Scientific instrument support system and method for mitigating spectral drift

Abstract

A scientific instrument support system, related methods, computing devices, and computer-readable media are disclosed. A method for reducing distortion in an optical emission spectrum acquired from an optical emission spectrometer is provided. The method may include acquiring a spectrum recorded using the spectrometer and one or more respective condition parameters indicative of operating conditions at the time the spectrum was recorded. The method may further include providing a model configured to output one or more transformation parameters for a transformation to be applied to the acquired spectrum in response to the one or more condition parameters. The transformation may be applied to the acquired spectrum in accordance with the acquired one or more transformation parameters to reduce distortion in the spectrum due to a mismatch between the operating conditions and a baseline operating condition.

Description

[Technical Field]

[0001] The present disclosure relates to reducing distortions in optical emission spectra caused by various operating conditions of an optical spectrometer. [Background technology]

[0002] Scientific instruments for acquiring optical spectra can include complex configurations of moving parts, sensors, input and output ports, energy sources, and consumable parts. As a result, the acquired spectra are sensitive to changes in the operating conditions of the scientific instrument, such as changes in temperature. Known approaches for correcting optical spectra for temperature changes are disclosed, for example, in International Publication No. 2021 / 018992(A1) and GB 2586046, which are incorporated herein by reference. These approaches include aligning the acquired spectrum to a reference spectrum using identified peaks in the acquired spectrum. To achieve high dispersion of light, echelle dispersive gratings can be used in optical emission spectroscopy. When used in conjunction with a second dispersive element, such as a prism or another grating, a high-resolution 2D spectrum known as an echelogram, also referred to as a full-frame or echelle spectrum, can be acquired.

[0003] The method may include obtaining a spectrum recorded using the spectrometer and one or more respective condition parameters indicative of operating conditions of the optical emission spectrometer or an environment of the optical emission spectrometer at and / or prior to recording the spectrum using the spectrometer. For example, in some embodiments, the one or more condition parameters may include a parameter indicative of temperature. The one or more condition parameters indicative of temperature may include, for example, at least one temperature measurement obtained from a temperature sensor attached to a mechanical structure of the optical emission spectrometer. In some embodiments, the one or more condition parameters may include a parameter correlated with or indicative of temperature. The one or more condition parameters may include one or more of: a parameter indicative of a heating current applied to a heating device to heat the optics and / or stabilize the temperature of the optics; a parameter indicative of a temperature of the environment of the spectrometer; a parameter indicative of a temperature of the optics; a parameter indicative of at least one temperature measurement obtained from a temperature sensor attached to a mechanical structure supporting one or more optical components of the optics; a parameter indicative of a radio frequency (RF) power of an RF generator to generate a plasma for use in acquiring the optical emission spectrum; a parameter indicative of a temperature of a power supply of the RF generator; a parameter indicative of a voltage of a power supply of the RF generator; a parameter indicative of a current of a power supply of the RF generator; a parameter indicative of a voltage applied to a heating pad attached to the optics; a parameter indicative of a current in a control board of the scientific instrument; a parameter indicative of a current in a camera printed circuit board (PCB); a parameter indicative of a current in a PCB of the optics; a parameter indicative of a temperature of one or more components in the PCB; a parameter indicative of whether an additional gas option is set; In some embodiments, the one or more condition parameters may include a condition change parameter that indicates a change in the condition, or a direction of the change in the condition may be provided as part of the input data for the model.

[0004]

[0013] Embodiments will be readily understood by the following detailed description taken in conjunction with the accompanying drawings, in which:

[0014] To facilitate this description, like reference numerals refer to like structural elements;

[0015] Embodiments are illustrated in the figures of the accompanying drawings, by way of example, and not by way of limitation, in which: [Brief explanation of the drawings]

