CN114786564B

CN114786564B - Real-time pain detection and pain management system

Info

Publication number: CN114786564B
Application number: CN202080075967.XA
Authority: CN
Inventors: 陈李魁; 沈子贵
Original assignee: Chengxi Technology Co ltd
Current assignee: Chengxi Technology Co ltd
Priority date: 2019-08-30
Filing date: 2020-08-31
Publication date: 2025-02-18
Anticipated expiration: 2040-08-31
Also published as: TWI856169B; CN114786564A; EP4021286A4; TW202133800A; WO2021037260A1; EP4021286A1

Abstract

The present invention provides a real-time pain detection system. The system includes: a device for acquiring biomedical signals related to pain of an object in need; a computing device for converting the biomedical signals acquired within a given period of time into signal data for measuring pain, analyzing the data to be divided into two or more models, the two or more models at least including a pain model defined by data showing a peak profile and a pain-free model defined by data showing a flat profile, and then measuring the pain state of the object based on the analysis result; a processing device for generating a pain index using the analysis result according to the needs or feelings of the object; and a display for displaying the pain state of the object.

Description

Real-time pain detection and pain management system

Cross Reference to Related Applications

The present application claims priority from U.S. provisional application No.62/894,532 filed 8/30 in 2019, the entire contents of which are incorporated herein by reference if fully set forth herein.

Technical Field

The present invention relates to a system and method for real-time pain detection by analysis of physiological signals, and a pain management system using the same.

Background

Pain is considered to be an unpleasant emotional and sensory experience that may be associated with real or potential tissue damage. Pain is now one of the most important clinical symptoms that can be used to detect the severity of a patient's injury. Pain is always subjective and each patient learns about the use of the word by previous experience with injury. It is therefore difficult to objectively measure the pain level by detecting a physiological condition or using a method of a parameter related to a physiological condition.

Some prior art references disclose pain monitoring/detection by measuring various physiological signals. For example, U.S. patent No.6,117,075 discloses a method and apparatus for determining depth of anesthesia (depth of anesthesia, DOA) by measuring skin temperature or photoplethysmographic pulse pressure to define and analyze oscillation modes, or correlation between simultaneous oscillation modes measured at different physical locations, to obtain an index of depth of anesthesia.

U.S. patent No.6,685,649 provides a method of monitoring a patient's condition under anesthesia or sedation. The method concentrates and relies on analyzing individual parameters related to the cardiovascular system, specifically using the time interval between the continuous waveforms, the pressure from the continuous waveforms, the time rate from the continuous waveforms.

Hitherto, it has been reported that physiological signals such as skin conductance, blood Pressure (BP), heart rate (HEART RATE, HR), electrocardiogram (Electro-Cardio-Gram, ECG), electroencephalogram (ElectroEncephaloGraph, EEG), photoplethysmogram (PhotoPlethysmoGraph, PPG), temperature have been used to determine the depth of anesthesia (DOA) or the degree of pain. However, medical studies have shown that the use of a combination of parameters from different physiological signals can significantly improve the achieved pain and painless classification performance compared to the use of any single signal alone (Guignard, clinical anaesthesiology, volume 20, phase 1: pages 161-180, 2006).

A set of physiological characteristics is reported to be used to develop a pain index or nociception index to determine the state of a patient. For example, U.S. patent No.9,498,138B2 provides an index called "PMD200" and relates to a system and method for monitoring by multidimensional analysis of a plurality of physiological signals to generate an index.

Under the care of a physician, it is considered unacceptable to subject a person who is acceptable for safety intervention to untreated severe pain. Pain management should be provided whenever there is a medical indication. Any pain management technique for the patient must be considered. Accurate prediction of pain intensity may provide valuable insight in the context of being able to be effectively used to ultimately determine the location of pain and formulate a reasonable treatment regimen based thereon. Thus, pain prediction can improve the quality of daily life of patients in health related fields such as rehabilitation, home healthcare and medical emergency services.

U.S. patent No.7,924,818 discloses an obstetric analgesic system for providing a short-acting analgesic for pain management during labor that is capable of effectively predicting uterine contractions in real time to coordinate the administration of the analgesic so that the peak effectiveness of the analgesic coincides with the intermittent pain of labor. However, in addition to pain associated with uterine contractions during labor, pain cannot be managed effectively by automated analgesic administration because of the unpredictable time and intensity of pain.

