METHOD AND APPARATUS FOR MONITORING A SEDATED PATIENT

TECHNICAL FIELD

The invention relates in general to medical technology, and in particular to a method and an apparatus for monitoring patients during surgery and general anaesthesia.

BACKGROUND OF THE INVENTION

During surgery it is very important to observe the patient's level of consciousness and awareness. Few reliable methods of observation exist today. In the field of medical technology there is a problem in producing physical measurements representing the activity in an individual's autonomous nervous system, i.e. in the part of the nervous system, which is beyond the control of the will.

Particularly, there is a special need to monitor the autonomous nervous system of a sedated, non-verbal patient, e.g. a patient in anaesthesia or an artificially ventilated patient, in order to detect if the patient needs more hypnotics because of awakening stimuli or more analgesia because of pain stimuli.

Tests have shown that the skin's conductance changes as a time variable signal which, in addition to a basal, slowly varying value (the so-called basal level or the average conductance level through a certain interval), also has a component consisting of spontaneous waves or fluctuations.

RELATED BACKGROUND ART

WO-00/72751 relates to an apparatus and a method for detecting pain in an individual by utilising spontaneous change in skin conductance. The method includes a monitoring process which comprises the steps of measuring and storing the skin conductance signal, deriving the frequency and amplitude of fluctuations in the skin conductance signal as analysis data, comparing the analysis data with threshold data, and activating a pain indicating warning signal if the comparison fulfils a predetermined condition. The analysis data is preferably derived expressing the frequency and amplitude of the fluctuations in the measuring signal, in particular by considering the measuring signal's local maximum values and minimum values in a time window consisting of an interval containing recently elapsed points of time. In the interval analysis data are formed for the amplitude by calculating the mean value of the differences from a minimum value to the following maximum value. Analysis data for the frequency are

preferably formed by counting the number of maximum values contained in the interval.

WO-03/94726 discloses a method and an apparatus for monitoring the autonomous nervous system of a sedated patient. In the method, a skin conductance signal is measured at an area of the patient's skin. Certain characteristics, including the average value of the skin conductance signal through a time interval and the number of fluctuation peaks through the interval, is calculated. Based on these characteristics, two output signals are established, indicating pain discomfort and awakening in the patient, respectively. The awakening signal is established based on the number of fluctuations and the average value through an interval.

WO-2005/117699 relates to a method and an apparatus for monitoring a sedated patient, in particular in order to establish an output signal which indicates the state of awakening in the patient. The method comprises the steps of providing a skin conductance signal measured at an area of the patient's skin, calculating an average value of the amplitude of fluctuation peaks in the skin conductance signal through a time interval, comparing the average value with a limit value, and establishing an output signal which indicates the state of awakening in the patient if the average value exceeds the limit value.

SUMMARY OF THE INVENTION

An object of the present invention is to provide a method and an apparatus for monitoring a sedated patient, in particular a method and an apparatus that provides a reliable output signal that indicates pain in the patient.

Another object of the invention is to provide a method and an apparatus for monitoring a sedated patient which substantially differs from the prior art.

According to the invention, the above objects are achieved by a method and an apparatus as defined in the appended, independent claims.

Further advantages and characteristics of the invention are indicated in the appended, dependent claims.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 is a block diagram illustrating a preferred embodiment of an apparatus for monitoring a sedated patient, in accordance with the invention.

Figure 2 is a flow chart illustrating an overall method for monitoring a sedated patient, in accordance with the principles of the invention.

Figure 3 is a flow chart illustrating the step of recognizing an activity in a high frequency range, in accordance with a first embodiment of the invention.

Figure 4 is a flow chart illustrating the step of recognizing an activity in a high frequency range, in accordance with a second embodiment of the invention.

Figure 5 is a flow chart illustrating the step of recognizing an activity in a high frequency range, in accordance with a third embodiment of the invention.

Figure 6 is a flow chart illustrating the step of recognizing an activity in a high frequency range, in accordance with a fourth embodiment of the invention.