[0005] [Figure 1] FIG. 10 is a diagram of an exemplary inductively coupled plasma optical emission spectrometer 1000 for obtaining optical emission spectra, according to some embodiments. [Figure 2] FIG. 1 is a block diagram of an exemplary scientific instrument support module for performing support operations, according to various embodiments. [Figure 3] 1 is an exemplary full-frame spectrum of the intensity distribution captured by the photodetector. [Figure 4] 10 is an example of an estimated warp profile for aligning a shifted spectrum to a reference spectrum according to some embodiments. [Figure 5] FIG. 1 is a flow diagram of an exemplary method for mitigating distortions in an optical emission spectrum, according to various embodiments. [Figure 6] FIG. 1 is a flow diagram of an exemplary method for obtaining training data, according to various embodiments. [Figure 7] FIG. 1 is a flow diagram of an exemplary method for training a machine learning model, according to various embodiments. [Figure 8] 1 illustrates exemplary input data for a machine learning model and exemplary output data for training and providing the model, according to various embodiments. [Figure 9A] 1 shows the prediction of lateral drift by a machine learning model in an exemplary embodiment. Left: actual drift (gray) and predicted drift (black). Right: prediction error (difference). [Figure 9B] 1 shows the prediction of vertical drift by a machine learning model in an exemplary implementation. Left: actual drift (gray) and predicted drift (black). Right: prediction error (difference). [Figure 9C]1 shows a comparison between actual and predicted lateral drift in an exploratory test of a machine learning model trained using only one condition parameter. [Figure 10] 1 is an example of a graphical user interface that can be used in implementing some or all of the assistance methods disclosed herein, according to various embodiments. [Figure 11] FIG. 1 is a block diagram of an exemplary computing device capable of implementing some or all of the scientific instrumentation methods disclosed herein, according to various embodiments. [Figure 12] FIG. 1 is a block diagram of an exemplary scientific instrument support system capable of implementing some or all of the scientific instrument support methods disclosed herein, according to various embodiments. DETAILED DESCRIPTION OF THE INVENTION

[0006] Disclosed herein are scientific instrument support systems, as well as related methods, computing devices, and computer-readable media. In some aspects, a method is implemented for reducing distortions in an optical emission spectrum obtained from an optical emission spectrometer. The optical emission spectrometer comprises optics for forming a spectrum; for example, the optical emission spectrometer may be an inductively coupled plasma optical emission spectrometer that can be configured to record an echelle emission spectrum of light emitted from a plasma. In some embodiments, the spectrometer may be a Raman spectrometer configured to record a spectrum of inelastically scattered photons. In some embodiments, the spectrometer may be an infrared spectrometer configured to record a spectrum of the interaction of infrared radiation interacting with an analyte. For the avoidance of doubt, references to "light" and "optical" will be understood to refer to any portion of the electromagnetic spectrum, for example, any one or more of the ultraviolet, visible, or infrared regions of the electromagnetic spectrum.

[0007] The method may include obtaining a spectrum recorded using the spectrometer and one or more respective condition parameters indicative of operating conditions of the optical emission spectrometer or an environment of the optical emission spectrometer at and / or prior to the time of recording the spectrum using the spectrometer. For example, in some embodiments, the one or more condition parameters may include a parameter indicative of temperature. The one or more condition parameters indicative of temperature may include at least one temperature measurement obtained, for example, from a temperature sensor attached to a mechanical structure of the optical emission spectrometer. In some embodiments, the one or more condition parameters may include a parameter correlated with or indicative of temperature. The one or more condition parameters may include one or more of: a parameter indicative of a heating current applied to a heating device to heat the optics and / or stabilize the temperature of the optics; a parameter indicative of a temperature of the environment of the spectrometer; a parameter indicative of a temperature of the optics; a parameter indicative of at least one temperature measurement obtained from a temperature sensor attached to a mechanical structure supporting one or more optical components of the optics; a parameter indicative of a radio frequency (RF) power of an RF generator to generate a plasma for use in acquiring the optical emission spectrum; a parameter indicative of a temperature of a power supply of the RF generator; a parameter indicative of a voltage of a power supply of the RF generator; a parameter indicative of a current of a power supply of the RF generator; a parameter indicative of a voltage applied to a heating pad attached to the optics; a parameter indicative of a current in a control board of the scientific instrument; a parameter indicative of a current in a camera printed circuit board (PCB); a parameter indicative of a current in a PCB of the optics; a parameter indicative of a temperature of one or more components in the PCB; a parameter indicative of whether an additional gas option is set; In some embodiments, the one or more condition parameters may include a condition change parameter that indicates a change in the condition, or a direction of the change in the condition may be provided as part of the input data for the model.