The analgesic nociception index (ANALGESIA NOCICEPTION INDEX, ANI) based on heart rate variability is reported to reflect varying degrees of acute pain. The purpose of this study was to compare the Analgesic Nociception Index (ANI) score to a numerical scoring scale (numeric RATING SCALE, NRS, 0-10) based on the self-assessment of pain in the rehabilitation room. A study of the Analgesic Nociception Index (ANI) is disclosed to evaluate new parameters for acute postoperative pain. However, it was concluded that the Analgesic Nociception Index (ANI) does not reflect the different states of acute postoperative pain measured on a numerical scoring scale (NRS) after general anesthesia of adults on the basis of sevoflurane (Ledowski et al, br. J. Anaesth., vol.111, 4: pages 627-9, 2013).

Accordingly, there remains a need to develop a system and method for detecting or monitoring or identifying the presence and severity of pain that does not rely on subjective assessment of pain whenever a patient's self-score for pain (e.g., sedated patient, very young child, person with learning disability) is not readily available, and a system and method for managing pain that is capable of determining the time of administration of an analgesic before pain begins.

Disclosure of Invention

Accordingly, the present invention provides a system and method for detecting or monitoring pain in a subject in real time after analysis and conversion using biomedical signals (e.g., heart rate).

In one aspect, the present invention provides a real-time pain detection system comprising means for acquiring biomedical signals related to pain of a subject in need thereof, computing means for converting the biomedical signals acquired over a given period of time into signal data for measuring pain, analyzing the data to separate into two or more models comprising at least a pain model defined by data showing a peak-like profile and a pain-free model defined by data showing a flat profile, further measuring a pain state of the subject based on the result of the analysis, processing means for generating a pain index from the subject's needs or sensation using the result of the analysis, and a display displaying the pain state of the subject.

In another aspect, the invention provides a method of pain management in a subject comprising obtaining a biomedical signal related to pain in the subject, converting the biomedical signal obtained over a given period of time into signal data for measuring pain, analyzing the data to separate into two or more models comprising at least a pain model defined by data showing a peak-like profile and a pain-free model defined by data showing a flat profile, and further measuring the pain state of the subject based on the result of the analysis, processing means for generating a pain index from the demand or sensation of the subject using the result of the analysis.

In another aspect, the present invention provides a pain management system comprising the real-time pain detection system of the present invention, an analgesic system for delivering an analgesic or performing a pain relief method, and means for communicating between the real-time pain detection system and the analgesic system. The system is capable of effectively predicting pain in real time to activate the analgesic system prior to pain, thereby basing the analgesic administration or pain relief method on the time and/or intensity of pain.

In a preferred embodiment of the invention, the biomedical signal is a heart rate related signal including, but not limited to, heart rate (HEART RATE, HR), pulse Rate (PR), heart rate variability (HEART RATE variability, HRV), and Electrocardiogram (ECG).

In one embodiment of the invention, an analgesic system is provided for administering a short-acting intravenous, transdermal, transmucosal, or intramuscular analgesic that provides improved pain relief timed to the pain cycle.

In certain embodiments of the invention, the analgesic may be a drug or a highly titratable, rapid onset and predictable agent with a short duration of biological activity.

In other embodiments, an analgesic system is provided for administration of an analgesic, or a Transcutaneous Electrical Nerve Stimulation (TENS) unit, or other method of impeding pain sensation, that is delivered/applied early to a patient to have an effect in the pain portion of the pain cycle.

According to the present invention, a method of monitoring pain in a subject includes acquiring biomedical signals related to pain during a process of predicting pain to establish a pain model of a patient and an optimal time for analgesic administration. In a related embodiment, the system of the present invention monitors time and intensity based on the collected biomedical signals.

In the present invention, the biomedical signal used may be any physiological signal related to heart rate. Currently available techniques that can analyze and monitor physiological signals related to heart rate include, but are not limited to, heart Rate (HR), pulse, heart Rate Variability (HRV), blood volume pulse (blood volume pulse, BVP), or Electrocardiogram (ECG). These signals reflect the level of autonomic nervous system activity, which is related to secretory activity of the heart muscle and internal organs.

