PHYSIOLOGICAL MEASUREMENT DEVICE, SYSTEM AND METHOD

Abstract

The present invention relates to a physiological measurement device, system and method. The device 11 receives a plurality of measurement signals s collected from a subject. i a plurality of input channels 15 configured to acquire a plurality of measurement signals s i and extracting a plurality of vector signals y from the plurality of filtered measurement signals using a digital filter. j The device 11 further comprises a recording unit 10 configured to calculate one or more stimulation signals w for electrically stimulating tissue of the subject. k a stimulation unit 20 configured to generate a plurality of vector signals y i and one or more stimulus signals w k a plurality of output channels 22, 32 configured to output one or more stimulus signals w k During or after the generation and output of the current filter coefficients, one or more stimulus signals w k and one or more stimulus signals w k A number of measurement signals s were collected while i a processing unit 27 configured to determine adjusted filter coefficients of the digital filter based on

Description

[Technical field]

[0001] In particular, the present invention relates to physiological measurement devices, systems and methods for taking and / or processing physiological measurements, such as electrophysiological measurements like an electrocardiogram (ECG) or an electroencephalogram (EEG). [Background technology]

[0002] Physiological measurement devices typically provide measurement results to a user, e.g. medical personnel, in the form of differential values ​​(often called "vectors" or, especially in the context of ECG or EEG, "leads"). The differential values ​​are derived from the differences of physical quantities measured at different locations on the patient's body. The measurement devices are designed to suppress as well as possible the common mode components of the signals measured at the different locations. Common mode interference is the main cause of artifacts in electrophysiological recordings. Such artifacts reduce the usefulness of the recordings for purposes such as diagnosis, treatment and non-medical analysis.

[0003] International Patent Application Publication No. WO2018 / 162365 describes a physiological measurement device that includes digital filters placed at the outputs of different analog-to-digital converters (ADCs) associated with different channels of the measurement device. The filters are calibrated to reduce common-mode interference. Each calibration method includes determining a reference input channel and minimizing the time-domain difference between the reference input channel and each input channel other than the reference input channel.

[0004] The article by Samiei Aria et al.: “A Bidirectional Neural Interface SOC With Adaptive IIR Stimulation Artifact Cancelers”, IEEE JOURNAL OF SOLlD-STATE CIRCUITS, IEEE,USA, vol. 56, no. 7, February 9, 2021 (2021-02-09), pp. 2142-2157, XP011863307, discloses a 180 nm CMOS bidirectional neural interface system-on-chip that allows simultaneous recording and stimulation with an on-chip stimulation artifact canceller. The front-end (FE) cancellation scheme incorporates a least mean square (LMS) engine that adjusts the coefficients of a two-tap infinite impulse response filter to replicate the stimulation artifact waveform and subtract it at the FE. Each recording channel contains a pair of adaptive infinite impulse response filters that allow cancellation of artifacts generated by the simultaneous operation of two on-chip stimulators. Summary of the Invention [Problem to be solved by the invention]

[0005] It is an object of the present invention to further reduce common mode interference and / or improve the performance of digital filters without compromising their common mode rejection capabilities. [Means for solving the problem]

[0006] In a first aspect of the present invention there is provided a physiological measurement device comprising: - a plurality of input channels configured to acquire a plurality of measurement signals collected from the subject; - a recording unit configured for filtering a plurality of measurement signals by using a digital filter and for calculating a plurality of vector signals from the filtered plurality of measurement signals, the vector signals representing differential signals being calculated by forming a linear combination of at least two filtered measurement signals; - a stimulation unit configured to generate one or more stimulation signals for electrically stimulating tissue of the subject; - a plurality of output channels configured to output a plurality of vector signals and one or more stimulus signals; and - a processing unit configured to determine, during or after the generation and output of the one or more stimulation signals, adjusted filter coefficients of the digital filter based on the current filter coefficients, the one or more stimulation signals and a plurality of measurement signals collected while the one or more stimulation signals are output for stimulation; having The digital filter is configured to, after the adjusted filter coefficients have been determined, filter the plurality of measurement signals using the adjusted filter coefficients.

[0007] In another aspect of the present invention, there is provided a physiological measurement system, the physiological measurement system comprising: - a plurality of measurement electrodes configured to collect a plurality of measurement signals from the subject; - a physiological measuring device according to any one of the claims; - a plurality of stimulation electrodes configured to apply one or more stimulation signals generated by the physiological measurement device to a subject to electrically stimulate tissue of the subject; and - an output unit configured to output a plurality of vector signals calculated by the physiological measurement device; has.

[0008] In further aspects of the present invention there is provided corresponding methods, a computer program having program code means which, when executed on a computer, causes the computer to perform the steps of the methods disclosed herein, and a non-transitory computer readable recording medium having stored thereon a computer program product which, when executed by a processor, causes the methods disclosed herein to be performed.

[0009] Preferred embodiments of the invention are defined in the dependent claims. The claimed methods, systems, computer programs and media are to be understood as having similar and / or identical preferred embodiments to the claimed systems, in particular as defined in the dependent claims and as disclosed herein.

[0010] One of the ideas of the present invention is the use of a stimulator element (also called a stimulation unit) for generating one or more stimulation signals for electrically stimulating tissues of a subject in a physiological measuring device. It has been found that the common mode interference generated by the stimulation unit offers the possibility to modify the input channel transfer function of the recorder element (also called a recording unit) to achieve a better common mode signal cancellation, both for interference generated by the device itself and for interference from external sources.

[0011] Common mode interference can be a particular problem in stimulator / recorders, since the operation of the stimulator element itself can generate large common mode signals at the input of the recorder element. This aspect becomes even more critical when the device needs to observe the instantaneous response to stimulation. This problem can be overcome according to the invention by using the interference caused by the operation of the stimulator to update the coefficients of the digital filter in such a way that the common mode rejection of the recorder element is improved. Thus, the physiological measurement device, system and method according to the invention exploit the fact that it is known when stimulation activity occurs and therefore when the input channel will pick up the (usually large) stimulation signal.

[0012] In one embodiment, the processing unit is configured to calculate the adjusted filter coefficients sample-by-sample (e.g., using least squares or recursive least squares algorithms, among others) or block-by-block (e.g., using optimization algorithms). Block-by-block algorithms are easy to use and have a well-established mathematical framework. They can easily work with various constraints and different cost functions. Their drawbacks are that they require all data being processed to be kept in memory, and the filter behavior may suddenly change as a result of updating the filter coefficients. The latter can be improved to some extent by constraints that limit the difference between the current and new filter coefficients. Online / sample-by-sample algorithms tend to occupy less memory / RAM, since less data is kept in memory. Their mathematical properties, such as convergence and stability, are more difficult to establish, especially for more complex cost functions. It is also more difficult to integrate constraints into sample-by-sample algorithms. Sample-by-sample algorithms tend to be a more elegant and economical solution, but applications are likely to use block-by-block algorithms unless the cost functions and constraints are very simple or a better mathematical theory emerges.

