Linearization of signals in measuring devices

Abstract

A method for pre-processing electrical sensing signals to compensate for common-mode to differential-mode conversion caused by nonlinearities in the signal transfer function, applying a linearization function to the input signals to cancel the nonlinearities. The method includes a calibration process in which a plurality of sets of standardized test signals are applied to the input channels of a measurement operation, and corresponding test outputs are measured for each set of test signals. Parameters of the linearization function are set based on a data set of acquired test outputs.

Description

[Technical Field]

[0001] The present invention relates to a method for linearizing input signals to a measuring device for calculating measured values ​​of biological signals. [Background technology]

[0002] Differential measurement is a type of measurement of a physical property (e.g., pressure, temperature, voltage), and the associated output from this measurement is the difference y of the property measured at two points x1 and x2, where the absolute values ​​at each point are irrelevant. The value of the physical property at each point is modeled as including a common-mode component c and a differential-mode component d. The differential-mode component d occurs with opposite signs at each point x. For example, the values ​​x of the property at two points, and the ideal output measurement y from these points are, for example, x1 = 0.5d + c x² = -0.5d + c y ideal =x1-x2=0.5d+c-(-0.5d+c)=d That is the case.

[0003] In an ideal scenario, calculating the difference between x1 and x2 removes the common-mode component c, leaving only the differential-mode component d.

[0004] However, in some actual system, signal x n Before performing the calculation of the measured values, the transfer function f n It is also subject to the constraints of the following: The transfer function arises from the operation of electrical components in the circuit, such as semiconductor components like diodes, thyristors, and transistors, as well as from the operation of more complex circuits, such as operational amplifiers and AD converters. y real =f1(x1)-f2(x2)=g(d,c)

[0005] The transfer function f(x) is a function of d and c, g, where y is the actual output value. This is called common-mode to differential-mode conversion (CM2DM), where c is usually much larger than d. Even a small amount of c appearing in the output affects the signal and masks part or all of d, significantly reducing the accuracy and reliability of the measurement.

[0006] Assuming the transfer function f is linear, CM2DM is (x n This can be reduced by making these transfer functions more equivalent (regardless of the value of ). However, if the transfer function is nonlinear, these input signals have different offsets and are located at different positions due to the nonlinearity, which can result in slightly different small signal gains for the input signals, so simply making the transfer functions more equivalent is not sufficient.

[0007] Figure 1 schematically shows exemplary linear transfer function 14 and nonlinear transfer functions 12a and 12b. As can be seen from the figure, in the nonlinear transfer function, the resulting gain for a signal is a nonlinear function of the input signal value (x axis). The y axis of the graph shows the output value corresponding to a given input signal value (x axis). The gain for small signals with at least a constant offset corresponds to the slope of the curve. Therefore, the gain applied by the transfer function depends on the offset of the input signal, as this offset changes the position of the baseline of this input signal along the nonlinear transfer function.

[0008] This offset is understood as part of the differential component d. The primary cause of the offset is the movement of the skin-electrode contact surface, which acts as an unintended battery generating potentials up to several hundred millivolts. Users of the measuring device are typically not interested in this offset but are interested in the component d, which is due to physiological causes. Physiological electrical activity most typically takes the form of an AC signal component. The aforementioned offset may in some cases be the “true” differential component, but this is irrelevant if the user is only interested in the physiological (AC) component of the signal.

[0009] The nonlinearity of the input channel transfer function is caused by components in the input path that exhibit significant nonlinear behavior, ranging from passive semiconductors (e.g., diodes or thyristors used for overvoltage protection) to operational amplifiers and analog-to-digital converters (ADCs).

[0010] The nonlinear behavior of the input channel can lead to significant CM2DM even if the linear components of the transfer function are kept the same.

[0011] As an example, Figure 2 illustrates how the nonlinear transfer function affects the differential signal calculated in the measurement system. Figure 2(a) shows the true differential signal d (simulated). Figure 2(b) shows two sensor signals x1 and x2, each including the common mode c, differential mode d, and their respective offsets.

[0012] Figure 2(c) shows the reconstructed differential signal 16 calculated by the measuring device, where each input signal is constrained by its transfer function (f1 and f2 respectively), and the differential signal is simply calculated as the difference between the two input signals. For ease of comparison, Figure 2(c) also shows the true differential signal 18 on the same axis. As can be seen from Figure 2(c), the calculated differential signal has a lot of noise with an amplitude that is roughly greater than the magnitude of the differential signal itself, resulting in a poor signal-to-noise ratio.

[0013] Differential measurement is widely used in electromedical applications. For example, well-known measurement modalities such as electrocardiogram (ECG) and electroencephalogram (EEG) recording (which measure the electrical activity of the heart and brain, respectively) are differential measurements. Other less well-known measurement modalities include electroneurography (NEG) and electromyography (EMG), which are also differential measurements. Newer electroanatomical imaging applications are also partly based on differential voltage measurement.

[0014] The signals of interest in electromedical applications are generally time-varying (AC) signals generated by the physiological activity of muscle or nerve tissue.

[0015] The source signal has a common-mode component c, a differential-mode component d, and an offset k, and the sampled signal is further constrained by its transfer function. The offset k is different for each source signal. The offset is caused, for example, by the electrochemical properties of the contact surface between the patient and the measurement system. For example, in the case of an ECG measuring device, this typically consists of adhesive electrodes attached to the patient's skin. The most common source of the common-mode component is power line interference, which can be introduced into the measurement through various possible paths / routes. In particular, interference can be introduced into the measurement by the patient (capacitive if the patient is near the interference source, or directly by conduction if the patient is in contact with electrical ground), through cabling (even more so if the cables are not completely shielded), or through measuring electronic equipment (the patient module is usually galvanically insulated, but still has a capacitive connection to ground).

[0016] Other sources of common-mode interference include, for example, electrical devices operating near or in contact with patients, measuring devices, or simply on the same power supply circuit, whose characteristics (frequency, amplitude, changes over time, etc.) differ significantly from typical power line interference patterns.

[0017] Common-mode interference is addressed in state-of-the-art devices by a processing operation in which a portion of the common-mode component is removed (eliminated) to compensate for the estimated interference contribution. The ratio of the common-mode component that is removed is known as the common-mode rejection ratio.

[0018] For example, known devices can address common-mode interference by means such as a driven-right-leg loop, shielding of recorders and cable wiring, use of precision components, and linear equalization of input channels.

[0019] In cases where interference occurs in a clinical environment, sometimes the device may include functions for performing sub-dermal treatment before attaching electrodes, verifying whether the electrodes are correctly attached to the patient, periodically replacing the electrodes, and checking and instructing the operator to remove electrical devices in the vicinity of the patient that may be sources of interference. If this function cannot reduce the interference, further suggestions such as rotating the patient or the device by 90 degrees may be made.