The term "heart rate" or "HR" or "pulse" as used herein refers to the heart rate (bpm) measured by the contraction (beats) of the heart per minute.

The term "heart rate variability" or "HRV" as used herein refers to a physiological phenomenon of change in the time interval between heart beats, which can be measured by a change in the heart beat interval.

The blood volume pulse (blood volume pulse, BVP) signal is from a photoplethysmograph (photoplethysmographic, PPG) sensor that emits infrared light through tissue to monitor blood volume in microvasculature and arteries. Thus, the change in BVP amplitude reflects transient sympathetic activation. Most PPG sensors can be placed anywhere on the body, the finger being the most common location to record BVP signals.

An Electrocardiogram (ECG) is an electrophysiological signal associated with electrical activity of the sinus node, reflecting cardiovascular activity. Furthermore, the response of ECG to external stimuli (e.g., pain stimulus versus pressure) can create great variability in the physiological signal of a given subject. Thus, we can use the ECG signal to extract general information about pain state or intensity.

In one embodiment of the invention, when ECG data is reliably detected at the beginning, the ECG data may be used as an effective precursor to defining pain and pain free models for coordinated delivery of an analgesic agent, thereby matching the pain relieving capacity of the analgesic agent to the pain cycle.

In an embodiment of the invention, the pain management system further comprises means for delivering a short-acting analgesic to the subject prior to pain such that the analgesic capacity of the analgesic peaks with pain. For example, the pain management system includes means for providing an audible or visual warning signal for notification.

In the present invention, the subject pain management system further provides means for triggering delivery of the analgesic.

In another embodiment of the invention, a pain management system is provided having an automatic analgesic delivery feature for automatically delivering an analgesic and/or adaptively changing the concentration of the analgesic based on a monitored biomedical signal related to heart rate (e.g., via a monitored ECG). In related embodiments of the invention, the pain management system may determine the degree of pain and vary the concentration of the analgesic based on the data. Such "degree of pain" refers to the time and/or intensity of pain, which may be determined by either (1) the current ECG, (2) the time history of the ECG, or (3) by the patient input system, and/or by some combination of the above (1) - (3), depending on the pain level of the subject, as a function of the subject's needs or feel.

In an embodiment of the present invention, the ECG signal is converted from the time domain to the frequency domain, then the data is split into two waveforms, and data analysis is performed to find the characteristics and differences of the two waveforms. Finally, a sign function and a value are estimated, which may represent the extent of the injury sensation.

In the present invention, a pain management system is provided that automatically delivers an analgesic prior to pain. The system preferably receives patient input to titrate the dose of analgesic and includes a respiratory monitor, such as a pulse oximeter, to monitor the patient's oxygen saturation to ensure safety. In related embodiments, the pain management system preferably controls the delivery of the analgesic while monitoring the patient's clinical condition with a pulse oximeter. In another related embodiment, the analgesic system preferably controls a transdermal, transmucosal, or intramuscular administration system.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention, as claimed.

Drawings

The foregoing summary, as well as the following detailed description of the invention, will be better understood when read in conjunction with the appended drawings. For the purpose of illustrating the invention, there is shown in the drawings embodiments which are presently preferred.

In the drawings:

fig. 1 provides a flow chart for a pain monitoring system according to the present invention.

Fig. 2 provides a pain module in which the blue signal is the FFT _Flat (i) standard deviation profile and the red signal is the first order of the FFT _Flat (i) standard deviation profile.

Fig. 3 shows a fast fourier transform of the ECG in one embodiment of the invention and is labeled FFT _Flat (i).

Fig. 4 shows the results of the ECG spike and gentle test data, wherein the upper graph is the ECG at gentle uterine contraction test data and the lower graph is the ECG at spike uterine contraction test data.

Fig. 5 shows the peak-like and flat raw FFT accumulation results, where in case 1 the red signal is flat FFT data and the blue signal is peak-like FFT data.

Fig. 6 shows the peak-like and flat FFT raw average distribution in case 1.