[0013] In another embodiment, the processing unit is configured to calculate the adjusted filter coefficients by minimizing a cost function that reflects the magnitude of the common-mode interference, where a simple measure such as the energy contained in the vector signal can be used as the magnitude.

[0014] In a practical implementation, the recording unit comprises an analog filter for analog filtering of a plurality of (preferably analog) measurement signals, an analog-to-digital converter for converting the filtered measurement signals into digital signals, in particular in single-ended mode, and a digital filter for digitally filtering the digital signals to generate a plurality of vector signals.

[0015] The processing unit may be configured to determine the adjusted filter coefficients after each stimulation event or after a predefined or arbitrary number of stimulation events. In this way, how frequently the filter coefficients should be updated can be pre-determined or individually controlled.

[0016] The physiological measurement device may further comprise a memory device configured to store a plurality of different sets of filter coefficients for an input channel, in which case the processing unit is configured to select one of the plurality of different sets of filter coefficients to determine the adjusted filter coefficients. In one configuration example, a plurality of vector signals may be defined, and the memory device may be configured to store a plurality of different sets of filter coefficients for an input channel, in which case at least some of the stored sets of filter coefficients, preferably each set, are associated with a different one of the plurality of vector signals.

[0017] In another embodiment, the processing unit is configured to calculate a plurality of vector signals, one of which is calculated based on a definition vector from a plurality of filtered measurement signals, the memory device is configured to store different sets of filter coefficients for an input channel, and the processing unit is configured to select one of the different sets of filter coefficients for calculating a particular vector signal.

[0018] In yet another embodiment, the processing unit is configured to generate samples of one or more stimulus signals applied to a number of input channels, input at least one definition vector describing a linear combination of samples of at least two input channels, optimize a measure calculated from the at least one vector signal, where the vector signal is based on samples of one or more stimulus signals and the definition vector, and obtain at least one set of filter coefficients for at least one digital filter associated with a specific input channel based on said optimization. According to this, the processing unit may preferably be configured to input a number of definition vectors and obtain at least one set of filter coefficients for a specific input vector. By optimizing the measure based on the definition vector as well as on samples of the measurement signal, the selection of a reference signal can be omitted, which results in a set of filter coefficients well suited for the individual digital input filters that allows good common-mode interference mitigation and has a particularly low negative impact on the quality of the desired (e.g. differential) component of the vector signal.

[0019] The processing unit may further be configured to optimize said measure by multiple optimization steps, in which a set of filter coefficients obtained by one optimization step is kept constant for a subsequent optimization step, by using at least one linear equality constraint on the filter coefficients and / or by using at least one non-linear inequality constraint to limit the gain of a particular digital filter at a given frequency. By basing the optimization on both samples of the measurement signal and on a definition vector, it is possible to define constraints for controlling characteristics of at least one input filter that may not be related or directly related to the reduction of common-mode interference. For example, the optimization may be subject to at least one linear equality constraint on the filter coefficients. Such a linear equality constraint may be applied to obtain a particular DC gain or a particular gain at the Nyquist frequency.

[0020] The method may include inputting a DC gain value of at least one input channel and determining a corresponding linear equality constraint based on the DC gain value. When setting the DC gain of one, more than one or all digital input filters based on the input DC gain value, differences between channels can be corrected. In one embodiment, a DC gain value of one input channel is inputted and a corresponding linear equality constraint can be determined based on the single DC gain value. The DC gain value can be determined by a separate measurement process performed before or in parallel with the method. Applying at least one linear equality constraint based on, for example, the DC gain value can prevent the optimization process from reaching a nonsensical solution of minimizing the target function by simply setting all filter coefficients to zero.

[0021] Furthermore, the optimization may be subject to at least one nonlinear inequality constraint to limit the gain of a particular input filter at a given frequency. Such constraints may provide at least some control over the frequency response of the digital input filter. These constraints may be defined, for example, to at least partially compensate for an undesirable frequency response of a portion of the analog front-end.

[0022] The optimization may also be subject to at least one bound constraint on the filter coefficients. Such a constraint limiting the range of the filter coefficients can avoid numerical overflow when storing the filter coefficients according to a particular digital format (e.g., a particular floating-point format or a particular fixed-point format). Such a constraint can further limit the gain of a particular input filter at a particular frequency, for example to avoid amplification of unimportant frequency ranges.

[0023] In other embodiments, the processing unit may be configured to sample one or more stimulation signals at an increased sampling rate compared to an operational sampling rate applied for performing physiological measurements by the physiological measurement device and / or at a resolution different from the measurement resolution of samples received for performing the physiological measurements. Such an adjusted sampling rate and / or resolution may be well suited for calibration, since higher frequencies can be taken into account more accurately, whereas a lower resolution may be acceptable during calibration since the amplitude of a test signal can be better controlled than the amplitude of a measurement signal captured during normal operation of the measurement device.

[0024] The processing unit may further be configured to optimize the measure by minimizing a penalty function of output samples corresponding to at least one vector signal calculated from the stimulus signal filtered according to the filter coefficients and from the definition vector, in particular by simultaneously minimizing a number of different penalty functions of the output samples or by minimizing a scalar target function of the different penalty functions.

[0025] A vector signal corresponding to a differential measurement is often also called a "vector". Furthermore, the applied stimulus signal may be the same for all input channels. Thus, a pure common-mode signal is applied to the channels of the measurement device. In this case, minimizing the norm of at least one vector signal corresponds to maximizing the reduction of the common mode. The output samples may correspond to a single, multiple, or all vector signals supported by the measurement device. If a dedicated set of filter coefficients is used for each vector signal or for predefined groups of multiple vector signals, the output samples may correspond to each vector signal or each group of vector signals.

[0026] The optimization may comprise simultaneously minimizing multiple penalty functions of the output samples, or minimizing a scalar target function of different penalty functions, in other words, the optimization may be a multi-objective optimization.

[0027] At least one penalty function is a convex or quasi-convex penalty function, preferably a norm. For example, the l-infinity (maximum) norm and another norm, preferably the l2 (Euclidean) norm, can be minimized simultaneously or based on a scalar target function. At least one penalty function can have a Huber loss function.

[0028] It is also possible that the optimization involves minimizing scalar target functions or other penalty functions of different norms in order to reduce the computational complexity of the optimization. Composing a target function from several different norms or penalty functions makes it possible to achieve multiple objectives or to trade off different objectives against each other. For example, using a weighted sum of the l-norm and the l2-norm minimizes not only the maximum deviation (quantified by the l-norm) but also the amount of energy in the error signal (quantified by the l2-norm), with the compromise balanced by the weights.