DETAILED DESCRIPTION OF THE INVENTION

Figure 1 is a block diagram illustrating a preferred embodiment of an apparatus for monitoring a sedated patient, in accordance with the invention.

Substantial parts of the apparatus' hardware structure is previously described in the Applicant's related patent application WO-03/94726, with particular reference to the block diagram in fig.1 and the corresponding, detailed description. The disclosure of this publication is hereby expressly incorporated by reference, the hardware structure and hardware components in particular.

On an area 2 of the skin on a body part 1 of the patient, sensor means 3 are placed for measuring the skin's conductance. The measurement arrangement is disclosed in closer detail in WO-03/94726.

The apparatus comprises a measurement converter 4; which in a preferred embodiment may include a synchronous rectifier and a low pass filter; which converts the measured skin conductance signal into a voltage. This voltage is further sent to control unit 5; which includes time discretization module 51 and analog-digital converter 52, which converts measurement data to digital form. The choice of circuits for time discretization and analog-digital conversion implies technical decisions suitable for a person skilled in the art. In the preferred embodiment, time discretization is done in an integrated circuit, which combines oversampling, filtering and discretization.

In the same way as in the related patent application WO-03/94726, the control unit 5 also includes other data storage 54, 55 and data processing units 53 interconnected to a digital bus 59.

Data processing unit 53 analyses the measured and digitized signal coming from unit 52. The signal is then analysed, using the method according to the invention, in order to establish at least a state of pain in the patient.

The control unit 5 is arranged to read time-discrete and quantized measurements for the skin conductance from the measurement converter 4, preferably by means of an executable program code, which is stored in the non-volatile memory 54 and which

is executed by the processing unit 53. It is further arranged to enable measurements to be stored in the read and write memory 55. By means of the program code, the control unit 5 is further arranged to analyze the measurements in real time, i.e. simultaneously or parallel with the performance of the measurements. The method or process performed by the control unit 5, in order to analyze the skin conductance signal, is distinctive and substantially different from the method/process disclosed in WO-03/94726.

In this context, simultaneously or parallel should be understood to mean simultaneously or parallel for practical purposes, viewed in connection with the time constants which are in the nature of the measurements. This means that input, storage and analysis can be undertaken in separate time intervals, but in this case these time intervals, and the time between them, are so short that the individual actions appear to occur concurrently.

The control unit 5 is further arranged to identify the fluctuations in the time- discrete, quantized measuring signal, by means of a program code portion which is stored in the non-volatile memory 54 and which is executed by the processing unit 53.

The processing unit 53, the memories 54, 55, the analog/digital converter 52, the communication port 56, the interface circuit 81 and the interface circuit 61 are all connected to a bus unit 59. The detailed construction of such bus architecture for the design of a microprocessor-based instrument is regarded as well-known for a person skilled in the art.

The interface circuit 61 is a digital port circuit, which derives output signals 71, 72 from the processing unit 53 via the bus unit 59 when the interface circuit 61 is addressed by the program code executed by the processing unit 53.

An active state of the first output signal 71 indicates that the analysis of the skin conductance measurement has detected that the patient is receiving awakening stimuli and may need more hypnotics. An active state of the second output signal 72 indicates the state of pain pain/discomfort in the patient.

In a preferred embodiment the display 8 comprises a screen for graphic visualization of the conductance signal, and a digital display for displaying the frequency and amplitude of the measured signal fluctuations. The display units are preferably of a type whose power consumption is low, such as an LCD screen and LCD display. The display means may be separate or integrated in one and the same unit.

The apparatus further comprises a power supply unit 9 for supplying operating power to the various parts of the apparatus. The power supply may be a battery or a mains supply of a known type.