Fig. 7 shows a comparison between the peak shape and the flat data in case 1.

Fig. 8 shows peak-like and flat raw FFT accumulation results, where in case 2, the red signal is flat FFT data and the blue signal is peak-like FFT data.

Fig. 9 shows the peak-like and flat FFT raw average distribution in case 2.

Fig. 10 shows a comparison between the peak shape and the flat data in case 2.

Fig. 11 shows the peak-like and flat raw FFT accumulation results, where in case 3 the red signal is flat FFT data and the blue signal is peak-like FFT data.

Fig. 12 shows the peak-like and flat FFT raw average distribution in case 3.

Fig. 13 shows a comparison between the peak shape and the flat data in case 3.

Fig. 14 shows peak-like and flat raw FFT accumulation results, where in case 4, the red signal is flat FFT data and the blue signal is peak-like FFT data.

Fig. 15 shows the peak-like and flat FFT raw average distribution in case 4.

Fig. 16 shows a comparison between the peak shape and the flat data in case 4.

Fig. 17 shows uterine contractions of the parturient.

Fig. 18, point pt_4 is time "12:04:30", point pt_13 is time, ECG data for a period of time is given, and parturient pain is predicted from 12:04:00 to 12:10:00 for a period of 10 seconds, g _ratio =1.5, alpha=1.

Fig. 19 shows the results of Anova statistical analysis of parturients numbered 0512, p=6.81633x10 ^-132, where point pt_12 is time "12:06:00", point pt_26 is time "12:08:10", and point pt_33 ends at time "12:09:20".

Figure 20 peak shape of parturient numbered 0512 was compared to the flat data.

Fig. 21, anova statistical analysis results for parturients numbered 0714, p=0.

Figure 22 peak shape of parturient numbered 0714 compared to flat data.

FIG. 23 provides a diagram of an analysis architecture of the present invention.

Detailed Description

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.

The present invention provides a novel analgesic system and method for managing pain. In accordance with the present systems and methods, biomedical data is monitored for use in coordinating delivery of pain management device methods to have a consistent effect with pain cycles.

The present invention provides a real-time pain detection system comprising means for acquiring biomedical signals related to pain of a subject in need thereof, computing means for converting the biomedical signals acquired in a given period of time into signal data for measuring pain, analyzing the data to separate into two or more models (including at least a pain model defined by data showing a peak-like profile and a pain-free model defined by data showing a gentle profile) and thereby measuring a pain state of the subject based on the result of the analysis, processing means for generating a pain index according to the demand or sensation of the subject using the result of the analysis, and a display displaying the pain state of the subject.

In another aspect, the invention provides a method for pain management in a subject, comprising obtaining a biomedical signal related to pain in the subject, converting the biomedical signal obtained over a given period of time into signal data for measuring pain, analyzing the data to separate into two or more models (comprising at least a pain model defined by data showing a peak-like profile and a pain-free model defined by data showing a flat profile), and further measuring the pain state of the subject based on the result of the analysis, processing means for generating a pain index from the subject's needs or sensation using the result of the analysis.

In operation, a monitoring device is used to collect biomedical data relating to pain, which is clinically relevant data relating to pain. A computing device is provided to analyze the collected biomedical signals and then transmit the biomedical data to create a pain model. The pain model may be defined by an algorithm for determining data such as pain onset, pain frequency, pain duration, pain intensity, time history of pain cycles, etc. Biomedical data relating to pain is then used to determine the time required to deliver the analgesic to match the analgesic's effectiveness to the pain. Based on the determined analgesic delivery time, an analgesic delivery apparatus is activated to deliver the analgesic to the patient.

In the present invention, an analysis architecture diagram is given, as shown in fig. 23:

Extraction of clinical biomedical data

Clinicians may currently use a variety of techniques to extract biomedical data related to pain, which may be used in accordance with the present invention to build pain models (including, for example, pain attacks, pain frequencies, pain durations, etc.).

In one embodiment, the detection of biomedical data related to pain used with the analgesic system of the invention may be performed using conventional methods or measurements. In the present invention, any available technique that can analyze and monitor physiological signals related to pain includes, but is not limited to, blood Volume Pulse (BVP), electrocardiogram (ECG), and skin conductance levels (skin conductance level, SCL). These signals reflect the level of autonomic nervous system activity, which is related to secretory activity of the heart muscle and internal organs.