[0029] In other embodiments, the processing unit is configured to use a plurality of measurement signals to modify and / or monitor the effect of the stimulation, in particular to determine one or more stimulation parameters including one or more of the amplitude, timing, and shape of one or more stimulation signals, thereby enabling closed-loop neuromodulation applications.

[0030] The Samiei et al. reference cited above describes the use of an adaptive digital filter to "learn" the shape of the artifacts generated by the stimulation pulse, and in a second step, using the trained filter as a predictor to generate an estimate of the current value of the applied stimulation pulse, outputting a signal through a digital-to-analog converter using this prediction, and subtracting this composite signal from the input signal in the analog domain. This reduces the amplitude of the stimulation artifacts in the analog domain before analog-to-digital conversion takes place. As a result, the required dynamic range (input range) of the analog front end (AFE) can be reduced.

[0031] The method disclosed by Samiei et al. operates in response to a stimulation signal, i.e., the system knows when a stimulation pulse is about to be applied and can begin signal synthesis and subtraction. Therefore, the method does not provide any benefit during periods when a stimulation pulse is not applied. Also, the adaptive digital filter used in the method is only used to generate artifact estimates of the synthesis pulse. It is not applied to output values, i.e., signals that are presented to a user or used as inputs for physiological processing algorithms.

[0032] The present invention uses single-ended amplification and sampling and cannot be used for differential amplification and sampling, whereas the method disclosed by Samiei et al. uses differential amplification and sampling and single-ended amplification and sampling is shown for simplicity.

[0033] The method disclosed by Samiei et al. aims to cancel certain types of artifacts when it is known when they occur, since the artifacts are generated by the system itself in the form of a stimulus pulse. In contrast, the present invention reduces the impact of stimulus pulse artifacts insofar as the artifacts have a common-mode signal component. Thus, the adjusted filter coefficients are applied continuously, not only in response to an applied stimulus pulse; i.e., the only filter coefficient updates occur in response to a stimulus pulse. The improved symmetry of the input channels helps to cancel all forms of common-mode interference, even when the system does not know when the interference will occur.

[0034] The method disclosed by Samiei et al. works by actively injecting the estimated inverse of the stimulation artifact into the analog signal path to cancel it. However, injecting a signal into the input path can lead to problems, such as leakage currents that can flow towards the patient, which can be easily detected. The present invention aims to avoid such problems, and does not inject any compensation signal into the analog signal path, and the reduction of interfering signal components can occur by increasing the symmetry of the overall (analog + digital) input channel behavior.

[0035] The method disclosed by Samiei et al. aims to reduce the required dynamic range of the analog front end (amplifier and analog-to-digital converter). Furthermore, the method aims to reduce both the common-mode and differential-mode components of the stimulus pulse artifacts. The present invention can instead have a large dynamic range input path, which makes the above even more effective. Furthermore, the present invention only cancels common-mode interference, but is not limited to artifacts that are found to occur as a stimulus pulse. The present invention will improve common-mode rejection even during periods when no stimulus pulse is applied.

[0036] In Samiei et al., a digital filter is used as a predictor to generate an estimate of the stimulus pulse artifact. Unlike the present invention, the adaptive digital filter is not applied to the measurement output, but only to the signal output through the digital-to-analog converter and injected into the analog signal path. Furthermore, the method disclosed by Samiei et al. takes an active approach to suppress one particular form of artifact and does not improve common-mode rejection by improving the symmetry of the single-ended input channel.

[0037] Finally, according to Samiei et al., the preferred method of forming the differential signal is in the analog portion of the circuit by using a differential amplifier. The differential signal is recovered by calculating a linear combination of the single-ended signals, e.g., the difference between the single-ended signals. Since the present invention, in contrast, utilizes single-ended amplification and sampling, an adapted digital filter using adjusted filter coefficients can be applied to each single-ended signal before a linear combination (e.g., the difference) of the single-ended signals is digitally calculated.

[0038] These and other aspects of the invention will be apparent from and elucidated with reference to the embodiments described hereinafter. [Brief description of the drawings]

[0039] [Figure 1] FIG. 1 shows a schematic diagram of one embodiment of a physiological measurement system according to the present invention. [Diagram 2] FIG. 2 shows a schematic diagram of an exemplary embodiment of a physiological measurement system according to the present invention. [Diagram 3] FIG. 3 shows a schematic diagram of one embodiment of a physiological measuring device according to the present invention. [Figure 4] FIG. 4 shows a schematic diagram of another embodiment of a physiological measuring device according to the invention. [Diagram 5] FIG. 5 shows a schematic diagram of one embodiment of a recording unit according to the invention. [Figure 6] FIG. 6 shows a schematic diagram of one embodiment of the method according to the invention. [Figure 7] FIG. 7 shows a schematic diagram of another embodiment of the method according to the invention. [Figure 8] FIG. 8 shows an example of a stimulation waveform. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0040] Closed-loop neuromodulation is an emerging technology proposed for the treatment and management of neurological conditions such as Parkinson's disease, epilepsy, and spinal cord injury. The technology is based on a device (also called a "stimulator / recorder") that can deliver stimulation signals to neural tissue and simultaneously record the electrical activity of the tissue to initiate, modify, and monitor therapeutic stimulation and its effects. Another application requiring both electrophysiological recording and stimulation capabilities is a bidirectional brain-machine interface. Measured electrophysiological signals are used to transfer information from the brain to the machine, and targeted electrical stimulation is used to transfer information from the machine to the brain.

[0041] Electrophysiological recorders record small potential differences between locations on or within a patient's body caused by physiological activity of electrically excitable cells such as nerve and muscle cells. The potential differences are small (in the microvolt to millivolt range) and are usually superimposed on relatively large common-mode signal components (millivolts to volts). Common-mode signals are generated by external sources (most commonly in the form of power line interference) and need to be suppressed to avoid artifacts in physiological recordings. Differential mode measurements theoretically provide good suppression of the common-mode signal components. However, common-mode to differential mode conversion occurs in practical devices due to mismatches in the input channels. For this reason, practical devices employ additional measures to suppress common-mode interference, such as a driven reference electrode feedback loop ("right foot drive" in electrocardiograms).

[0042] The present invention provides an apparatus, system and method for electrophysiological stimulation and measurement of (electro)physiological measurement signals. The present invention uses common mode artifacts generated by the stimulation as a calibration signal to update the digital filter coefficients of the recorder components. This adjustment improves the common mode rejection of the recorder components, thereby improving the rejection of both the artifacts generated by the stimulation and the interference from external sources.