[0008] It will be understood that when referring to condition parameters indicative of the operating conditions at the time of recording, no particular timing precision is implied. That is, the condition parameters obtained at the time of recording affect or include information about the state of the spectrometer at the time of recording, e.g., information that affects or includes the mechanical configuration of optical components as affected by environmental and optics temperatures, which affect distortions in the recorded spectrum. Thus, such a time "at the time of recording" may somewhat precede or follow the actual exact time of recording the spectrum, so long as the condition parameters at such time affect or include information about distortions of the recorded spectrum relative to baseline conditions.

[0009] The method may further include providing a machine learning model configured to output, in response to the one or more condition parameters, one or more transformation parameters of a transformation to be applied to the acquired spectrum to mitigate distortions in the spectrum due to a mismatch between the operating conditions and the baseline operating conditions. For example, in some embodiments, the one or more transformation parameters of the transformation may define all or a portion of a transformation to be applied to the acquired spectrum. For example, in some embodiments, the transformation includes at least one of a translation, a rotation, a scaling operation, a centering operation, a normalization operation, a twisting operation, a stretching operation, and a deformation field. The one or more condition parameters may be applied as input to the machine learning model, and the one or more transformation parameters may be obtained as an output of the machine learning model. Such embodiments allow distortions to be mitigated over a wide range of operating conditions, resulting in useful analytical results being obtained over a wider range of operating conditions, compared to conventional approaches that require a spectrometer to operate within a constrained range of stable operating conditions to obtain useful analytical results. Typically, reaching this constrained range of stable operating conditions requires the spectrometer to operate for a significant period of time to allow operating conditions to stabilize before performing any analysis, wasting energy, time, and expensive consumables such as argon gas used in plasma generation. By enabling the use of the spectrometer over a wider range of varying operating conditions, this wasted "start-up" period can be reduced or eliminated. Furthermore, while varying operating conditions traditionally may require analytical results to be "thrown away" (e.g., due to changes in the temperature of the spectrometer chamber), various of the embodiments disclosed herein allow valid and analytically useful spectra to be acquired during these varying conditions, improving throughput and instrument availability. Additionally, because the transformation parameters are estimated based on the operating conditions, the described methods avoid the need to adapt the transformation for each acquired spectrum, thereby enabling distortion mitigation in a computationally efficient manner.

[0010] In some embodiments, the input data for the machine learning model may include a time series of one or more condition parameters at each of a plurality of time points. For example, the time series may have a 15-minute time slot. It is understood that the number of optical emission spectra acquired throughout the time slot of the time series may vary, and any suitable number of optical emission spectra may be acquired. For example, in some embodiments, an optical emission spectrum may be acquired at each of a plurality of time points. In an example where the time series is 15 minutes, an optical emission spectrum may be acquired every minute, and one or more condition parameters may be acquired every minute or more frequently, and an average, standard deviation, mean, and slope may be provided for each time step. In some embodiments, the optical emission spectrum may be acquired only at the last time point or at any time point. In some embodiments, an optical emission spectrum is not acquired within the time slot of the time series; instead, an optical emission spectrum is acquired at a predetermined time after the last time point of the time series. A time slot of the time series of one or more condition parameters may define any period of time that includes or is earlier than the last of the one or more time points at which one or more optical emission spectra are acquired. In those embodiments in which the input data includes a time series, it is understood that the output data may or may not include a time series of transformation parameters. In some embodiments, the time period includes time points up to and including the time of acquiring an optical emission spectrum. For example, in some embodiments, the input data includes a time series of one or more condition parameters at each of a plurality of time points from a predetermined amount of time (e.g., 15 minutes) before acquiring the optical emission spectrum up to and including the time of acquiring the optical emission spectrum. In these embodiments, the output data may include one or more transformation parameters for a transformation applied to the optical emission spectrum. In some embodiments, the input data includes a time series of one or more condition parameters over a predetermined time period (e.g., 15 minutes), with an optical emission spectrum being acquired at each time point, or at a subset of the time points. In some embodiments, the time series may be used to predict the transformation parameters as output at a single time point.In such embodiments, the one or more condition parameters may further include one or more of the mean, standard deviation, maximum-minimum difference, or slope of change for the condition parameter for each time interval prior to the time of image recording. For example, [0-1 min, 1-2 min, ..., 14-15 min]. In some embodiments, the output data may include a time series of one or more transformation parameters at each or a subset of the multiple time points. In these embodiments, the method may include applying a transformation to the optical emission spectrum acquired at each time point according to the one or more transformation parameters at that time point of the acquired instance of the output data. A transformation according to each transformation parameter may be applied to each spectrum at each time point in the time series.