In the present invention, biomedical data related to pain may be any clinical data, including data for determining the absence, presence and intensity of pain (Cruccu et al, 2010; et al, 2011), such as the digital pain score scale (Numeric PAIN RATING SCALES, NPRS), the language score scale (Verbal RATING SCALES, VRS), and the visual analog scale (Visual Analog Scales, VAS), (Frampton, hughes-Webb, 2011). These self-reported scales have been well applied and validated in cancer patients (Caraceni et al, 2002). In addition, the mcgill pain questionnaire (MCGILL PAIN Questionnaire, MPQ) and a brief pain scale were also used to assess the broader pain perception in the multidimensional scale (Frampton and Hughes-Webb, 2011). While self-describing pain provides an important clinical reference and has proven to be an effective method for properly treating patients suffering from pain in most cases (Brown et al, 2011). In addition, pain assessment for the recognition and prediction of human behavior may also be used, including vocalization (Puntillo et al, 2004), body movement (Young et al, 2006), and facial expressions (Lucey et al, 2011; kaltwang et al, 2012; irani et al, 2015). Despite behavioural methods, they may also not be suitable for individuals suffering from paralysis or other behavioural affecting dyskinesias. By observing an individual's face, a number of features associated with emotional states, including pain states, can be extracted. Measurements focused on various bio-physiological signals may also be used, such as heart rate variability (De Jonckheere et al, 2010, 2012; faye et al, 2010; logier et al, 2010), skin conductance or skin electrical activity (Harrison et al, 2006; treister et al, 2012), electromyography (Oliveira et al, 2012), electroencephalography (Nir et al, 2010; huang et al, 2013), and functional magnetic resonance imaging (functional magnetic resonance imaging, fMRI), (Marquand et al, 2010; brown et al, 2011). Pain assessment methods performed by multi-modal signals have been identified as very effective, some even significantly better than single signal mode (Werner et al, 2014; Et al, 2015). Pain intensity is quantitatively measured from a variety of physiological signals obtained from the wearable sensor. Automatic recognition of pain intensity from physiological signals, such as Electromyography (EMG) and body movement and support vector machines (Support Vector Machines, SVM) and random forests (Random Forests, RF) may also be included as classifiers to recognize three pain intensities (Olugbade et al 2015). Kachele et al use Electromyography (EMG), skin Conductance Level (SCL), and Electrocardiogram (ECG) in combination with unsupervised and semi-supervised learning to establish a personalized continuous pain intensity recognition system (Kachele et al, 2016).

Establishment of pain model

The system of the present invention includes a computing device for analyzing the collected biomedical signals to define pain and pain-free models (e.g., ECG data). In a preferred embodiment, the computing means for defining pain and painless models includes means for receiving and analyzing sensor inputs to accurately determine pain onset, pain frequency, pain duration, pain intensity, pain cycle history time, and the like. The system of the present invention may include a graphical user interface to display biomedical data related to pain, pain models, and to enable user interaction.

In one embodiment, the system of the present invention further comprises an intelligent system that can use biomedical data related to pain generated by the computing device to provide biomedical clinical data to determine the onset of a pain cycle. Furthermore, the intelligent system may be provided in the analgesic system of the present invention to enable real-time assistance in providing pain management support (i.e., the type of analgesic to be administered, the likelihood of delivery over a period of time, etc.).

According to the invention, the computing device is preferably a digital signal processor that can (1) automatically, accurately, and in real-time extract biomedical signals, such as ECG signals, from the sensor inputs, (2) evaluate the quality of the biomedical data provided by the processor in view of environmental noise, and (3) determine the onset of pain, pain frequency, pain duration, pain intensity, etc., based on the biomedical data.

Biomedical signals (i.e., ECG signals, etc.) collected in accordance with the present invention are transmitted from the data extraction device to the computing device for signal processing. The computing device may also be responsible for maintenance of the collected biomedical data and maintenance of the analgesic system itself. The computing device may also detect user input and operate according to the user input via user interface devices known to those skilled in the art.