[0043] Fig. 1 shows a schematic diagram of an embodiment of a physiological measurement system 100 according to the invention. The system comprises a number of measurement electrodes 101 arranged to collect a number of measurement signals (here also called input signals, preferably analogue input signals) from a subject. The system further comprises a physiological measurement device 11, which may be implemented in software and / or hardware, for example as or in a processor or computer. Details of the physiological measurement device 11 will be described below. Furthermore, a number of stimulation electrodes 102 are provided, which are arranged to supply one or more stimulation signals generated by the physiological measurement device 11 to the subject, to electrically stimulate tissues of the subject. An output unit 103 is provided for outputting a number of vector signals calculated by the physiological measurement device 11. The output unit 103 can generally be any means for outputting, preferably in visual form, the determined number of vector signals calculated by the physiological measurement device. For example, the output unit 103 can be a display, a touch screen, a computer monitor, a screen of a smartphone or tablet, a printer etc.

[0044] The physiological measuring device 11 can be an electrophysiological measuring device such as an ECG or an EEG. Various configurations are possible, especially with regard to the number of channels N and the number of output signals M.

[0045] A relatively simple example is an ECG or EEG device with two electrodes connected to a single output signal (single vector) and two inputs (two channels). The only definition vector required for such a measurement device 11 may be v1=[1,-1].

[0046] According to another example of a physiological measurement system 110 shown in Fig. 2, a 12-lead ECG is provided with 9 measurement electrodes connected to N=9 inputs of the measurement device 11. Thus, such a measurement device has 9 channels. The electrodes may include 3 limb electrodes (right arm RA, left arm LA, left leg LL), 6 chest electrodes (V1 to V6), which are connected to 9 different inputs of the physiological measurement device 11 via measurement cables 35.

[0047] The output vector signals y1, …, y 12 may be defined according to the 12 leads that are visualized to the medical personnel by the output device 103. The leads and respective vector signals include lead l (LA-RA), lead ll (LL-RA), lead lll (LL-LA), aVR (RA-0.5*(LA+LL)), aVL (LA-0.5*(RA+LL)), aVF (LL-(RA+LA)), and (Vn-WCT); n=1,...,6. WCT corresponds to Wilson's central terminal voltage, which is approximated as (RA+LA+LL) / 3.

[0048] Two exemplary stimulation electrodes S1, S2 are provided, attached to the subject's body in this example and implanted in the subject's body in other embodiments. Separate electrodes are preferably used for recording the measurement signal and injecting the stimulation signal, which simplifies the circuitry and eliminates the presence of dual-purpose electrodes that require special consideration when updating the filter coefficients.

[0049] Stimulation electrodes are used, for example, in neuromodulation therapy (neuromodulation), which offers the possibility of treating different pathologies. The term "neuromodulation" is essentially the electrical stimulation of the nervous system to regulate or modify a particular function (movement disorder, pain, epilepsy, etc.) and can be done in different ways, such as via stimulation on the skin surface, peripheral nerve stimulation, cortical stimulation, or deep brain stimulation. A pulse generator (either as part of the physiological measurement device 11 or as an external component) is usually used to generate and modify the different stimulation settings.

[0050] The present invention is not limited to the physiological measurement device shown and described. For example, 15-lead or 18-lead ECGs may be provided in accordance with the present disclosure. High-density EEGs with 32 or more channels may also be provided in accordance with the present disclosure. Furthermore, different types, numbers and configurations of stimulation electrodes may be used depending on the desired application.

[0051] 3 shows a schematic diagram of an embodiment of a physiological measurement device 11 according to the invention. The device receives a number of measurement signals s i Preferably, these input channels 15 are directly connected to the measurement electrodes 101 (preferably by wired connection, for example via one or more measurement cables).

[0052] The recording unit 10 records a number of measurement signals s by using digital filters. i and calculates a plurality of vector signals yj (also called difference signals) from the filtered measurement signals. The recording unit 10 is preferably configured to capture electrophysiological signals from a human or animal and sample said signals in single-ended mode with respect to a common stable reference. This allows individual input channel specific digital filters to be applied to the recorded signals before the differences / vectors / leads are calculated.

[0053] The stimulation unit 20 generates one or more stimulation signals w for electrically stimulating tissue of the subject. k The stimulation unit is preferably configured to supply a stimulation signal to electrically excitable tissue (nerve or muscle tissue) of a human or animal.

[0054] A plurality of output channels 22, 32 outputs a plurality of vector signals y i (to an output 103, e.g. a display) and one or more stimulus signals w k (preferably to a stimulating electrode 102 which is wired via one or more signal cables or the like).

[0055] The processing unit 27 receives one or more stimulus signals w j The adjusted filter coefficients x' of the digital filter during or after the generation and output of i the current filter coefficient x i , 1 or more stimulus signals w k and one or more stimulus signals w k A number of measurement signals s were collected while i The processing unit 27 may be implemented in hardware and / or software by a computer or processor, for example as one or more programmed microprocessors. The processing unit 27 is further preferably configured to control the recording unit 10 and the stimulation unit 20 and their functions.

[0056] The processing unit can thus update the coefficients of the individual channel-specific digital filters while or after the stimulation unit generates the stimulation signal, thereby minimizing the appearance of common-mode stimulation artifacts in the difference (vector) signal. Different algorithms can be used to update the filter coefficients, such as online sample-by-sample optimization (Least Mean Squares (LMS) or Recursive Least Squares (RLS)), or block-by-block optimization.

[0057] The present invention is based on the insight that devices with integrated stimulation functionality may generate significant common mode signals when generating the stimulation signal. Devices intended to provide the stimulation signal are not subject to very restrictive limitations, which means that the stimulation signal can be large enough to be used to compensate for input channel calibration / mismatch. Thus, the present invention exploits signals generated for the primary purpose of stimulating tissue for the secondary purpose of updating channel-specific digital filter coefficients.

[0058] Fig. 4 shows a schematic diagram of another more detailed embodiment of a physiological measuring device 11 according to the invention. The recording unit 10 of the physiological measuring device 11 has a number of channels. For simplicity, only three channels are shown in Fig. 4, but in general there are N channels. Each channel receives an analog input signal (measurement signal) s i The multi-channel analog-to-digital converter 21 has analog inputs 15 for inputting individual input signals s (i=1, ..., N). The inputs 15 of each channel are connected to an analog front end 17. The analog front end 17 includes analog input circuits 19 associated with each channel. The analog input circuits 19 may include one or more amplifiers, impedance converters, analog filters, etc. The input circuits of each channel are connected to a multi-channel analog-to-digital converter 21. The multi-channel analog-to-digital converter 21 receives the individual input signals s preprocessed by the individual analog input circuits 19. i , the input signal s of each channel i Sample c i = 1,…,N.