[0020] Nonlinearity is simply considered to affect the output gain, and the limits of the manufacturing test are set to maintain this output gain within an acceptable range. This ensures that the device complies with applicable standards and regulations, but still allows for the occurrence of significant CM2DM due to the nonlinearity of the transfer function. SUMMARY OF THE INVENTION PROBLEMS TO BE SOLVED BY THE INVENTION

[0021] State-of-the-art electrophysiological recorders do not consider the effect of the transfer function of possible non-linear input channels on the device's common-mode rejection ratio. A method for addressing this non-linearity would be valuable. MEANS FOR SOLVING THE PROBLEMS

[0022] The present invention is defined by the claims.

[0023] According to one aspect of the present invention, a computer-implemented method for performing on a measurement device is provided. The method has a differential measurement operation, and the differential measurement operation includes a step of reading signals from at least two input channels of the measurement device, wherein the signals from each input channel are received from respective sensor elements, a step of applying a linearization function to each of the input channels, wherein the linearization function has adjustable parameters, after applying the linearization function, a step of applying a calculation algorithm to the input channels to derive at least one output measurement value, and a step of outputting a data signal indicating the output measurement value. The method further has a calibration operation performed at a time different from the differential measurement operation, and the calibration operation

[0024] includes a step of applying a plurality of sets of calibration input signals of a constant value to the input channels of the measurement device, a step of applying the calculation algorithm to each set of the calibration input signals without applying the linearization function to form a calibration data set and recording each output measurement value, and a step of fitting the adjustable parameters of the linearization function based on the calibration data set. The method is thus based on applying linearization to the input signal channels, where the parameters of the linearization function are configured based on a calibration operation in which a standardized test signal is applied to the input channels of the measurement device and the corresponding output measurement signals are recorded. Based on the relationship between the test input and the measured test output, an appropriate set of parameters of the linearization function can be determined. The method further has a calibration operation performed at a time different from the differential measurement operation, and the calibration operation

[0025] The signals may be electrophysiological signals.

[0026] The signals may be electrophysiological signals.

[0027] In some embodiments, the linearization function may be a polynomial function.

[0028] In some embodiments, it is assumed that the signal received in each input channel represents a true sense signal processed by a nonlinear transfer function, where the fitting of tunable parameters of the linearization function involves an optimization process that minimizes the error between the true output measurement for a set of calibrated input signals, as would be obtained if the nonlinear transfer function were not applied to the input signals, and the actual output measurement recorded in the calibration dataset.

[0029] It should be noted that the nonlinearity of the transfer function is typically caused primarily by the influence of electrical components within the measuring device itself, rather than by the sensing device located outside the measuring device. Since cables are modeled as a network of linear components such as resistors and capacitors, the asymmetry introduced by the external parts of the measurement setup is primarily a linear effect. For this reason, the calibration process can be achieved by applying a constant value calibration reference signal directly to the input port. Nonlinearity is induced within the electronics of the measuring device itself (e.g., from protection diodes / thyristors, operational amplifiers, A / D converters, and power supplies).

[0030] In some embodiments, the calculation algorithm involves applying one or more vectors to the input channels, each vector defining a mapping from a set of input channels to an output measurement. The output measurement is a linear combination of the input channels, the weights of which are defined by the elements of the vectors. Calibration of the parameters of the linearization function is performed in part on the one or more vectors. In other words, calibration of the linearization function is performed in part on the calculation algorithm used to calculate the output measurement. For example, simply put, calibration of the parameters of the linearization function may involve a process of minimizing the difference or error between the expected output obtained by applying the vectors to the calibration input signal and the actual output obtained by applying the vectors to the calibration input signal.

[0031] In some embodiments, the measurement operation may involve the application of multiple output vectors to derive multiple different output measurements.

[0032] In some embodiments, the measuring device may have three or more input channels. In some embodiments, each of the multiple output vectors defines a mapping from only a subset of the input channels to its respective output measurement value.

[0033] In some embodiments, the measurement operation involves referencing a dataset of multiple different linearization functions, where each linearization function is associated with only a subset of the multiple output vectors, and only the linearization function associated with a given output vector is applied to a given input channel before each output vector is applied. In other words, different input channels may be constrained by different linearization functions, and each input channel may be processed (operating separately or in parallel) using two or more linearization functions depending on the particular output vector to be applied.

[0034] For example, a simple illustrative case is a system with four input channels, where the computation algorithm has the application of two vectors: the first vector calculates the difference between the first and second input channels, and the second vector calculates the difference between the third and fourth vectors.

[0035] In this case, the optimal linearization of the first and second input channels to which the first vector is applied is performed separately from the linearization of the third and fourth input channels to which the second vector is applied. This is because, for example, the pair of the first and fourth input channels is not used together when calculating one of the two vectors, so the calibration procedure does not need to consider how well the linearity of the first and fourth input channels match.

[0036] In some embodiments, each input channel is assumed to have a differential mode component, a common mode component, and an offset, where the calibration procedure involves referring to a default probability distribution for the offsets of the input channels. This probability distribution is used either explicitly (actively) or implicitly (implicitly) in the calibration algorithm to guide the selection or determination of an appropriate estimated offset for use in the linearization function.

[0037] For example, the calibration procedure first records a calibration dataset obtained by applying a computational algorithm to each set of calibration input signals without applying a linearization function, and this calibration dataset has the respective calibration output values ​​for each calibration input signal applied and for each of the various possible offset values. This calibration procedure then, as an example, involves generating a cost function, which is calculated as the sum of the absolute differences in the slopes of these calibration output values ​​as a function of the calibration input values ​​for all possible combinations of offsets. In other words, for each possible combination of offsets of the input channel, each set of calibration output values ​​is obtained as a function of the calibration input signal applied. Each of these sets of calibration output values ​​defines the slope (as a function of the calibration input value) with respect to the calibration input value applied. The cost function is calculated based on the difference of these slopes for different combinations of offsets, and the optimization (fitting procedure) is based on minimizing this cost function (thereby linearizing the input channel).

[0038] When the device is operating normally, if all possible offset combinations are equally likely to occur, then further improvement of the cost function is impossible. Optimal fitting of the linearization function results in the slope of the input channel's transfer function (i.e., the calibrated output value as a function of the applied calibration input signal) being identical for all combinations of input channel offsets. This is equivalent to linearizing each individual input channel.

[0039] For example, if there is prior knowledge that some combinations of offsets are more likely (or less likely) than others due to the presence of additional circuits such as a Drive Right Leg (DRL) circuit, this knowledge can be applied to give more weight to the more likely offset combinations and less weight to the less likely combinations. Depending on the type of nonlinearity and the linearization function, this can reduce the CM2DM for predicted operating conditions at the expense of worsening the CM2DM under less likely or abnormal operating conditions.

[0040] In some embodiments, the linearization function and calibration operation are configured such that the linearization function being calibrated provides a more optimal linearization for the set of offsets having the highest probability in the probability distribution, and a less optimal linearization for the set of offsets having the lowest probability in the probability distribution.

