculates the noise signal of the target user's overlapping non-voice stage, and uses the least squares method to perform signal fitting to obtain the noise signal curve of the target user's non-voice stage. Then, the noise signal curve of the target user's non-voice stage is extended and expanded by copying and expanding until it has the same length as the selected sound curve information, thereby obtaining a modified noise signal curve. Finally, the signal value of the original signal curve is subtracted from the signal value corresponding to the modified noise signal curve to obtain a modified sound curve. The original signal curve is the average sound curve information obtained based on the average value of the sound curve information collected by multiple collectors.

[0057] Furthermore, as an optional embodiment of the present invention, the recognition module 105 is also used to input the modified sound curve into the VGGish recognition model to identify the physiological needs of the target user; send the physiological needs to the target terminal so that the caregiver can provide care for the target user based on the physiological needs, and record the care information of the target user on the target terminal.

[0058] Specifically, the embodiment of the present invention recognizes the physiological needs of the target in real time according to the modified sound curve after denoising. The recognition model is trained by transfer learning. The main model uses the pre-trained VGGish and completes the training according to the manually pre-labeled crying training set. After the training is completed, the model is deployed on the hospital server, and then the operation mode of the nursing information tracking and recording system in the hospital is as follows:

[0059] First, the target user's voice information and related physiological characteristics are collected in real time through the bed, and uploaded to the hospital server via wired or wireless means. Then the hospital server denoises the sound signal in the above way, and then identifies the target user's physiological needs based on the recognition model. Secondly, the physiological needs information of the target user is sent to the mobile terminal of the relevant nursing staff, and the nursing information of the target user is recorded at the same time; when the nursing staff receives the nursing reminder, they can realize fast nursing, improve nursing efficiency, reduce guesswork and trial and error, greatly reduce the nursing pressure at night and during busy hours, and finally realize real-time monitoring of the target user.

[0060] The embodiment of the present invention can determine the target user's unvoiced sound analysis window based on the target user's sound information, and based on the noise signal in the target user's unvoiced sound analysis window. The noise signals of each collector are fitted to obtain the noise signal curve of the unvoiced period, and the average sound curve information of the target user at each moment is corrected using the noise signal curve to obtain a corrected sound curve. Finally, the recognition model identifies the target user's physiological needs based on the corrected sound curve. In this way, the target user's sound signal is denoised by the target user's sound characteristics, which improves the noise reduction effect of the child's sound and accurately and effectively identifies the child's physiological needs.

[0061] Embodiment 2:

[0062] Corresponding to the information tracking and recording system for pediatric in-hospital nursing provided in the above embodiment, based on the same technical concept, the embodiment of the present invention also provides an information tracking and recording system for pediatric in-hospital nursing, Figure 3 A schematic diagram of a pediatric in-hospital nursing information tracking and recording system provided by another embodiment of the present invention is shown in FIG. Figure 3 The information tracking and recording system for pediatric in-hospital nursing may have relatively large differences due to different configurations or performances, and may include one or more processors 301 and memory 302, wherein the memory 302 is used to store computer programs that can be run on the processor 301, and the processor 301 is used to execute the program stored in the memory 302 to obtain the sound curve information of the target user collected by each collector in the pediatric hospital at each moment; determine the possibility that the target user has not spoken in the current sound analysis window of the collector based on the sound information of the sound curve information in the sound analysis window; and when the possibility is greater than a first threshold, determine the current sound analysis window The target user does not make a sound in the to-be-determined sound analysis window; according to the signal overlap of each to-be-determined sound analysis window, determine whether the sound signal of each sampling point value in the to-be-determined sound analysis window of the target user is a noise signal; in the case of a noise signal, use the least squares method to fit the noise signal of each collector to obtain the noise signal curve of the unsound period; based on the noise signal curve, correct the average sound curve information of the target user at each moment to obtain a corrected sound curve; input the corrected sound curve into the recognition model to identify the physiological needs of the target user, so that the nursing staff can provide care for the target user based on the physiological needs. Among them, the memory 302 can be a short-term storage or a persistent storage. The application stored in the memory 302 may include one or more modules (not shown in the figure), and each module may include a series of computer executable instructions in the information tracking and recording system for pediatric in-hospital nursing.

[0063] Furthermore, the processor 301 can be configured to communicate with the memory 302 to execute a series of computer executable instructions in the memory 302 on the pediatric in-hospital care information tracking and recording system. The pediatric in-hospital care information tracking and recording system can also include one or more power supplies 303, one or more wired or wireless network interfaces 304, one or more input and output interfaces 305, and one or more keyboards 306.

[0064] Specifically in this embodiment, the information tracking and recording system for pediatric in-hospital nursing includes a processor, a communication interface, a memory and a communication bus; wherein the processor, the communication interface and the memory communicate with each other through the bus; the memory is used to store computer programs; the processor is used to execute the programs stored in the memory to achieve the above Figure 1 The various steps in the method embodiment are similar to those in the method embodiment, and have the beneficial effects of the above method embodiments. To avoid repetition, the embodiments of the present invention will not be described in detail here.