[0059] In the illustrated embodiment, the multi-channel analog-to-digital converter 21 has multiple analog-to-digital converters (ADCs 23), each of which is associated with a particular channel. In other embodiments, a single ADC 23 can be associated with multiple channels, and the multi-channel analog-to-digital converter 21 can have a multiplexer for selectively connecting one channel to each ADC. When using a multiplexer, one ADC 33 can be used to convert the input signal of each channel in turn. A multiplexer can be used to sequentially sample several input channels with one ADC. For example, in a possible implementation of an ECG device, there can be two ADCs operating to sequentially sample five ECG input channels per ADC for a total of ten input channels.

[0060] The digital interface of the multi-channel analog-to-digital converter 21 is connected to a processing unit 25 of the measuring device 11 and receives the individual input signals s i Sample of c i is transferred to the processing unit 25.

[0061] The processing unit 25 may comprise a first processor 27 and a first memory device 29, which together constitute a computing device operable to perform the various processing steps required to perform measurements by the physiological measurement device 11. The processing unit 25 comprises a number of digital input filters 31. The number of these input filters may correspond to the number of channels N, with one digital input filter 31 assigned to one particular channel. Each input filter 31 is represented by a column vector x i It can be implemented as a finite impulse response (FIR) filter with filter coefficients denoted by (i=1,…,N). i The length K of each digital filter 31 corresponds to the number of filter coefficients of each digital filter 31. The digital filters 31 are also called input channel specific digital filters (ICSFs). These filters i Sample of ci before a differential output signal is derived from these samples.

[0062] The present disclosure is not limited to a particular filter topology. Instead of an FIR filter, other filters may be applied, such as an infinite impulse response (IIR) filter. The filter 31 may be implemented in software, i.e., a computer program stored in the memory device 29 is executed by the first processor 27 to calculate the filter coefficients x i It can also be programmed to perform a filter algorithm based on the filter coefficients x i may be stored in a programmable, non-volatile portion of the first memory device 29. The processing unit 25 is accessible by a processor (processing unit) 27 (which may be implemented as a separate or common processing unit with the processing unit 25), which processes the sample c i Read the filter coefficients x i This allows the user to program and adjust the

[0063] The processing unit 25 receives the individual input signal samples c filtered by the individual digital input filters 31. i 1 or more output signals from y j (j=1, ..., M) to calculate the output signal y j is the filtered sample of multiple channels c ~ i can be calculated by forming a linear combination of at least two input signals s i This linear combination is used to measure the specific output signal y j Definition of vector v j can be specified by, where 1 T *v j = 0, and 1 is a column vector of 1s, i.e. 1 T = [1,…,1], and “*” represents a scalar product. The output signal is y j =c ~ i *vj The linear combination described by the definition vector specifies a differential measurement. The differential measurement can be based on two or more channels. In some applications, such as ECG or EEG, the vector signals are often also called "leads".

[0064] In physiology, voltage measurements are almost always differential. Other physical quantities such as pressure or temperature are usually measured as absolute values ​​(e.g., a temperature of 37°C or an intra-arterial pressure of 100 mmHg), but in some cases differential measurements may be important (cardiac output can be measured by thermodilution, which requires measuring the temperature difference at two points in a blood vessel). Thus, the techniques described here can also be applied to differential measurements that are not voltage measurements.

[0065] For simplicity, FIG. 4 shows only one digital filter 31 per input channel. However, for one input channel i, each sample c i It is also possible to provide multiple digital filters 31 for processing the signal c. Different digital filters 31 may be associated with different output signals. Thus, the filtered signal c i ~j A particular variant of j can be calculated.

[0066] FIG. 5 shows an exemplary embodiment of the recording unit 10. In this example, the output signal y1 is generated by subtracting the filtered input signal s2 from the other filtered input signal s1 (definition vector [1,-1]). Thus, the output signal y1 corresponds to a signal obtained by differential measurement based on the two input signals s1 and s2. When performing physiological measurements, the input signals s1, s2 are subject to common-mode interference. In an ideal situation, subtracting the two signals s1, s2 from each other would completely eliminate the common-mode part. However, in a practical implementation of the measuring device 11, the signal y1 contains a common-mode part to some extent due to differences in the electrical characteristics of the components 19, 21, 23, etc. of the signal paths associated with the individual channels.

[0067] In electrophysiology, the voltages generated between points on the surface of a patient due to muscle or nerve cell activity are measured. The signals of interest are usually in the microvolt (EEG) to millivolt (ECG) range, but the common mode component of the signal can be tens or hundreds of millivolts. In a medical environment, there are many possible sources of interference (including common mode interference sources) ranging from other medical equipment, electronic communication or networking equipment, and power line noise. For the output signal yj to be useful for diagnosis and treatment, it is necessary for the electrophysiological measurement device 11 to have a high common mode rejection ratio (CMRR). Although standardized test procedures exist for determining the CMRR in the range of 50 Hz or 60 Hz power line frequencies, common mode interference should also be mitigated at other frequencies. Frequencies where common mode interference should be mitigated may include frequencies outside the ECG / EEG band, especially if the recording device performs some kind of out-of-band processing (e.g., for electrode impedance measurements or ECG pacemaker pulse detection).

[0068] As mentioned above, the recording unit 10 records the potentials from several electrodes. These electrodes are usually in contact with the patient's tissue by an implanted electrode array (external electrodes, e.g. glue or needles, can also be used). Each electrode is connected to an analog input circuit within the device. The analog input circuit usually includes a low pass filter, but can also include other filter characteristics such as a high pass filter or a band stop filter. Due to the use of real components, the analog input circuit is subject to component tolerances that lead to impedance mismatches. This is one cause of common-mode to differential mode conversion. Another cause of common-mode to differential mode conversion is the impedance of the tissue / electrode boundary. This impedance can also change slowly over time as the electrode / tissue boundary degrades.

[0069] After passing through the analog input circuitry, each electrode signal (also called measurement signal or input signal) is converted to digital form using an analog-to-digital converter. This conversion is done in single-ended mode, which means that the electrode signal is referenced to an internal stabilized potential. After analog-to-digital conversion, each electrode signal passes through a digital filter with adjustable coefficients. After this filter, vectors ("difference" / "lead") are formed from the individual electrode signals, and these vectors contain information used to evaluate the progress of the treatment.

[0070] The stimulation unit 20 outputs a stimulation signal w via an output channel 22. k (k=1,...,K) to the stimulation electrodes, which activates electrically excitable tissue, typically neural tissue. The stimulation waveforms typically have a relatively large amplitude to ensure tissue excitation. The stimulation signals also induce a large common-mode signal at the inputs of the recording unit 10, particularly at any input electrodes not used as stimulation pulse outputs.