[0041] For example, the signal received in each input channel can be assumed to represent a true sensed signal processed by a nonlinear transfer function, where fitting the adjustable parameters of the linearization function involves an optimization process that minimizes the error between the true output measurements for a set of calibrated input signals, as would be obtained if the nonlinear transfer function were not applied to the input signals, and the actual output measurements recorded in the calibration dataset. In this case, a more optimal linearization for a set of possible offsets means a smaller average error in the calibrated output measurements for input signals with the set of possible offsets, and a less optimal linearization for a set of possible offsets means a larger average error in the calibrated output measurements for input signals with the set of possible offsets.

[0042] This allows for the consideration of all possible scenarios, but provides a smooth degradation of the linearization quality as the offset changes from a more common set to a less common set.

[0043] Further aspects of the present invention provide a computer program product having coding means configured to cause a processor to perform a method according to any example or embodiment described above or further below, or according to any claim of this application, when executed on the processor. The coding can cause the processor to perform a measurement operation of the method when the processor is operably coupled to two or more sensor elements that function as sources of input signal channels. The coding can cause the processor to perform a calibration operation of the method when the processor is sequentially ported / supplied with a set of calibration input signals.

[0044] A further aspect of the present invention provides a processing apparatus having an input / output unit and one or more processors. The one or more processors are capable of operating in at least a first mode and a second mode. In the first mode, the one or more processors are adapted to perform a differential measurement operation, which includes at least, A step of reading signals from at least two input channels of the measuring device, wherein the signals from each input channel are received from the respective sensor elements, A step of applying a linearization function to each of the input channels, wherein the linearization function has adjustable parameters, The steps include: applying the linearization function to the input channel to derive at least one output measurement value, and applying the calculation algorithm to the input channel after applying the linearization function; Step 1: Output a data signal indicating the output measurement value. It has.

[0045] In the second mode, one or more processors are configured to perform a calibration operation, and the calibration operation is The step of applying a set of constant value calibration input signals to the input channel of the measuring device, The steps include: applying the calculation algorithm to each set of calibration input signals without applying the linearization function to form a calibration dataset, and recording each output measurement value; Steps to fit the adjustable parameters of the linearization function based on the calibration dataset. It has.

[0046] Further aspects of the present invention provide a measuring device having at least two signal input ports for simultaneously reading at least two input signal channels from each sensor element, and a processing device according to any embodiment or example outlined in the present disclosure, or according to any claim of this application.

[0047] Further aspects of the present invention provide a system having a measuring device according to any embodiment or example outlined in this disclosure as described above, or according to any claim of this application. The system further comprises a signal measuring device having at least two sensor elements for connection to the signal input port.

[0048] These and other aspects of the present invention will become apparent from and be explained with reference to the embodiments described below. [Brief explanation of the drawing]

[0049] For a better understanding of the present invention and to more clearly illustrate how the present invention is carried out, the accompanying drawings are referenced merely as examples. [Figure 1] Figure 1 shows exemplary linear and nonlinear transfer functions. [Figure 2] Figure 2 shows differential and common-mode interference. [Figure 3] Figure 3 shows an exemplary system and processing apparatus according to one or more embodiments of the present invention. [Figure 4] Figure 4 schematically shows an exemplary processing workflow of a measurement operation performed by the processing apparatus according to one or more embodiments of the present invention. [Figure 5] Figure 5 schematically shows an exemplary processing workflow of a calibration operation performed by the processing apparatus according to one or more embodiments of the present invention. [Figure 6] Figure 6 shows an exemplary probability distribution for the offset of a pair of two input signal channels. [Figure 7] Figure 7 provides a more detailed example of the output signals obtained in low-probability, medium-probability, and high-probability offset states when the probability density of the offset is not considered in the linearization. [Figure 8] Figure 8 provides a more detailed example of the output signals obtained in low-probability, medium-probability, and high-probability offset states when the probability density of the offset is considered in the linearization. [Figure 9] Figure 9 shows an exemplary processing flow following an example where a single linearization function is applied to each input signal channel and a single output vector is applied to both input channels. [Figure 10] Figure 10 shows a further exemplary processing flow following a further example where two output vectors are applied to different selected combinations of a set of three input channels. [Modes for carrying out the invention]

[0050] The present invention will be described with reference to the drawings.

[0051] The detailed descriptions and specific examples illustrate exemplary embodiments of the apparatus, systems, and methods, but these are for illustrative purposes only and should not be understood as limiting the scope of the invention. These and other features, aspects, and advantages of the apparatus, systems, and methods of the invention will be better understood from the following description, the appended claims, and the appended drawings. The drawings are for illustrative purposes only and are not drawn to a specific scale. Also, the same reference numerals are used throughout the drawings to indicate the same or similar parts.

[0052] The present invention provides a method for preprocessing an electrically sensing signal to compensate for a common-mode to differential-mode conversion caused by the nonlinearity of the signal transfer function. The nonlinearity is neutralized by applying a linearization function to the input signal. This method includes a calibration process in which multiple sets of standardized test signals are applied to the input channel of the measurement operation, and for each set of test signals, a corresponding test output is measured. The parameters of the linearization function are set based on a dataset of acquired test outputs.

[0053] As a further example of this concept, according to one exemplary set of preferred (but not exhaustive) embodiments, whose details are further elaborated in subsequent sections, one or more linearization functions h before calculating the difference or vector that defines the output measurement n A method is provided that attempts to solve the problem of common-mode to differential-mode conversion due to the nonlinearity of the transfer function of the input channel by providing a measuring device that applies (x) to the input signal. The linearization function uses parameters determined in a calibration procedure by applying a constant sequence of input signals to the input of the device and recording the (raw / unlinearized) output. The dataset of calibrated input and calibrated output values ​​is used together with information about the vectors that the device uses when calculating the measurements, and preferably a predetermined statistical distribution of predicted signal offset values ​​(or combinations of offset values) assumed to be included as components in each input signal, to calculate linearization parameters that provide optimal linearization for likely combinations of channel offset values ​​and suboptimal linearization for rare combinations.

[0054] Figure 3 shows an exemplary processing apparatus 22 and system according to one or more embodiments of the present invention.

[0055] The processing unit 22 has an input / output (I / O) unit 26 and one or more processors 24. In this example, the I / O has two input channels 32a, 32b and one output channel 36. However, in other examples, the I / O may have three or more input channels and three or more output channels.

[0056] Each sensor element 52a and 52b is connected to its respective input channel 32a and 32b. Each sensor element is adapted to acquire biosignals at a specific location on the subject's body. For example, each of these sensor elements may be an ECG electrode. The signals from each sensor element 52a and 52b supply their respective input signals 54a and 54b to the input channels 32a and 32b.

[0057] One or more processors 24 (only one is shown in Figure 3, but in practice an assembly of multiple processors may be used) receive input signals from input channels 32a, 32b, process these signals to derive one or more output measurements 38, which are then exported as measurement data signals 42 via output channel 36.

[0058] One or more processors 24 are particularly adapted to perform a computer implementation method including differential measurement and calibration operations, which are performed at different times.

[0059] Figure 4 schematically shows the processing workflow for differential measurement operation. Figure 5 schematically shows the processing workflow for calibration operation.

[0060] The differential measurement operation has the following steps, which are performed by one or more processors 24.