In certain embodiments, the computing device includes memory capacity large enough to perform the operations of the algorithm according to the invention. The memory capacity of the present invention may support loading computer program code via a computer readable storage medium, wherein the program contains source code for performing the operational algorithm of the present invention. Alternatively, the memory capacity may support direct programming of the CPU to execute the operational algorithms of the present invention. The standard bus configuration may transfer data between the CPU, memory, ports, and any communication devices.

Communication devices such as wireless interfaces, cable modems, satellite links, microwave repeaters, and conventional telephone modems may transmit biomedical data from the computing device to the vendor over the network. Networks that may be used to transmit biomedical data include, but are not limited to, local area networks, internal networks, and the open internet.

In accordance with the present invention, a novel obstetric analgesic system is provided that includes a patient-controlled analgesic (patient controlled analgesia, PCA) function that enables a patient to automatically administer an analgesic after signaling about the onset of pain.

In a common form of Patient Controlled Analgesia (PCA) for the present invention, a mechanical device is provided for the subject that includes a reservoir and a patient operable pump. The pump dispenses an ascending dose of analgesic from the reservoir into the subject's Intravenous (IV) system, according to patient demand. The device may also include a lockout interval feature that prevents patient re-administration for a period of time to ensure overdose is avoided.

The systems for pain management according to the present invention include analgesic systems including, but not limited to, intravenous, subcutaneous, intramuscular, intra-articular, parenteral, intraperitoneal, intranasal, inhalation, oral, rectal, intravaginal, topical, nasal, ocular, topical, transdermal, sublingual, epidural, intrathecal delivery of an analgesic (e.g., analgesic, anesthetic, sedative, tranquilizer, or combination of anesthetic antagonists) or electrical stimulation of spinal nerves (e.g., transdermal Electrical Nerve Stimulation (TENS)).

According to the present invention, the analgesic may be delivered automatically based on the established shrinkage data. In certain embodiments, the analgesic agent that causes sensory loss is delivered automatically by any one of a local block, a paracervical block, a pudendum block, epidural anesthesia and analgesia, spinal anesthesia and analgesia, and inhalation anesthesia.

The invention is illustrated in the following embodiments and examples.

The symbols are defined as follows:

TPeak (i) time of peak in uterine contractions;

TFlat (j) time of gentle in uterine contractions;

10000 sampling points, and the total duration is 20 seconds;

ECG (n) ECG recordings of parturients;

f, frequency = 1000ms/512 sampling points;

Fcut shear frequency.

1. Implementation of pain model

1.1 Data extraction

The uterine contractions of each parturient were compared to the ECG patterns in its labor. Peak time in the uterine contractions plot is labeled TPeak (i). The gentle time in the uterine contractions is marked TFlat (i). When the uterine contractions of the parturient are in peak or gentle time T (i), we captured 10000 ECG signals during T (i). The ECG sampling frequency was 512Hz, totaling 10000pts of 20 seconds. We labeled ECG _Peak (i) for peak uterine contractions and ECG _Flat (i) for gentle uterine contractions.

1.2 Data processing

First, we collect 10000 ECG (i) signal points. Next, we perform a fast fourier transform (Fast Fourier Transform, FFT) and obtain the resulting FFT _Peak(i)、FFT_Flat (i). Third, the "cut frequency" F _cut is set to focus on the large difference between peak-like and flat FFT results and compare its pain statistics. Finally, we split the EEG data into two groups, "peak-like" and "flat" and calculate each mean and standard deviation.

In the four maternal examples, each "peak-like" showed a distinct difference from the average of "flat". Both "peak" and "flat" groups were tested using ANOVA and a strong significant difference was obtained.

1.3 Model implementation

Based on the assumption that the highest peak of the uterine contraction map is the time point of pain and the lowest flat point of the uterine contraction map is the time of no pain. From the results of FFT _Peak (i) defining a pain model and FFT _Flat (i) defining a pain-free model, these two FFT packet signals are clearly very different. This means that an efficient frequency domain analysis map can be obtained by FFT converting the 20 second ECG signal. We can compare the FFT (i) data with a flat general pattern and judge the occurrence of pain.