The apparatus may advantageously be adapted to suit the requirements regarding hospital equipment, which ensures patient safety. Such safety requirements are relatively easy to fulfill if the apparatus is battery-operated. If ₅ on the other hand, the apparatus is mains operated, the power supply shall meet special requirements, or requirements are made regarding a galvanic partition between parts of the apparatus (for example, battery operated), which are safe for the patient and parts of the apparatus, which are unsafe for the patient. If the apparatus has to be connected to external equipment, which is mains operated and unsafe for the patient, the connection between the apparatus, which is safe for the patient and the unsafe external equipment requires to be galvanically separated. Galvanic separation of this kind can advantageously be achieved by means of an optical partition. Safety requirements for equipment close to the patient and solutions for fulfilling such requirements in an apparatus like that in the present invention are well-known to those skilled in the art.

Figure 2 is a flow chart illustrating an overall method or process for monitoring a sedated patient, in accordance with the principles of the invention. The process is advantageously performed by a control unit or a processing device in an apparatus for monitoring a sedated patient, such as the control unit 5 or more specifically the processing device 53 in fig. 1.

The process starts at the initial step 200.

Next, in the measurement providing step 210, a measurement signal u(t) representing the skin conductance measured at an area of the patient's skin is acquired through a measurement interval. The duration of the measurement interval may advantageously be in the range [2s, 30s], more advantageously in the interval [4s, 20s] and most advantageously approximately 10 seconds.

Next, in the recognizing step 220, a particular activity is recognized in a high frequency range of the measurement signal acquired in step 210. The actual content of the recognizing step 220 is further elaborated with reference to four embodiments illustrated in figures 3, 4, 5, and 6 below.

Next, in the test step 230, the process determines if the activity is recognized. If this test is false, the process repeats at step 210. If the test is true, the process continues at step 240.

In the output signal setting step 240 the output signal 72 is set as indicating pain in the patient.

As illustrated, and for simplicity, the process ends at the terminating step 290. However, it will be advantageous to reiterate the process from step 200, selecting another measurement interval in step 210, such as advancing the interval 1 second.

Figure 3 is a flow chart illustrating a process that forms the step of recognizing an activity in a high frequency range, in accordance with a first embodiment of the invention.

The recognizing process starts at the initial step 300.

Next, the filtering step 310 is performed. In this step a high pass filtered signal is derived from the measurement signal by means of a high pass filter. The high pass filter advantageously has a cutoff frequency in the range [0.5Hz, 2.0Hz] ₅ more advantageously in the range [0.8Hz, 1.5Hz], and most advantageously approximately 1.0Hz. Advantageously, the high pass filter is a Bessel type filter with an order of at least two, or alternatively three, four or more.

The filtering step 310 further comprises to analyse said high pass filtered signal. To this end, a frequency spectrum analysis is performed.

More specifically, the analysis comprises a spectrum calculating step 330 of calculating a frequency spectrum representing the frequency content of the high pass filtered signal. The analysis advantageously comprises a Fast Fourier Transform process.

Further, the centroid calculating step 340 is performed, wherein an amplitude value of a centroid of said frequency spectrum is calculated.

The centroid (or spectral centroid) may be considered as the two-dimensional center of mass of the area defined by the frequency spectrum curve. Consistent with the invention, an approximation to the exact centroid may be used as the calculated centroid.

Next, the comparison step 350 is performed, wherein the centroid amplitude is compared with a predetermined limit value. The predetermined value is advantageously in the range [0.003uS, 0.01 OuS], more advantageously in the range [0.004uS, 0.007uS] and most advantageously approximately 0.005uS.

If the comparison indicates that the centroid amplitude exceeds the limit value, step 360 is performed, wherein the activity is established as recognized. Else, step 370 is performed, wherein the activity is established as not recognized.

The activity recognizing process, corresponding to step 220 illustrated in fig. 2, is completed at the terminating step 390.

Figure 4 is a flow chart illustrating a process that forms the step of recognizing an activity in a high frequency range, in accordance with a second embodiment of the invention.

This recognizing step starts at the initial step 400.

Next, the filtering step 410 is performed. In this step a high pass filtered signal is derived from the measurement signal by means of a high pass filter. The high pass filter advantageously has a cutoff frequency in the range [0.5Hz, 2.0Hz], more advantageously in the range [0.8Hz ₅ 1.5Hz], and most advantageously approximately 1.0Hz.