[0061] This measurement operation includes the step of reading signals from at least two input channels (channel A 32a and channel B 32b) of the measuring device. The signals from each channel are received from the respective sensor elements 52a and 52b. The measuring device may include analog-to-digital conversion means that are applied to the received input signals and digitize these input signals.

[0062] The measurement operation further comprises the step of applying a linearization function 62 having adjustable parameters to each input channel.

[0063] The measurement operation further includes the step of applying a linearization function 62 to the input channel to derive at least one output measurement value 38, and then applying a calculation algorithm 64 to the input channel. The input channel supplied to the calculation algorithm is preferably a single-ended input channel, and the output from the measurement algorithm is preferably a single-ended output signal.

[0064] A data signal 42 indicating the output measurement value 38 is output. This data signal is output via the output channel or port 36 of the input / output unit 26.

[0065] The measurement operation can be performed entirely or partially using software running on hardware (ASIC, FPGA) and / or a microprocessor (microcontroller, digital signal processor).

[0066] One or more processors may be further configured to perform a calibration operation having the following steps shown in Figure 5.

[0067] This calibration operation involves applying a plurality of n sets 72 of constant value calibration input signals 74a, 74b to the input channels 32a, 32b of the measuring device 22. These n sets are applied one by one in sequence. The calibration input signals of each set are applied to the input channels simultaneously. The constant signal value of the calibration input signals is sometimes referred to elsewhere as the "calibration point".

[0068] The calibration operation further comprises the steps of applying a calculation algorithm 64 to each pair 72 of calibration input signals 74a, 74b without applying a linearization function to form a calibration dataset 44, and recording each output measurement value 38. Each output measurement value 38 may be output via the output channel 36 of the I / O 26 in the formation of a data signal 42. The calibration dataset can be stored in a local data store of the device 22 or in a remote data store.

[0069] The calibration operation further includes the step of fitting adjustable parameters of the linearization function 62 based on the calibration dataset 44.

[0070] With respect to measurement and calibration operations, the processing unit 22 is capable of operating in at least a first and a second mode, in which the processing unit is adapted to perform a measurement operation, and in which the processing unit is adapted to perform a calibration operation.

[0071] The calibration operation is performed at a different time than the measurement operation. The calibration operation only needs to be performed once, for example, at the factory or during pre-use initialization, before use for measurement. Alternatively, the calibration operation may be performed on-site using the measuring device. It may be performed repeatedly.

[0072] The measuring device 22 may include a user interface that allows the user to provide user input to control the selection of a first or second mode. These mode selection options may be available only in a specific initial state or mode of the device, such as when the device first enters a setting or setup mode. In some examples, the measuring device may adapt to enter a calibration mode in response to receiving a default control signal from a separate secondary device via a data connection. The data connection may be facilitated by a dedicated control or setting port included in the measuring device. The data connection may be detachable from the port of the measuring device. For example, the user may be supplied with a separate device that is connectable to the measuring device and automatically prompts the execution of the calibration step. In some examples, the calibration operation may be an algorithm or computer program stored in the secondary device, in which case the provided invention may have the measuring device in combination with this secondary device.

[0073] The function to perform calibration operations on-site after manufacturing allows for periodic recalibration, taking into account the aging of components and the time-dependent changes in the nonlinearity of the transfer function.

[0074] In further examples, the two modes may not be freely selectable by the end user. For instance, the device may be set to measurement operation mode at the time of shipment, where the calibration mode is accessible only via a dedicated data input port and / or using a key available at the factory.

[0075] In some examples, the measuring device is an ECG measuring device. In some examples, the measuring device may have a user interface for displaying the measured value 38 and / or for receiving user control commands.

[0076] The calculation algorithm 64 applied in the measurement operation comprises applying one or more vectors to the input channels, each vector defining a mapping from the set of input channels to an output measurement, where the output measurement is a linear combination of the input channels and the weights defined by the elements of the vectors. The one or more vectors applied in the measurement operation are stored in a local data store, which may be pre-stored, for example, during manufacturing or setup and retrieved during the execution of the measurement operation.

[0077] The aforementioned vector can define a linear combination of input channels, where the sum of the weight coefficients is zero. For example, in the simplest case, the output measurement can simply be the difference between two input signals. In this case, the coefficients are -1 and 1, and their sum is zero.

[0078] In some embodiments, the measurement operation may have the application of multiple output vectors to derive multiple different output measurement values ​​38.

[0079] In some embodiments, the measuring device has three or more input channels, where each of the plurality of output vectors defines a mapping from only a subset of these input channels.

[0080] The minimum list of the features described above, when combined, results in a measuring device that is more resistant to interference and provides a clean, uninterrupted output signal in environments where devices not using the method of the present invention exhibit significantly degraded output signals due to interference. This leads to more accurate measurements, which in turn leads to more accurate diagnoses and subsequent treatments.

[0081] Further details regarding linearization will be explained.

[0082] According to a simple example, the one or more linearization functions h n (x) may also be a polynomial function, where the tunable parameters of the linearization function are, for example, h(x) = ax 3 + bx 2 + cx + d are the coefficients of a polynomial such as this. Since the number of multiplication operations required to calculate the value of the polynomial is only slightly small, the calculation efficiency is good. Based on the preferable balance between calculation efficiency (low order) and accuracy (high order), a low-order or high-order polynomial can be used.

[0083] One or more linearization functions are preferably monolithic functions, but since these functions need to show sufficient smoothness to operate efficiently, instead, piecewise functions such as, for example, cubic splines or higher-order splines may be used.

[0084] In a further embodiment, the one or more linearization functions are based on other functions as the basis of h n (x), for example, trigonometric functions, for example, h(x) = a * tanh(b * x + c) + d such as the hyperbolic tangent tanh(x) can be used.

[0085] The calculation of the coefficients based on the compiled calibration dataset 44 is given as a convex optimization problem. This can be easily solved using an automatic function fitting algorithm and solution (solver) such as, for example, a regression algorithm.

[0086] Here, one implementation according to at least one set of these embodiments will be described in more detail. In this set of embodiments, the calculation of the linearization parameters utilizes one or more vectors used in calculating the measurement values in the measurement operation.

[0087] Briefly speaking, during the calibration operation, the following information, the value of each pair of calibration input signals, the calculations performed on these signals, and the measurement results provided as output It is known that, based on knowledge of the calculations performed on these signals (i.e., applying vectors to the input signals), it is theoretically possible to identify what the “predicted” output measurement would be in the absence of CM2DM interference. Thus, the difference between the predicted output and the measured output can be used to derive the fitting of a linearization function. In other words, this fitting of the linearization function aims to minimize the error between the “true” predicted measurement output based on known inputs and known calculation functions 64 and the actual measurement output during calibration. Since multiple sets of calibration signals are applied, and ideally define differences that increase or decrease regularly, the nonlinearity of the error is shown in the calibration dataset, and this is appropriately taken into account in the fitting of the linearization function.