[0065] It should be noted that the information tracking and recording system for pediatric in-hospital care provided in the embodiment of the present invention and the information tracking and recording system for pediatric in-hospital care provided in the embodiment of the present invention are based on the same application concept, so the specific implementation of this embodiment can refer to the implementation of the aforementioned information tracking and recording system for pediatric in-hospital care, and has the same or similar beneficial effects, and the repeated parts will not be repeated.

[0066] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0067] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. An information tracking and recording system for pediatric in-hospital nursing, characterized in that: The information tracking and recording system for pediatric in-hospital care includes: An acquisition module is used to acquire the sound curve information of the target user at each time collected by each collector in the pediatric hospital; A determination module, configured to determine the possibility that the target user has not spoken in the current sound analysis window of the collector according to the sound information of the sound curve information in the sound analysis window; Among them, the method for obtaining the possibility that the target user has not made a sound is as follows: calculating the absolute value of the first difference between the sound energy of the current sound analysis window and the sound energy of each sound analysis window adjacent to the current sound analysis window, and superimposing the absolute values ​​of each of the first differences to obtain a first superimposed value; determining a first number of positive numbers in which the first difference between the sound energy of the current sound analysis window and the sound energy of each sound analysis window adjacent to the current sound analysis window is a first number, and a second number of each sound analysis window adjacent to the current sound analysis window; calculating a second difference between the second number and a predetermined value; determining a first ratio between the first number and the second difference; determining a first product between the first superimposed value and the first ratio as the degree of increase of the sound energy of each sound analysis window adjacent to the current sound analysis window; normalizing the first difference to obtain a first normalized value, and calculating a second product between the first normalized value and the degree of increase; normalizing the second product to obtain the possibility; The determining module is further configured to determine, when the probability is greater than a first threshold, a to-be-determined sound analysis window in which the target user has not spoken in the current sound analysis window; The determination module is further used to determine whether the sound signal of each sampling point value of the collector in the sound analysis window to be determined in which the target user has not made a sound is a device noise signal according to the signal overlap part of each sound analysis window to be determined; A fitting module, for fitting the device noise signals of each collector to obtain a device noise signal curve during a silent period, when the device noise signal is a device noise signal; A correction module, used for correcting the average sound curve information of the target user at each time based on the device noise signal curve to obtain a corrected sound curve; The recognition module is used to input the modified sound curve into a recognition model to identify the physiological needs of the target user, so that the caregiver can provide care for the target user based on the physiological needs.

2. The information tracking and recording system for pediatric in-hospital nursing according to claim 1, characterized in that: The determining module is further used to determine the sound energy of the sound analysis window of the sound curve information of the collector according to the sound information of the sound curve information in the sound analysis window; The possibility that the target user has not spoken in the current sound analysis window of the collector is determined by using the sound energy of the target user in the current sound analysis window and the sound energy of the sound analysis window adjacent to the current sound analysis window.

3. The information tracking and recording system for pediatric in-hospital nursing according to claim 2, characterized in that: The determination module is further configured to superimpose the sound information of the sound curve information in the sound analysis window to obtain the sound energy of the sound analysis window of the sound curve information.

4. The information tracking and recording system for pediatric in-hospital nursing according to any one of claims 1 to 3, characterized in that: The determination module is further used to determine the signal overlap portion of each of the to-be-determined sound analysis windows, and calculate the variance of the time interval of each sampling point value in the signal overlap portion; Determine, based on the variance, the probability that the sound signal of each sampling point value of the collector within the to-be-determined sound analysis window in which the target user does not make a sound is a device noise signal; When the probability is greater than the second threshold, it is determined that the sound signal of the sampling point value is a device noise signal.

5. The information tracking and recording system for pediatric in-hospital nursing according to claim 4, characterized in that: The determination module is further used to calculate a first sum between a predetermined value and the variance; The reciprocal of the first sum is determined as the probability.

6. The information tracking and recording system for pediatric in-hospital nursing according to claim 1, characterized in that: The fitting module is also used to, in the case of device noise signals, superimpose the device noise signals of each collector into the same sequence in chronological order, and use the least squares method to fit the device noise signals in the sequence to obtain the device noise signal curve during the silent period.

7. The information tracking and recording system for pediatric in-hospital nursing according to claim 1, characterized in that: The correction module is further used to calculate the average value of the sound curve information collected by each collector to obtain the average sound curve information; The device noise signal curve is extended and expanded until the length is the same as the sound curve information to obtain a modified device noise signal curve; The average sound curve information is subtracted from the correction device noise signal curve to obtain the corrected sound curve.

8. The information tracking and recording system for pediatric in-hospital nursing according to claim 1, characterized in that: The recognition module is further used to input the modified sound curve into the VGGish recognition model to identify the physiological needs of the target user; The physiological needs are sent to a target terminal, so that a nursing staff can provide care for the target user based on the physiological needs, and the nursing information of the target user is recorded in the target terminal.

9. The information tracking and recording system for pediatric in-hospital nursing according to claim 1, characterized in that: The normalizing the second product to obtain the possibility includes: performing linear normalization on the second product to obtain the possibility.

Images