[0071] To reduce common-mode interference, a digital filter 31 is provided for each output signal y jTo this end, a processing unit 27 shown in FIG. 4 is applied to calculate the filter coefficients x i Determine / adjust the individual output signals y j It is possible to minimize the common mode component of the

[0072] As the physiological measurement device 11, and in particular the stimulation unit 20, controls the timing and waveform of the stimulation, the single-ended electrode data as the stimulation occurs can be recorded as a measurement signal. In this case, the updated filter coefficients x' i The processing unit calculates the single-ended data samples, prior knowledge of expected signal and noise waveforms, and the current filter coefficients x i This can be done sample-by-sample using algorithms such as Least Mean Squares (LMS) or Recursive Least Squares (RLS), or block-by-block using an optimization algorithm.

[0073] The prior knowledge can be known characteristics of the physiological waveform, such as frequency domain characteristics or statistical characteristics. Frequency domain characteristics would be the knowledge that the energy of the signal of interest is mostly contained within a known frequency range. Statistical characteristics vary between different measurement types. For example, ECG signals are sparse, meaning that many samples of the ECG waveform are equal to or close to zero. The derivative of the ECG signal with respect to time is also sparse. That is, the slope and curvature of the ECG signal are often close to zero. The sparseness of the derivative means that the signal has structure in time. In statistical terms, sparseness means that the probability density function is leptokurtic / super-Gaussian. The probability density function of the ECG signal is also usually asymmetric, resulting in a skewness value that is not close to zero. Furthermore, some time domain characteristics of the ECG signal are known, for example, the absolute value of the slope is usually less than 350 mV / s.

[0074] Other measurements may have different characteristics. For example, EMG signals are also usually sparse (discernible parts of activity and inactivity) and therefore leptonic. EEG signals are usually more random; that is, they are dense rather than sparse. Measurements that can pick up a single action potential are sparse because excitable cells have a refractory period with no activity between each action potential.

[0075] Such characteristics can be integrated into a cost function or constraint to find filter coefficients that result in an output vector signal that matches the known characteristics. Similarly, known characteristics of an interference signal (i.e., a stimulus signal, etc.) can be used to find filter coefficients that minimize the similarity of the output vector signal to the interference. For example, line noise interference is typically dense (values ​​away from zero most of the time) and plagioclase / sub-Gaussian in shape.

[0076] 4, a processing unit 27 and a memory device 29 are provided. The processing unit 27 may include input / output (IO) circuitry configured to access the memory device 29, in particular the non-volatile portion thereof, and to access the current filter coefficients x i and the filter coefficient x i The IO circuitry is configured to access data present in the measurement device 11 and / or perform IO operations to control the measurement device to perform the methods described herein.

[0077] The cost function to be minimized can be chosen to reflect the magnitude of the common-mode interference. In one embodiment, this may mean a simple measure such as the energy contained in the vector signal. The optimization may also be subject to constraints, for example to ensure that the differential mode gain (responsible for accurate reproduction of the signal of interest) is not reduced, or to limit the values ​​of the calculated filter coefficients to avoid overflow. Typically, the calculation of the filter coefficients is performed while the device is in use on a patient.

[0078] In one embodiment, as shown in Figures 4 and 5, each channel has exactly one digital filter 31. The filter samples C ~ i is the value of each output signal y corresponding to a particular vector or lead. j In a different embodiment, at least one channel has at least two sets of filter coefficients corresponding to different digital filters. The filters are used to calculate one or more particular output values. For example, a first digital filter is used to calculate a first output signal and a second digital filter is used to calculate a second output signal. Both digital filters are configured to filter the input signal of one channel. In one embodiment, each output signal y j has a dedicated set of digital filters 31. In other words, the set of filter coefficients corresponding to each digital filter 31 is j is obtained for a particular vector corresponding to

[0079] Fig. 6 shows a flow chart of an embodiment of a physiological measurement method 200 according to the invention, which is carried out, for example, by the physiological measurement device 11. In a first step 201, a number of measurement signals collected from a subject are acquired (received or retrieved). In a second step 202, the number of measurement signals is filtered using a digital filter and a number of vector signals are calculated from the filtered number of measurement signals. In a third step 203, one or more stimulation signals are generated for electrically stimulating tissues of the subject. In a fourth step 204, the number of vector signals and the one or more stimulation signals are output. In a fifth step 205, during or after the generation and output of the one or more stimulation signals, adjusted filter coefficients of the digital filter are determined based on the current filter coefficients, the one or more stimulation signals and the number of measurement signals collected while the one or more stimulation signals are output for stimulation. The method then proceeds to step 201 and, in step 202, the adjusted filter coefficients are used for filtering.

[0080] It should be noted that the adjustment of the filter coefficients is performed only when the stimulus signal is generated and output. Furthermore, the adjustment of the filter coefficients may not be performed at all times when the stimulus signal is generated and output, but may be performed only at regular or irregular intervals, randomly, or in response to a specific trigger or command.

[0081] Fig. 7 shows a flow chart of another embodiment of a physiological measurement method 300 according to the invention. In a first step 301, the device is checked if it is ready or prepared to emit a stimulation signal. If it is, in a second step 302, a stimulation signal is generated. In parallel with this, in a third step 303, single-ended electrode signals are recorded during (and optionally after) stimulation, in a fourth step 304, updated channel-specific digital filter coefficients are calculated, and in a fifth step 305, the updated channel-specific digital filter coefficients are applied. Finally, in a sixth step 306, the method is terminated and waits for the next stimulation event. During this waiting time, electrode signals from the measurement electrodes can be recorded and processed with the updated filter coefficients.

[0082] FIG. 8 shows various examples of stimulation signals. The stimulation signal can generally be a rectangular biphasic or monophasic pulse, which is often used for electrical stimulation. Other waveforms such as decaying exponential, trapezoidal, or sinusoidal can be used as well. The invention does not imply or presuppose any particular stimulation waveform. The stimulator design can use almost any waveform that achieves the desired stimulation without any compromise or compromise in terms of calibration. Usually, this means a rectangular pulse followed by a longer, shorter section with reversed polarity (to make the average current over time zero). It can also mean a rectangular pulse followed by a similar rectangular pulse with reversed polarity ("biphasic stimulation"). The two pulses can also be separated by a short equipotential section. More complex stimulation patterns can be formed by sequences of such pulses.

[0083] Typically, all input channels are fed with one stimulation signal. Only a single stimulation signal is applied to the patient through the stimulation output electrodes. Although the appearance of the signal at each of the input electrodes is slightly different due to differences in the input transfer functions, each input channel signal is still a rendering of the same stimulation signal. Therefore, there are at least two single-ended input channel signals.

[0084] The stimulation signal is typically much larger in amplitude than the bioelectrical signal picked up by the recording unit. Depending on the application, the stimulation signal is delivered via electrodes that are used just for this purpose, or via electrodes that are also used as recorder inputs outside of the periods when stimulation occurs.