The filtering step 410 further comprises to analyse said high pass filtered signal. To this end, a time domain analysis is performed in this embodiment.

More specifically, the analysis comprises a detecting step 430, wherein peaks and valleys in said high pass filtered signal are detected through the measurement interval.

Advantageously, the pair generating step 440 and the discarding step 450 are included after the detecting step 430. Alternatively, the comparison step 460 may be performed directly after the completion of the detecting step 430.

In step 440, pairs of corresponding peaks and valleys are generated.

In step 450 is performed, a pair of corresponding peak and valley which complies with a predetermined criterion is discarded, i.e. not considered as a valid peak and a valid valley. More particularly, this criterion is fulfilled when a difference between said corresponding peak and valley is less than a predetermined value in the range [0.01OuS, 0.04OuS], more advantageously in the range [0.015uS, 0.03OuS] and most advantageously approximately 0.02OuS.

Next, the comparison step 460 is performed, wherein the number of peaks or valleys, or advantageously the number of valid pairs of peaks and valleys, in the measurement interval, is compared with a predetermined limit value. The number is advantageously calculated as a rate, i.e. an average number per second, by dividing the total number of peaks/valleys by the duration of the measurement interval.

The predetermined value is advantageously in the range [0.04, 1.0] peaks or valleys per second, more advantageously in the range [0.06, 0.7] peaks or valleys per second, and most advantageously approximately 0.2 peaks or valleys per second.

If the comparison indicates that the number of peaks and/or valleys exceeds the limit value, step 470 is performed, wherein the activity is established as recognized. Else, step 480 is performed, wherein the activity is established as not recognized.

The activity recognizing process, corresponding to step 220 illustrated in fϊg. 2, is completed at the terminating step 490.

Figure 5 is a flow chart illustrating a process that forms the step of recognizing an activity in a high frequency range, in accordance with a third embodiment of the invention.

In this embodiment, no high pass filter is used in order to recognize an activity in a high frequency range of the measurement signal through the measurement interval. Instead, the measurement signal is fed to two low pass filters with different characteristics, the resulting filtered signals are analysed, and the activity is established as a logic operation based upon the results of the two analyses.

More specifically, the process that forms the step of recognizing an activity starts at the initial step 500. Two parallel or concurrent process branches are then performed, explained below. The first and second process branches may alternatively be performed sequentially, as the result of one does not influence on the result of the other.

The first parallel process branch continues at the filtering step 510, wherein a first low pass filtered signal is provided by low pass filtering the measurement signal using first low pass filter characteristics. The first low pass first filter characteristics comprises a cutoff frequency in the range [3Hz, 7Hz], more advantageously in the range [4Hz, 6Hz], and most advantageously approximately 5Hz. Advantageously, the second low pass filter is a Bessel type filter with an order of at least two, or alternatively three, four or more.

The filtering step 510 is followed by a step of analysing said first low pass filtered signal. To this end, a frequency spectrum analysis is performed in this embodiment.'

More specifically, the analysis comprises a spectrum calculating step 530 of calculating a frequency spectrum representing the frequency content of the high pass filtered signal. The analysis advantageously comprises a Fast Fourier Transform process.

Further, the centroid calculating step 540 is performed, wherein an amplitude value of a centroid of said frequency spectrum is calculated.

The centroid (or spectral centroid) may be considered as the two-dimensional center of mass of the area defined by the frequency spectrum curve. Consistent with the invention, an approximation to the exact centroid may be used as the calculated centroid.

Next, the comparison step 550 is performed, wherein the centroid amplitude is compared with a predetermined limit value. The predetermined value is

advantageously in the range [0.003uS, 0.01OuS], more advantageously in the range [0.004-uS, 0.007uS] and most advantageously approximately 0.005uS.

If the comparison indicates that the centroid amplitude exceeds the limit value, step 560 is performed, which establishes that either pain or awakening, or both, is detected in the patient.