[0088] Therefore, in summary, it is assumed that the signal received in each input channel represents the true sensed signal processed by a nonlinear transfer function, where fitting the tunable parameters of the linearization function to the calibration dataset involves minimizing the error between the theoretical true output measurement for each calibration input signal, which would be obtained if the nonlinear transfer function were not applied to the input signal, and the actual output measurement recorded in the calibration dataset.

[0089] As a further example, with respect to calibration input signals, for the linearization of a single input channel, the set of calibration points (i.e., signal values ​​of the calibration input signals) is simply the set of calibration input signals for that single channel. For example, if the input range is 0V to 1.0V, the calibration signal values ​​can be selected as 0V, 0.25V, 0.5V, 0.75V, and 1.0V. This list of voltages corresponds to the set of calibration points. This also applies to multi-channel systems where each input channel is linearized independently of the others.

[0090] In the case of multi-channel linearization that takes into account the combination of input channels, the set of calibration points consists of combinations of calibration input signal values. In the example using two channels and the above list of voltages, all sets of calibration points are the total combinations of {0V, 0.25V, 0.5V, 0.75V, 1.0V} × {0V, 0.25V, 0.5V, 0.75V, 1.0V}. 2 = 25 (also known as the Cartesian product), consisting of (0V, 0V), (0V, 0.25V), ..., (0.25V, 0V), (0.25V, 0.25V), ..., (1.0V, 1.0V). The size of this set increases exponentially with the number of input channels, and in a system with 10 input channels (e.g., a 12-inductor ECG recorder) and 5 test voltages, all sets of calibration points contain approximately 10 million elements.

[0091] In the simplest case, all of these combinations involve the process of fitting a linearization function. However, while state-of-the-art processors can handle such a data load, the calibration process can be accelerated by using only a subset of possible calibration points for fitting the linearization function. Based on knowledge of the vectors used in the measurement operation (for example, the knowledge that an ECG measurement operation does not form a vector containing more than two input chest leads), elements are excluded from the set of data points.

[0092] These elements can also be excluded based on the knowledge that they do not present conditions under which the device is expected to generate a valid output, or that they do not actually occur because, for example, a drive right leg (DRL) circuit exists that averages the input channels to 0.5V. This means that, for example, combinations of input signals such as (1.0V, 1.0V) and (0V, 0V) can be ignored in calculations. For example, during normal operation, the combinations (0V, 1.0V), (0.25V, 0.75V), (0.5V, 0.5V), (0.75V, 0.25V), and (1.0V, 0V) are the most appropriate, and any other combinations represent a temporary or abnormal / faulty state.

[0093] As mentioned above, one difficulty is that in actual use, the sensor measurement input signals 54a and 54b are expected to each have specific offsets in addition to the standard differential-mode and common-mode components. These offsets depend on the physical characteristics of the device setup, as described above. If the offsets are not taken into account in the linearization function itself, these offsets will cause corresponding nonlinear distortions in the output measurement signal, thus creating significant difficulties in processing operations involving the application of nonlinear functions (i.e., the linearization function h(x)). However, these offsets cannot be known in advance. Therefore, the most appropriate solution is to use the estimated probability distribution of the input channel offsets and set or adjust the parameters of the linearization function so that offsets within at least the highest probability range of this probability distribution do not cause very large distortions in the output measurement signal. Ideally, the linearization function is also adjusted so that the resulting distortion with respect to the offsets changes relatively smoothly as a function of the input signal offsets. In other words, as the (possible) offset of the signal moves from a high-probability range to a low-probability range, it is preferable that there is no large jump or discontinuity in the output measurement signal, but rather that the resulting signal distortion increases smoothly, and therefore the accuracy of the signal deteriorates smoothly.

[0094] Accordingly, according to the set of embodiments described above, it is assumed that each input channel has a differential mode component, a common mode component, and an offset, where the calibration procedure involves referring to a predetermined probability distribution for the offsets of these input channels.

[0095] In other words, the calibration dataset 44 of calibration input and calibration output values ​​is used together with information about the vectors the device uses when calculating the measurements, and a default statistical distribution of the offset values ​​(or combinations of offset values) of the input signals that are assumed to be included as components in each input signal, to calculate linearization parameters that provide optimal linearization for the most likely combinations of channel offset values ​​and suboptimal linearization for the rarer combinations.

[0096] In particular, in some embodiments, the linearization function and calibration operation are configured such that the linearization function being calibrated provides a more optimal linearization for the set of offsets with the highest probability in the probability distribution, and a less optimal linearization for the set of offsets with the lower probability in the probability distribution.

[0097] In this situation, a more optimal linearization for a set of possible offsets means that, after calibration, the average error in the output measurement for an input signal having the set of possible offsets is smaller, and a less optimal linearization for a set of possible offsets means that, after calibration, the average error in the output measurement for an input signal having the set of possible offsets is larger.

[0098] In applications using only a single channel, the quality of linearization is measured by the deviation of the transfer function (i.e., the function of the calibrated output value as a function of the applied calibrated input signal) from an ideal linear function.

[0099] However, in the types of applications described in the disclosure of this invention, while simply linearizing each input channel in this way may provide a significant improvement over the uncalibrated state, given the degrees of freedom of the linearization function, it may not be optimal in terms of reducing the conversion from common mode to differential mode.

[0100] In particular, optimal linearization is achieved when CM2DM is not present in the predicted operating conditions. Predicted operating conditions may be constrained by factors such as the presence of a drive right leg (DRL) feedback loop (see below) in the input signal path, which causes the sum of the input signal potentials to be fixed. The optimization algorithm assigns greater importance to CM2DM by reducing it when the sum of the input electrode potentials is close to this fixed value, and less importance or no importance at all when the sum differs from the predicted value.

[0101] As an example, Figure 6 shows an exemplary probability distribution for the offset of a pair of input channels having only a first input channel (channel A) and a second input channel (channel B), where the system setup includes a driven right leg (DRL) circuit. The graph shows, (Channel A + Channel B) / 2 = 0.5 The scaled probability density function for the given system is shown.

[0102] To elaborate further, electrophysiological measurement devices often implement a feedback circuit that feeds back to the patient the negative sum of the potentials of the input electrodes (a constant positive or negative offset) via a designated electrode. In ECG, this is often the right bundle branch electrode, and therefore the circuit is called a driving right bundle branch “DRL” circuit or right bundle branch driven “RLD” circuit, but generally, electrodes attached to any part of the body can be used. In other measurements, the terms “active electrode,” “reference electrode,” or “driving electrode” may be used, and the term “feedback circuit” may be used instead of DRL circuit.

[0103] The primary purpose of a DRL circuit is to reduce common-mode signals. This generally works very effectively at low frequencies, but its effectiveness decreases at higher frequencies due to instability in the control loop and limitations on the current flowing from the device to the patient.

[0104] Another effect of the DRL circuit is establishing an operating point. The DRL circuit brings the potentials of the input electrodes to a state where the sum of their potentials is approximately equal to a constant offset. Under normal conditions, this operating point is reached quickly (within 1 second). Therefore, in this scenario, the input channel voltages are assumed to be close to the calibration points of (1.0V, 0V), (0.75V, 0.25V), (0.5V, 0.5V), (0.25V, 0.75V), and (0V, 1.0V) with respect to a constant offset of 0.5V in this example.