[0085] The present invention takes advantage of the fact that stimulation signals often appear as large common-mode components at the input electrodes. However, because the analog parts of the input channels (G1, G2 in FIG. 5) are slightly different, the original common-mode components are no longer equal at the ADC. This is also called common-mode to differential-mode conversion. When the difference / vector is formed using the raw ADC signals, some of the common-mode components appear in the differential signal.

[0086] The device according to the invention has access to the raw ADC samples as well as a calculated difference signal y, which is the difference of the ADC signal after it has been processed by an equalization filter. For samples affected by a stimulus signal, the device can search for filter coefficients that minimize some measure of the magnitude of y (e.g., the l2 norm, an upper-bound norm) subject to a set of constraints. The result is the updated filter coefficients.

[0087] According to the present invention, the adjusted filter coefficients of the digital filter are determined during or after the generation and output of one or more stimulation signals based on the current filter coefficients, the one or more stimulation signals, and a number of measurement signals collected while the one or more stimulation signals are output for stimulation. In one embodiment, this information is used to determine a cost function and constraints for an optimization problem. In this case, the optimization problem can be solved with respect to the updated coefficients. For example, deviations of the updated coefficients from the current coefficients can be penalized in the cost function or limited by constraints. This prevents the optimizer from making large jumps in the coefficient values.

[0088] In another example, the cost function penalizes some measure of the amplitude of the measurement signal, for example the l2 / Euclidean norm. Because the measurement signal during the stimulus pulse contains a significant amount of common-mode interference, minimizing that amplitude over the filter coefficients will result in updated coefficients that provide better common-mode rejection even when no stimulus signal is applied.

[0089] In other examples, deviations of the updated coefficients from their initial or nominal values ​​may be penalized or limited, eliminating solutions that deviate too far from what is known to be acceptable.

[0090] In another example, one could penalize the maximum value of the covariance of the stimulation signal and the measurement signal, which would be an advanced implementation compared to just penalizing the amplitude of the measurement signal during stimulation.

[0091] In another example, the onset and duration of the stimulus event is known and used to trigger the recording of measurement data that may contain a large common mode component from the stimulus pulse.

[0092] In the following, an exemplary implementation of the optimization problem, cost function and constraints is described. The following notation will be used: a, b, c: scalars b, x, y: column vectors A, B, X: Matrix 0,1: vectors of all zeros and all ones, respectively b T , x T :Transposed vector ||…||2: l2 norm, Euclidean norm ||…||1: l1 norm ||…|| ∞ : Upper bound norm, Chebyshev norm, uniform norm A typical FIR filter is: b=[b0b1…b m ] T (filter coefficient vector, length m+1); x(n) = [x(n)x(n-1)…x(nm)] T (input data vector); y(n)=b0x(n)+b1x(n-1)+…+b m x(nm)=b T x(filter output); It is defined as follows:

[0093] The difference between the two filtered signals is:

number

number

[0094] The optimization problem for updating the filter coefficients can be implemented as follows: Record x1 and x2 during the stimulus event. From this data, form a matrix X by placing the contents of x1 and x2 in a matrix. Each x represents a common mode component (e.g., x 1,cm ) and the differential mode component (x 1,dm ), so the matrix X can also be considered as X=X dm +X cm where X cm is a matrix norm of X during a stimulus event. dm significantly larger than

[0095] A typical optimization goal may be to minimize ||X·b|| with respect to b, since during a stimulus event the differential mode signal is small while the common mode signal is large and the change in norm with respect to b (the slope) is dominated by the common mode signal component. Constraints and regularization are used to avoid non-useful solutions (e.g. b=0) while also imposing desired properties on the filter coefficients, such as minimal attenuation of the signal of interest.

[0096] An example cost function is:

number

[0097] Constraints can be used to impose desired properties on a solution. For example, T b k An equality constraint of the form -c=0 can be used to limit the gain of the equalization filter bk to c at f=0 Hz. Since the signal of interest is usually located near 0 Hz (for the entire frequency range), one such constraint for each equalization filter ensures that the equalization filter does not significantly attenuate the signal of interest.

[0098] If you do not know exactly what gain you need at f=0Hz, 1 T b k -c≦0 -1 T b k +d≦0 We can constrain the gain at f=0 Hz to lie in the closed interval [d;c] using a pair of inequality constraints: This gives the optimization algorithm some freedom as far as the gain at 0 Hz is concerned.

[0099] The gain at frequencies other than 0Hz is ||Wf·bk||-c≦0 can be restricted by an inequality constraint of the form

number

[0100] Boundary constraints on the filter coefficients of the form bi<=c and bi>=d (usually d=-c, c positive) can be used to restrict the values ​​of the coefficients to the range of the data type used to store them, or to smaller absolute values ​​to avoid numerical overflow during the calculation of the filter output values. Many commercial and freely available optimization packages accept boundary constraints as inputs, resulting in more efficient computations compared to listing the boundary constraints among the general constraint inputs. More elaborate boundary constraint settings can be used to restrict the deviation of the optimization results from some nominal or initial filter coefficient vector.

[0101] The optimization problem should be formulated as a convex optimization problem if possible, which means that the cost function to be minimized should be convex, the inequality constraints should be convex, and the equality constraints need to be affine ("linear" is not entirely correct, but is often colloquially used synonymously with "affine").

[0102] In summary, the present invention improves the input channel symmetry of an electrophysiological recorder that is part of, combined with, or used with an electrophysiological stimulator. The stimulation operation of such a device generates significant undesirable common mode signal components. Since the timing and waveform of the stimulation are known, digital filter coefficients for the individual input channels can be calculated that correct for the input channel asymmetry present. This makes the device, and in particular the recorder, less sensitive to common mode interference from any source, such as interference caused by the functioning of the stimulator, but also from unknown external sources. The reduced sensitivity to interference improves the therapeutic effect of therapeutic devices, such as closed-loop neuromodulation devices, and the reliability of non-therapeutic devices, such as brain-machine interfaces.

[0103] The invention is particularly applicable to electrophysiological stimulators which also include an electrophysiological recording function and in which the stimulation effect produces a significant common mode signal at the input of the recorder element.

[0104] One group of devices are closed-loop neuromodulation devices. These devices deliver therapeutic stimulation only when and to the extent necessary, and require monitoring of neural signals to determine the need, extent and success of treatment. Improved artifact and interference rejection improves monitoring of treatment efficacy and disease progression, ultimately improving the likelihood of treatment efficacy. In this context, measurement signals collected from the subject may be used to modify and / or monitor the efficacy of the stimulation. For example, stimulation parameters, amplitude, timing and / or shape of the stimulation signal may be adjusted or controlled based on information obtained from the measurement signals that may be indicative of the efficacy of the stimulation.