Else, step 570 is performed, which establishes that neither pain nor awakening is detected in the patient.

The result of the steps 550, 560 and 570 is that a first detection signal is established, which is indicative of pain, awakening, or both in the patient. This first detection signal is passed on to the activity establishing step 580, explained below.

The second parallel process continues from the initial step 500 at the filtering step 610, wherein a second low pass filtered signal is provided by low pass filtering the measurement signal using second low pass filter characteristics. The second low pass filter characteristics comprises a cutoff frequency in the range [0,3Hz, 1,0Hz], more advantageously in the range [0,5Hz, 0,8Hz], and most advantageously approximately 0,65Hz. Advantageously, the second low pass filter is a Bessel type filter with an order of at least two, or alternatively three, four or more.

The filtering step 510 is followed by a step of analysing said first low pass filtered signal. To this end, a frequency spectrum analysis is performed.

More specifically, the analysis comprises a spectrum calculating step 630 of calculating a frequency spectrum representing the frequency content of the high pass filtered signal. The analysis advantageously comprises a Fast Fourier Transform process.

Further, the centroid calculating step 640 is performed, wherein an amplitude value of a centroid of said frequency spectrum is calculated.

The centroid (or spectral centroid) may be considered as the two-dimensional center of mass of the area defined by the frequency spectrum curve. Consistent with the invention, an approximation to the exact centroid may be used as the calculated centroid.

Next, the comparison step 650 is performed, wherein the centroid amplitude is compared with a predetermined limit value. The predetermined value is advantageously in the range [0.003uS, 0.01OuS], more advantageously in the range [0.004uS, 0.007uS] and most advantageously approximately 0.005uS.

If the comparison indicates that the centroid amplitude exceeds the limit value, step 660 is performed, which establishes that awakening is detected in the patient. Else,

step 670 is performed, which establishes that awakening is not detected in the patient.

The result of the steps 650, 660 and 670 is that a second detection signal is established, which is indicative awakening in the patient. This second detection signal is passed on to the activity establishing step 580, explained below.

In the activity establishing step 580, the activity is established as recognized if the first detection signal is indicative of pain or awakening, and if the second detection signal is not indicative of awakening.

The activity recognizing process, corresponding to step 220 illustrated in fig. 2, is completed at the terminating step 590.

Figure 6 is a flow chart illustrating a process that forms the step of recognizing an activity in a high frequency range, in accordance with a fourth embodiment of the invention.

In the same way as with the embodiment of fig. 5, no high pass filter is used in order to recognize an activity in a high frequency range of the measurement signal through the measurement interval. Instead, the measurement signal is fed to two low pass filters with different characteristics, the resulting filtered signals are analysed, and the activity is established as a logic operation based upon the results of the two analyses.

More specifically, the process that forms the step of recognizing an activity starts at the initial step 700. Two parallel or concurrent process branches are then performed, explained below. The first and second process branches may alternatively be performed sequentially, as the result of one does not influence on the result of the other.

The first parallel process branch continues at the filtering step 710, wherein a first low pass filtered signal is provided by low pass filtering the measurement signal using first low pass filter characteristics. The first low pass first filter characteristics comprises a cutoff frequency in the range [3Hz, 7Hz], more advantageously in the range [4Hz, 6Hz], and most advantageously approximately 5Hz. Advantageously, the second low pass filter is a Bessel type filter with an order of at least two, or alternatively three, four or more.

The filtering step 710 is followed by a step of analysing said first low pass filtered signal. To this end, a time domain analysis is performed.

More specifically, the analysis comprises a detecting step 730, wherein peaks and valleys in said high pass filtered signal are detected through the measurement interval.

Advantageously, the pair generating step 740 and the discarding step 750 are included after the detecting step 430. Alternatively, the comparison step 760 may be performed directly after the completion of the detecting step 730.

In step 740, pairs of corresponding peaks and valleys are generated.