[0105] This is because, for example, any positive integer power of x is monotonic for x >= 0 (where x is the value of the input signal to the input channel), making it a reasonable choice to use a polynomial as the linearization function.

[0106] However, it should be noted that the ADC connected between the input port and the processor often uses bipolar sampling, and the midpoint of the ADC range is 0V. This must be taken into consideration when selecting and applying the linearization function.

[0107] Referring to Figure 6, the measurement range is 0 to 1, with the midpoint being 0.5. The right-foot drive (RLD) operates to maintain the average of the two channels at 0.5. In this particular case, this means that combinations of input channel offsets that sum to zero are more likely, and combinations of input channel offsets that sum to a non-zero value are less likely. This is because the latter combinations occur only during DRL control loop settling (which is a short, transient period) or under abnormal operating conditions. In some embodiments, this is taken into consideration when determining the probability distribution of the offsets. For example, the optimization process may ignore the quality of linearization for the above situation, especially if the linearization is further improved for more normal situations.

[0108] Furthermore, if all electrodes are properly and similarly mounted and used for the same duration, the DC offsets they generate will be approximately equal. On the other hand, if these electrodes have different aging degradation or mounting quality, the DC offsets of these electrodes are more likely to be different. The optimization process can more strongly optimize linearization for situations where x1 ≈ 0.5 and x2 ≈ 0.5 (where x1 and x2 are the signals from the first and second input channels), but situations where this does not hold indicate an inadequate measurement setup, but should not be completely ignored as they do not indicate an abnormal operating condition.

[0109] The probability distribution function of the offset is determined in advance and stored for later use. It can be determined by empirical testing, theoretical / analytical modeling, or simulation.

[0110] It is not essential for fitting a linearization function to consider the probability distribution of the input signal channel offset. However, if the calculation of the linearization parameters does not consider the probability distribution function, the result may (paradoxically) be worse under probable conditions and better under unprobable conditions.

[0111] This is illustrated, for example, in Figures 7 and 8. Each graph in each figure shows the “true” input signal 82 from the sensor element, the actual recorded (uncorrected) input signal 84, and the corrected input signal 86 after the application of the linearization function.

[0112] Figure 7 shows the results for a set of different probability offset conditions in a case where the linearization function is not fitted to account for the predicted probability distribution of signal offsets. Figure 8 shows the results for the same set of different probability offset conditions in a case where the linearization function is fitted to account for the predicted probability distribution of signal offsets. In the scenario in Figure 7, it can be seen that common-mode interference suppression is actually significantly better in the less likely offset scenario than in the more likely offset scenario. In contrast, in the case in Figure 8, signal correction is best in the more likely offset scenario and worst in the less likely offset scenario (which is clearly a favorable result).

[0113] The probability density function can be used to fit a linearized function, and in the cost function used for fitting / optimization, higher weights can be applied to more likely operating conditions, and lower weights to less likely or abnormal operating conditions. In this context, operating conditions refer, for example, to more likely and less likely combinations of input signal offsets, and / or more likely or less likely combinations of input signal values.

[0114] These applied weights cause the optimization process to improve the linearization of states with higher weights at the expense of states with lower weights.

[0115] To further illustrate this, we mentioned above that to increase the speed of computing the linearization function fitting, we completely exclude certain less likely combinations of possible input channel signals from the calibration dataset. However, the use of probability distributions allows for a more sophisticated approach. This probability distribution allows different combinations of input signals and / or different combinations of offset values ​​to be assigned different weights in the optimization operation (i.e., fitting the parameters of the linearization function) according to their estimated probabilities under normal operating conditions.

[0116] For example, continuing with the two channel examples mentioned above, the input voltages near the calibration point (0.5V, 0.5V) represent the operating conditions when the electrode is new and properly attached to the patient's skin, whereas the input voltages near the calibration points (1.0V, 0V), (0V, 1.0V), (0.25V, 0.75V), and (0.75V, 0.25V) occur when the electrode has deteriorated due to prolonged use (the latter is generally less likely than the former). Furthermore, the input voltages near the calibration points (0.25V, 0.5V), (0.75V, 0.5V), (0.5V, 0.25V), and (0.5V, 0.75V) represent temporary states during which the device should still be producing a noise-free output.

[0117] To mitigate CM2DM caused by channel nonlinearity, the optimization algorithm used in the calibration operation can determine the parameters of a linearization function that makes the slope (or a reasonable approximation of the slope) of the output measurement as similar as possible, as a function of the applied set of input channel signals, at each calibration point (where the calibration point means a specific combination of input channel signals).

[0118] One way to mathematically describe this cost function is the sum of the absolute differences in these slopes at each calibration point.

[0119] For example, the calibration procedure first records a calibration dataset obtained by applying a computational algorithm to each set of calibration input signals (i.e., at each calibration point) without applying a linearization function, and this calibration dataset has respective calibration output values ​​for each applied calibration input signal and for each of the various possible offset values. The calibration procedure then, for example, generates a cost function calculated as the sum of the absolute differences in the slopes of the calibration output values ​​as a function of the calibration input values ​​for all possible combinations of offsets. In other words, for each possible combination of offsets of the input channel, each set of calibration output values ​​is obtained as a function of the applied calibration input signal. Each of these sets of calibration output values ​​defines a slope (as a function thereof) with respect to the applied calibration input value. The cost function is calculated based on the differences in these slopes for various combinations of offsets, and the optimization (fitting procedure) is based on minimizing this cost function (thereby linearizing the input channel).

[0120] To optimize the process for the bias that minimizes CM2DM at the most common operating point, the absolute difference in slope at each calibration point can be multiplied by a weighting coefficient correlated with the probability of operating near that calibration point. This probability can be expressed as a probability density function (PDF). The probabilities used in the probability distribution function can be estimated.

[0121] Each calibration point corresponds to a set of calibration input signals to which it is applied, where each calibration input signal includes, for example, the sum of the main signal component and the calibration offset component. In this way, the calibration can take into account various possible combinations of input signal values ​​(in differential mode) and offsets.

[0122] Regarding the probability distribution of the offset, this can be determined before the measurement and calibration operations. Different embodiments differ in the assumptions made with respect to the probability density function of these offsets. In some cases, when additional data is not available, a normal (i.e., Gaussian) distribution of the offset is assumed. In other cases, more complex probability distributions (various functions, including, for example, asymmetric distributions) can be used. The probability distribution used may be calculated based on theoretical assumptions, or it may be based on an empirically obtained dataset of signal offsets, which is then statistically analyzed to identify the probability distribution.

[0123] The value of the probability density function at the calibration point is used directly as a weighting coefficient when calculating the calibration coefficient, or the weighting coefficient is derived as a function of the probability density function.