[0105] The second group of devices are bidirectional brain-machine interfaces: improved artifact and interference rejection increases the accuracy and reliability of the information collected from the brain and transmitted to the machine.

[0106] While the invention has been illustrated and described in detail in the drawings and foregoing description, such illustration and description are to be considered illustrative or exemplary and not restrictive. That is, the invention is not limited to the disclosed embodiments. Other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims.

[0107] In the claims, the word "comprise" does not exclude other elements or steps, and the singular does not exclude a plurality. A single element or other unit may fulfill the functions of several items recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.

[0108] The computer program may be stored / distributed on a suitable non-transitory medium, such as an optical storage medium or a solid-state medium, provided together with or as part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless communication systems.

[0109] Any reference signs in the claims should not be construed as limiting the scope.

Claims

1. a plurality of input channels for acquiring a plurality of measurement signals collected from a subject; a recording unit for filtering the plurality of measurement signals using a digital filter and calculating a plurality of vector signals from the filtered plurality of measurement signals, the vector signals representing differential signals being calculated by forming a linear combination of at least two filtered measurement signals; a stimulation unit for generating one or more stimulation signals for electrically stimulating tissue of the subject; a plurality of output channels for outputting the plurality of vector signals and the one or more stimulus signals; a processing unit that determines adjusted filter coefficients of the digital filter during or after generating and outputting the one or more stimulus signals based on current filter coefficients, the one or more stimulus signals, and the plurality of measurement signals collected while the one or more stimulus signals are being output for stimulation; A physiological measurement device having: the digital filter filters the plurality of measurement signals using the adjusted filter coefficients after the adjusted filter coefficients are determined; Physiological measuring device.

2. the processing unit calculates the adjusted filter coefficients sample by sample, in particular using a least squares or recursive least squares algorithm, or block by block, in particular using an optimization algorithm; Physiological measurement device according to claim 1 .

3. the processing unit calculates the adjusted filter coefficients by minimizing a cost function that reflects the magnitude of common-mode interference. Physiological measurement device according to claim 1 .

4. the recording unit comprises an analog filter for analog filtering of the plurality of measurement signals, an analog-to-digital converter for converting the filtered measurement signals into digital signals, in particular in single-ended mode, and a digital filter for digitally filtering the digital signals to generate the plurality of vector signals; Physiological measurement device according to claim 1 .

5. the processing unit determining adjusted filter coefficients after each stimulus event or after a predetermined or any number of stimulus events; Physiological measurement device according to claim 1 .

6. a memory device for storing a plurality of different sets of filter coefficients for one input channel; the processing unit selecting one set of the plurality of different sets of filter coefficients to determine the adjusted filter coefficients. Physiological measurement device according to claim 1 .

7. the processing unit calculates a plurality of vector signals, wherein one of the plurality of vector signals is calculated from a plurality of filtered measurement signals based on a definition vector; the memory device stores different sets of filter coefficients for one input channel; the processing unit selecting one set of the different sets of filter coefficients to calculate a particular vector signal; Physiological measurement device according to claim 1 .

8. The processing unit generating samples of one or more stimulus signals applied to the plurality of input channels; inputting at least one definition vector describing a linear combination of samples of at least two input channels; optimizing a measure calculated from at least one vector signal, wherein the vector signal is based on samples of the one or more stimulus signals and the definition vector; obtaining at least one set of filter coefficients for at least one digital filter associated with a particular input channel based on the optimization; Physiological measurement device according to claim 1 .

9. the processing unit inputs a plurality of definition vectors and obtains at least one set of filter coefficients for a particular input vector; 9. The physiological measurement device of claim 8.

10. The processing unit By multiple optimization steps where the set of filter coefficients obtained by one optimization step is kept constant for subsequent optimization steps, and / or By using at least one linear equality constraint on the filter coefficients, and / or By using at least one nonlinear inequality constraint to limit the gain of a particular digital filter at a given frequency, optimizing said measure; 9. The physiological measurement device of claim 8.

11. The processing unit at an increased sampling rate compared to the operating sampling rate applied to perform physiological measurements by the physiological measurement device; and / or at a resolution different from a measurement resolution of the samples received to perform the physiological measurements; sampling the one or more stimulus signals; 9. The physiological measurement device of claim 8.

12. the processing unit optimizes the measure by minimizing a penalty function of output samples corresponding to at least one vector signal calculated from the stimulus signal filtered according to the filter coefficients and from the definition vector, in particular by simultaneously minimizing a plurality of different penalty functions of the output samples or by minimizing a scalar target function of different penalty functions, 9. The physiological measurement device of claim 8.

13. a plurality of measurement electrodes for collecting a plurality of measurement signals from the subject; A physiological measurement device according to claim 1; a plurality of stimulation electrodes configured to apply one or more stimulation signals generated by the physiological measurement device to the subject to electrically stimulate tissue of the subject; an output unit that outputs a plurality of vector signals calculated by the physiological measurement device; A physiological measurement system comprising:

14. In the physiological measurement device according to claim 1, obtaining a plurality of measurement signals collected from a subject; filtering the plurality of measurement signals using a digital filter and calculating a plurality of vector signals from the filtered plurality of measurement signals, the vector signals representing differential signals and calculated by forming a linear combination of at least two filtered measurement signals; generating one or more stimulation signals to electrically stimulate tissue of the subject; outputting the plurality of vector signals and the one or more stimulus signals; determining adjusted filter coefficients of the digital filter during or after the generation and output of the one or more stimulus signals based on current filter coefficients, the one or more stimulus signals, and the plurality of measurement signals collected while the one or more stimulus signals are being output for stimulation; after the adjusted filter coefficients have been determined, filtering the plurality of measurement signals using the adjusted filter coefficients; A computer program having program code means for causing a computer to execute the program.

15. The physiological measurement device of claim 1, when executed by a processor, obtaining a plurality of measurement signals collected from a subject; filtering the plurality of measurement signals using a digital filter and calculating a plurality of vector signals from the filtered plurality of measurement signals, the vector signals representing differential signals and calculated by forming a linear combination of at least two filtered measurement signals; generating one or more stimulation signals to electrically stimulate tissue of the subject; outputting the plurality of vector signals and the one or more stimulus signals; determining adjusted filter coefficients of the digital filter during or after the generation and output of the one or more stimulus signals based on current filter coefficients, the one or more stimulus signals, and the plurality of measurement signals collected while the one or more stimulus signals are being output for stimulation; after the adjusted filter coefficients have been determined, filtering the plurality of measurement signals using the adjusted filter coefficients; A non-transitory computer-readable recording medium that stores a computer program for executing the above.

Images