We calculate the mean and standard deviation of FFT _Peak (i) and FFT _Flat (i). From the mean distribution we use one standard deviation as the observation range for the effective pain model. To increase the variance, speed up the calculation and reduce the judgment time, we calculated the first derivative of the peak shape (defined as pain model) and the gentle (defined as painless model) standard deviation distribution. The variance of each frequency is shown and a threshold for pain occurrence time may be set in the experience data.

ECG signals are randomly acquired from a patient, their FFT (i 1) distribution is calculated, and then the difference in average between FFT (i 1) and FFT _Flat (i) is calculated. Then, the ratio of the difference to the standard deviation of FFT _Flat (i) is calculated. If the difference in low frequency is over 200% or the total frequency accumulation is over 100%, then this ECG (i) can be determined over the duration of the pain. The results of this pain module analysis are shown in figure 1.

2. Operating function

2.1 Extraction of Peak uterine contractions

A. ECG (n) is extracted at TPeak (i) and labeled n (TPeak (i))

B. Pain data showing peak-like contours were obtained:

c. The ECG _Peak (i) frequency domain data is calculated by a Fast Fourier Transform (FFT) and defined as FFT _Peak (i). From the distribution of FFT _Peak (i), we can find the pain difference between pain and painless.

D. valid data from-f_cut to f_cut is retained, a constant value of the sample point.

The absolute value of FFT _Peak (i) represents the energy of this time of occurrence

Fig. 2 shows a fast fourier transform of the ECG, labeled FFT _Peak (i).

2.2 Extraction of gentle uterine contractions

A. ECG (n) is extracted at TFlat (i) and labeled n (TFlat (i))

B. pain data were obtained over a gentle period:

c. The ECG _Flat (i) frequency domain data is calculated by a Fast Fourier Transform (FFT) and defined as FFT _Flat (i). From the FFT _Flat (i) distribution we can find the pain difference between pain and painless.

The absolute value of FFT _Flat (i) represents the energy at that time of occurrence

Fig. 3 shows a fast fourier transform of the ECG, labeled FFT _Flat (i).

Anova1 statistical test

P=anova 1 (|fft _Peak(i)|;|FFT_Flat (i) |, category)

As shown in fig. 3, the signals are divided into two groups, with p-value less than 0.05.

2.3 Establishing the final model and threshold

FFT (k) represents continuous ECG data with periodic samples s_pt. To enhance the characteristics of the FFT (k) distribution, we consider the first order derivative as FFT' (k). Comparing FFT '(k) and FFT' _Flat (k) to obtain pain index g (k), wherein α can be set by the painful individual sensation of the subject:

Judging function:

wherein X and Y are defined as "pain" and "painless", respectively, and

Wherein g _ratio and g _normal thresholds can be easily set from each fertility history.

In one embodiment of the present invention, the values of X and Y are defined as 0 and 1, respectively, representing "most painful" and "painless". In another embodiment of the invention, more than two values may be used for pain measurements. For example, the values X and Y are defined as 5 (maximum pain) and 0 (no pain), respectively, so the pain level can be expressed as 5 (maximum pain), 4 (more pain), 3 (moderate pain), 2 (mild pain), 1 (less pain), and 0 (no pain).

In one embodiment of the present invention, ECG data showing a peak profile (peak shape) and a gentle profile (gentle) are shown in fig. 4, including ECG at gentle uterine contraction test data (upper graph) and ECG at peak uterine contraction test data (lower graph).

2.4 Test results of four parturients

Case 1

The peak-like and flat raw FFT accumulation results are shown in fig. 5, including flat FFT data displayed as red and peak-like FFT data displayed as blue. The peak-like and flat FFT raw average distribution is shown in fig. 6.

An Anova statistical analysis was performed for case 1, p=1.7961 x10 ^-47. A comparison between the peak-like data and the flat data in case 1 is given in fig. 7.

Case 2

The peak-like and flat raw FFT accumulation results are shown in fig. 8, including flat FFT data displayed as red and peak-like FFT data displayed as blue. The peak-like and flat FFT raw average distribution is shown in fig. 9.

An Anova statistical analysis was performed for case 2, p=6.81633x10 ^-132. A comparison between the peak-like data and the flat data in case 2 is given in fig. 10.