[0124] When two or more output vectors are applied when calculating output measurements, it is advantageous to use slightly different linearization functions (or at least different fitting / optimization versions of the same linearization function) for these different output vectors. For example, the specific mapping behavior applied by a given output vector results in different nonlinearities in the output measurement error. Therefore, having a linearization function that is specific to one or more of these different output vectors improves error reduction in output measurements.

[0125] Accordingly, according to some embodiments, the measurement operation can refer to a dataset of multiple linearization functions (or multiple different fitting / optimization versions of the same linearization function, i.e., having different sets of fitted parameters), where each linearization function is associated with only a subset of the multiple output vectors, and only the linearization function associated with a given output vector is applied to a given input channel before each output vector is applied.

[0126] As an addition or alternative, in some examples, the measurement operation may use multiple different linearization functions to apply to different subsets of one or more input channels. A different linearization function may exist for each input channel. This allows for consideration of differences in nonlinear transfer functions along different input channel paths.

[0127] As an example, Figures 9 and 10 show two exemplary processing flows of a measurement operation. In the example in Figure 9, there are two input signals supplied to the measuring device, namely a first input signal 54a and a second input signal 54b. Each input signal is assumed to be constrained by its respective nonlinear transfer functions 102a and 102b, which introduce an error, which is a nonlinear function of the input signal itself, into the signal. An analog-to-digital converter (ADC) 60 is applied to the signals received by the measuring device. In the example in Figure 9, each received input signal is processed using different linearization functions 62a and 62b. These two signals are then provided as input to a calculation algorithm 64 that generates an output measurement value 38 consisting of a linear combination of the corrected input signals.

[0128] The example in Figure 10 is similar except that there are three input signals 54a, 54b, and 54c, each assumed to be constrained by different nonlinear transfer functions 102a, 102b, and 102c. In this example, two different output measurements 38a and 38b are obtained by applying two different calculation algorithms 64a and 64b. The first calculation algorithm receives only a subset of the input signals, in particular the first input signal 54a and the second input signal 54b (after linearization). The second calculation algorithm 64b also receives only a subset of the input signals, in particular the second input signal 54b and the third input signal 54c (after linearization). Each of the calculation algorithms 64a and 64b is associated with a different subset of several linearization functions 62a, 62, 62c, and 62d. In particular, the linearization functions 62a and 62b are associated with a first calculation algorithm 64a, where the first input signal 54a is processed using the first linearization function 62a before being applied to the first calculation algorithm 64a, and the second input signal 54b is processed using the second linearization function 64b before being applied to the first calculation algorithm 64a. The linearization functions 62c and 62d are associated with a second calculation algorithm 64b, where the second input signal 54b is processed using the third linearization function 62c before being applied to the second calculation algorithm 64b, and the third input signal 54c is processed using the fourth linearization function 64d before being applied to the second calculation algorithm 64b.

[0129] For example, as mentioned above, when performing the calibration to minimize the error between the predicted output and the actual output by applying the measurement vector to one or more input calibration signals, the output measurement vector is taken into consideration.

[0130] The exemplary embodiments described so far have specifically described only systems with two input channels. When there are only two input channels, only one vector is possible to calculate the measurement, which corresponds to calculating the difference between the two input channels.

[0131] As a further illustrative example, consider a system with three input channels labeled x1, x2, and x3. In this case, three calculation vectors are possible to calculate the difference between each individual signal and each of the other two signals. However, since the calculation vectors can also use "virtual" points such as x1-(0.5*x2-0.5*x3), theoretically, an infinite number of calculation vectors are possible in practice. Which vectors to use depends on the application, i.e., which physiological signals are being measured. For example, in the case of a 12-lead ECG measurement device, 12 vectors are typically calculated corresponding to the difference signals between different combinations of the nine input electrodes (three limb electrodes RA, LA, LL and six chest electrodes V1...V6). Other measurement modalities, such as EEG, can calculate vectors between all electrode pairs.

[0132] Therefore, by using the knowledge of the vectors used in the calculation procedure, the number of sets of calibration points that must be considered for each vector during the calibration operation can be reduced. If two calibration points (where the calibration points represent a specific set of input channel signals) differ only with respect to the input channels used by a given vector, then only one of these calibration points needs to be considered in the cost function for that vector.

[0133] For example, in an ECG with multiple chest electrodes (V1, V2, ...), a vector is not used to calculate the difference between two different chest electrodes.

[0134] The principles of the present invention can be applied to any type of sensing or measuring device. Examples include electromedical measuring devices such as ECG measuring devices, electroanatomical imaging systems, EEG recorders, AED devices, fetal / obstetric ECG recorders, and electroanatomical imaging systems. However, the applications are not limited to electromedical devices and can be used in other fields where differential measurements are acquired. For example, embodiments can be usefully utilized in non-clinical applications such as smartwatches or fitness trackers with ECG functionality. Embodiments of the present invention can be usefully utilized in any differential measuring device where the conversion from common mode to differential mode via the nonlinearity of the input channel transfer function becomes problematic.

[0135] In summary, one particularly advantageous (but not exhaustive) set of features of embodiments of the present invention is summarized below.

[0136] The objective of at least one set of embodiments is to address the problem of common-mode to differential-mode conversion due to the nonlinearity of the transfer function of the input channel by providing a measuring device that applies a linearization function h(x) to the input signal before calculating the difference or vector which is the output. The linearization function uses parameters determined in a calibration procedure by applying a constant sequence of input signals to the input of the device and recording the (raw / uncalibrated) output. Information in the input and output value pair is used together with information about the vector used by the device and the statistical distribution of offset values ​​(or combinations of offset values) to calculate linearization parameters that provide optimal linearization for likely combinations of channel offset values ​​and less optimal linearization for rare combinations. Since the rare combinations usually indicate some problem with the measurement, such as poor contact of skin electrodes, this provides graceful degradation of the output measurement when the offset condition changes from good / common to poor / uncommon.

[0137] The embodiments of the present invention described above utilize a processing apparatus. This processing apparatus may generally have a single processor or multiple processors. This processing apparatus may be located within a single housing device, structure, or unit, or it may be distributed among multiple different devices, structures, or units. Accordingly, reference to the processing apparatus adapted or configured to perform a particular step or task may correspond to a step or task performed by any one or more of the processing components, either individually or in combination. Those skilled in the art will understand how such a distributed processing apparatus can be implemented. The processing apparatus includes a communication module or input / output unit for receiving data and outputting data to further components.

[0138] One or more processors in a processing unit are implemented in many ways using software and / or hardware to perform various required functions. Typically, a processor uses one or more microprocessors programmed with software (e.g., microcode) to perform the required functions. This processor may be implemented as a combination of dedicated hardware for some functions and one or more programmed microprocessors and associated circuits for other functions.

[0139] Examples of circuits used in various embodiments of this disclosure include, but are not limited to, conventional microprocessors, application-specific integrated circuits (ASICs), and field-programmable gate arrays (FPGAs).

[0140] In various implementations, the processor may be associated with one or more storage media, which are volatile and non-volatile computer memories such as RAM, PROM, EPROM, and EEPROM. These storage media may be encoded with one or more programs that perform the required functions when executed on one or more processors and / or controllers. The various storage media may be mounted within the processor or controller, or they may be transportable so that one or more programs stored in these storage media can be loaded into the processor.