In step 750, a pair of corresponding peak and valley which complies with a predetermined criterion is discarded, i.e. not considered as a valid peak and a valid valley. More particularly, this criterion is fulfilled when a difference between said corresponding peak and valley is less than a predetermined value in the range [0.01OuS ₅ 0.04OuS], more advantageously in the range [0.015uS, 0.03OuS] and most advantageously approximately 0.02OuS.

Next, the comparison step 760 is performed, wherein the number of peaks or valleys, or advantageously the number of valid pairs of peaks and valleys, .in the measurement interval, is compared with a predetermined limit value. The number is advantageously calculated as a rate, i.e. an average number per second, by dividing the total number of peaks/valleys by the duration of the measurement interval.

The predetermined value is advantageously in the range [0.04, 1.0] peaks or valleys per second, more advantageously in the range [0.06, 0.7] peaks or valleys per second, and most advantageously approximately 0.2 peaks or valleys per second.

If the comparison indicates that the number of peaks/valleys exceeds the limit value, step 770 is performed, which establishes that pain or awakening is detected in the patient. Else, step 780 is performed, which establishes that pain or awakening is not detected in the patient.

The result of the steps 760, 770 and 780 is that a first detection signal is established, which is indicative of awakening or pain in the patient. This first detection signal is passed on to the activity establishing step 790, explained below.

The second parallel process branch continues at the filtering step 810, wherein a second low pass filtered signal is provided by low pass filtering the measurement signal using first low pass filter characteristics. The first low pass first filter characteristics comprises a cutoff frequency in the range [0.3Hz, 1.0Hz], more advantageously in the range [0.5Hz, 0.8Hz], and most advantageously approximately 0.65Hz. Advantageously, the second low pass filter is a Bessel type filter with an order of at least two, or alternatively three, four or more.

The filtering step 810 is followed by a step of analysing said first low pass filtered signal. To this end, a time domain analysis is performed in this embodiment.

More specifically, the analysis comprises a detecting step 830, wherein peaks and valleys in said high pass filtered signal are detected through the measurement interval.

Advantageously, the pair generating step 840 and the discarding step 850 are included after the detecting step 830. Alternatively, the comparison step 860 may be performed directly after the completion of the detecting step 830.

In step 840, pairs of corresponding peaks and valleys are generated.

In step 850, a pair of corresponding peak and valley which complies with a predetermined criterion is discarded, i.e. not considered as a valid peak and a valid valley. More particularly, this criterion is fulfilled when a difference between said corresponding peak and valley is less than a predetermined value in the range [0.01OuS, 0.04OuS], more advantageously in the range [0.015uS, 0.03OuS] and most advantageously approximately 0.02OuS.

Next, the comparison step 860 is performed, wherein the number of peaks or valleys, or advantageously the number of valid pairs of peaks and valleys, in the measurement interval, is compared with a predetermined limit value. The number is advantageously calculated as a rate, i.e. an average number per second, by dividing the total number of peaks/valleys by the duration of the measurement interval.

The predetermined value is advantageously in the range [number per second, number per second], more advantageously in the range [number per second, number per second], and most advantageously approximately [number per second].

If the comparison indicates that the number of peaks/valleys exceeds the limit value, step 870 is performed, which establishes that awakening is detected in the patient. Else, step 880 is performed, which establishes that awakening is not detected in the patient.

The result of the steps 860, 870 and 880 is that a second detection signal is established, which is indicative of awakening in the patient. This second detection signal is passed on to the activity establishing step 790, explained below.

In the activity establishing step 790, the activity is established as recognized if the first detection signal is indicative of pain or awakening, and if the second detection signal is not indicative of awakening.

The activity recognizing process, corresponding to step 220 illustrated in fig. 2, is completed at the terminating step 890.

The above detailed description of the invention has been presented for purposes of illustration. It is not exhaustive and does not limit the invention to the precise form disclosed. Several modifications and adaptations of the present invention will be

apparent to those skilled in the art from consideration of the specification and practice of the invention.

The invention has been primarily described with reference to human patients. It should be appreciated that the invention also may be used with animals.