Case 3

The peak-like and flat raw FFT accumulation results are shown in fig. 11, including flat FFT data displayed as red and peak-like FFT data displayed as blue. The peak-like and flat FFT raw average distribution is shown in fig. 12.

An Anova statistical analysis was performed for case 3, p=3.86697x10 ^-31. A comparison between the peak-like data and the flat data in case 3 is given in fig. 13.

Case 4

The peak-like and flat raw FFT accumulation results are shown in fig. 14, including flat FFT data displayed as red and peak-like FFT data displayed as blue. The peak-like and flat FFT raw average distribution is shown in fig. 15.

An Anova statistical analysis was performed for case 4, p=0. A comparison between the peak-like data and the flat data in case 4 is given in fig. 16.

Real-time detection of pain signals in patients by continuous ECG data

We continuously collect ECG data from the machine. Each s_pt period data may transmit ECG data from the time domain to the frequency domain. We can obtain a pain index to estimate the pain level of the patient. In practical applications, we can set s_pt/2 to be the period of the monitored rate.

For example, we set s_pt=10240 as periodic data in each calculation sample, and the sampling rate of the heart rate monitor is 512Hz. Thus, we can set the period to s_pt/2=5120, which represents the pain trend over 10 seconds of calculation. This parameter needs to meet enough heart rate ECG samples and the observation time is not too long for the same duration. By comparing the peak shape in g (k) with the flat results, we can easily set g _ratio and the alpha value, typically we can set g _ratio =1.5, alpha=1.

Although this description contains many specifics, these should not be construed as limitations on the scope of the invention or of what may be claimed, but rather as descriptions of features specific to particular implementations or examples of the invention. Certain features that are described in this specification in the context of separate implementations or examples can also be implemented in combination in a single implementation. Conversely, various features that are described in the context of a single implementation or example can also be implemented in multiple implementations or examples, respectively, or in any suitable subcombination.

Claims

1. A real-time pain detection system comprising means for acquiring pain-related biomedical signals in a subject, computing means for converting the pain-related time-domain biomedical signals acquired during each given period of time into frequency-domain signal data for measuring pain by frequency-domain conversion (frequency domain conversion) and analyzing the relation of the frequency-domain signal data to predetermined model data having a correlation of the frequency-domain biomedical signals with pain, the predetermined model data comprising a pain model defined by at least a peak-like profile in the frequency-domain data and a pain-free model defined by a flat profile in the frequency-domain data, the computing means generating a pain index using the results of the analysis and determining a pain state of the subject based on the pain index, and a display displaying the pain state of the subject.

2. The real-time pain detection system of claim 1, wherein the biomedical signal is from an Electrocardiogram (ECG).

3. A pain management system comprising the real-time pain detection system according to claim 1 or 2, and an analgesic system for automatically delivering a drug to the subject in accordance with the pain index and the pain state.

4. The pain management system of claim 3, further comprising means for communicating between the real-time pain detection system and the analgesic system.

5. The pain management system of claim 4, wherein the analgesic system is configured to administer an analgesic via intravenous, transdermal, transmucosal, or intramuscular administration.

6. A method of detecting pain, comprising:

converting the pain-associated time domain biomedical signal of a subject acquired during each given period of time into frequency domain signal data for measuring pain by frequency domain conversion (frequency domain conversion);

Analyzing the frequency domain signal data in relation to predetermined model data having a correlation of the frequency domain biomedical signal and pain, the predetermined model data comprising at least a pain model defined by a peak-like profile in the frequency domain data and a pain-free model defined by a flat profile in the frequency domain data, and

A pain index is generated using the results of the analysis and a pain state of the subject is determined based on the pain index.

7. The method of claim 6, wherein the biomedical signal is from an Electrocardiogram (ECG).

8. The method according to claim 7,

Wherein said converting comprises deriving a first-order differential fast fourier transform (FFT') of said biomedical signal to obtain said frequency domain signal data,

Wherein the predetermined model data in the analysis comprises FFT' model data, and

Wherein the pain index (g (k)) is generated using the following formula:

Where FFT' (k) is the first derivative of the continuous ECG data FFT (k), and α can be set by the pain sensation of the subject.