[0141] Modifications of the disclosed embodiments can be understood and practiced by those skilled in the art in carrying out the claimed invention by examining the drawings, this disclosure, and the appended claims. In the claims, the term “having” is not precluding other elements or steps, and the absence of a statement that there are multiple elements is not precluding that there are multiple elements.

[0142] A single processor or other unit can perform the functions of several of the items enumerated in the claims.

[0143] The mere fact that certain means are enumerated in different dependent claims does not indicate that combinations of these means cannot be used advantageously.

[0144] Computer programs can be stored / distributed on suitable media, such as optical or solid-state media supplied together with or as part of other hardware, but they can also be distributed in other forms, such as via the Internet or other wired or wireless telecommunications systems.

[0145] When the term “adapted to” is used in a claim or specification, it means that the term “adapted to” is the same as the term “configured to.”

[0146] No reference numeral in a claim should be construed as limiting its scope.

Claims

1. A computer-based method for performing measurements on a measuring device, wherein the method is A step of reading signals from at least two input channels of the measuring device, wherein the signals from each input channel are received from the respective sensor elements, A step of applying a linearization function to each of the input channels, wherein the linearization function has adjustable parameters, The steps include: applying the linearization function to the input channel to derive at least one output measurement value, and applying the calculation algorithm to the input channel after applying the linearization function; Step 1: Output a data signal indicating the output measurement value. A differential measurement operation having, The step of applying a set of constant value calibration input signals to the input channel of the measuring device, The steps include: applying the calculation algorithm to each set of calibration input signals without applying the linearization function to form a calibration dataset, and recording each output measurement value; Steps to fit the adjustable parameters of the linearization function based on the calibration dataset. Having, A calibration operation performed at a different time than the differential measurement operation. It has, The signals received in each input channel are assumed to represent the true detection signals processed by a nonlinear transfer function. The step of fitting the adjustable parameters of the linearization function involves minimizing the error between the theoretical true output measurement for each calibration input signal, which would be obtained if the nonlinear transfer function were not applied to the input signal, and the actual output measurement recorded in the calibration dataset. A method used by computers.

2. The method according to claim 1, wherein the linearization function is a polynomial function.

3. The method according to claim 1, wherein the calculation algorithm comprises applying one or more vectors to the input channels, each vector defining a mapping from the set of input channels to an output measurement, the output measurement being a linear combination of the input channels and weights defined by the elements of the vectors, and calibration is performed partially on the one or more vectors.

4. The method according to claim 3, wherein the differential measurement operation has the application of a plurality of vectors to derive a plurality of different output measurements.

5. The method according to claim 4, wherein the measuring device has three or more input channels, and each of the plurality of vectors defines a mapping from only a subset of the input channels.

6. The method according to claim 5, wherein the differential measurement operation involves referencing a dataset of multiple linearization functions, each linearization function being associated with only a subset of the multiple vectors, and only the linearization functions associated with a given vector are applied to a given input channel before each of the vectors is applied.

7. The method according to claim 1, wherein each input channel is assumed to have a differential mode component, a common mode component, and an offset, and the calibration operation involves referring to a predetermined probability distribution for the offset of the input channel.

8. The method according to claim 7, wherein the linearization function and the calibration operation are configured such that the calibrated linearization function provides a more optimal linearization for the set of offsets having the highest probability in the probability distribution, and a less optimal linearization for the set of offsets having a low probability in the probability distribution.

9. A computer-based method for being performed on a measuring device, wherein the method is: A step of reading signals from at least two input channels of the measuring device, wherein the signals from each input channel are received from the respective sensor elements, A step of applying a linearization function to each of the input channels, wherein the linearization function has adjustable parameters, The steps include: applying the linearization function to the input channel to derive at least one output measurement value, and applying the calculation algorithm to the input channel after applying the linearization function; Step 1: Output a data signal indicating the output measurement value. A differential measurement operation having, The step of applying a set of constant value calibration input signals to the input channel of the measuring device, The steps include: applying the calculation algorithm to each set of calibration input signals without applying the linearization function to form a calibration dataset, and recording each output measurement value; Steps to fit the adjustable parameters of the linearization function based on the calibration dataset. Having, A calibration operation performed at a different time than the differential measurement operation. It has, The input channels are assumed to each have a differential mode component, a common mode component, and an offset, and the calibration operation involves referring to a predetermined probability distribution for the offset of the input channels. The linearization function and the calibration operation are configured such that the calibrated linearization function provides a more optimal linearization for the set of offsets with the highest probability in the probability distribution, and a less optimal linearization for the set of offsets with a low probability in the probability distribution. The signals received in each input channel are assumed to represent the true detection signals processed by a nonlinear transfer function. The step of fitting the adjustable parameters of the linearization function includes an optimization process that minimizes the error between the true output measurements for a set of calibration input signals, as would be obtained if the nonlinear transfer function were not applied to the input signals, and the actual output measurements recorded in the calibration dataset. A more optimal linearization for a set of possible offsets means that, after calibration, the average error in the output measurement for an input signal having the set of possible offsets is smaller, and A less-than-optimal linearization for the aforementioned set of possible offsets means that, after calibration, the average error in the output measurement for an input signal with the aforementioned set of possible offsets is larger. A method used by computers.

10. A computer program configured to cause the processor to perform the method described in claim 1 when executed on the processor.

11. Input / Output section, and One or more processors In a processing apparatus having, The one or more processors are capable of operating in a first mode in which the one or more processors perform differential measurement operations, and the first mode is A step of reading signals from at least two input channels of a measuring device, wherein the signals from each of the input channels are received from the respective sensor elements, A step of applying a linearization function to each of the input channels, wherein the linearization function has adjustable parameters, The steps include: applying the linearization function to the input channel to derive at least one output measurement value, and applying the calculation algorithm to the input channel after applying the linearization function; Step 1: Output a data signal indicating the output measurement value. It has, The one or more processors are capable of operating in a second mode configured to perform calibration operations, and the second mode is The step of applying a set of constant value calibration input signals to the input channel of the measuring device, The steps include: applying the calculation algorithm to each set of calibration input signals without applying the linearization function to form a calibration dataset, and recording each output measurement value; Steps to fit the adjustable parameters of the linearization function based on the calibration dataset. It has, The signals received in each input channel are assumed to represent the true detection signals processed by a nonlinear transfer function. The step of fitting the adjustable parameters of the linearization function involves minimizing the error between the theoretical true output measurement for each calibration input signal, which would be obtained if the nonlinear transfer function were not applied to the input signal, and the actual output measurement recorded in the calibration dataset. Processing device.

12. At least two signal input ports for simultaneously reading at least two input signal channels from each sensor element, and The apparatus according to claim 11 A measuring device having the following features.

13. The measuring device according to claim 12, and A signal measuring device having at least two sensor elements for connection to the aforementioned signal input port. A system that has

Images