[0011] Because the dependence of the spectral distortion on operating conditions may exhibit some degree of hysteresis or history dependency, using a time series of operating conditions as input can improve the accuracy of the predicted transformation parameters. This is because the history dependency can be accounted for by the model compared to models that use only instantaneous operating conditions, where time information is lost. This is true whether the model output includes the transformation parameters at a single point in time or at multiple points in time. In some embodiments, the element of history dependency can be accounted for in the input data by including parameters indicative of changes or direction of change in the operating conditions instead of or in addition to using a time series as input.

[0012] In some embodiments, the spectrum may be an image, and the transformation may include operations that vary across multiple locations within the image; for example, in some embodiments, the operations may include applying a distortion field (also known as a warp field) to the image to align the image. As a result, a wide range of local distortions can be captured. Compared to global affine transformations such as translation, rotation, and scaling, such local non-affine transformations can capture and mitigate a wider range of spectral distortions, thus enabling more accurate identification of peaks and corresponding analytes. Specifically, the local transformations can capture functional groups in different portions of a spectrum that are distorted differently, thereby increasing the likelihood of accurate identification. While examples related to spectra represented as digital images are used throughout, it should be understood that a spectrum can be represented in any suitable manner, for example, in terms of a set of identified intensity peaks or troughs and their respective locations, e.g., the two-dimensional coordinates of each of the intensity peaks in the case of an eschelogram or other two-dimensional spectrum.

[0013] The machine learning model may be any suitable machine learning model. For example, a random forest or a gradient boosting decision tree (GBDT) learning algorithm, both of which include multiple decision trees, use "bagging" and "boosting" techniques, respectively. In one embodiment, the XGBoost algorithm may be used. The XGBoost algorithm is a well-known decision tree-based ensemble machine learning algorithm that, like GBDT, uses gradient boosting. XGBoost is described in the paper "Chen, T. and Guestrin, C., 2016, August. XGBoost: A scalable tree boosting system." in Proceedings of the 22nd ACM SIGCDD International Conference on Knowledge Discovery and Data Mining (pp. 785-794), which is incorporated herein by reference. A machine learning model using the XGBoost algorithm may be trained to provide one or more transformation parameters for a transformation applied to acquired spectra in response to a time series of one or more condition parameters.

[0014] In some embodiments, the machine learning model includes a feedforward neural network. In embodiments where the input data includes a time series, one or more condition parameters at each of the multiple time points can be applied together as inputs to the feedforward neural network. In some embodiments, the machine learning model can include a recurrent neural network, such as an LSTM recurrent neural network. In some embodiments, the machine learning model can include a Transformer model.

[0015] In one aspect of the present disclosure, a method for training a machine learning model includes obtaining a training dataset including training data pairs. Each training data pair includes values ​​of one or more condition parameters for an environment of the optical emission spectrometer that differ from a respective operating condition or a baseline operating condition of the optical emission spectrometer, and respective values ​​of one or more transformation parameters. It will be understood that the values ​​of the condition parameters and / or transformation parameters may be scalar or non-scalar values ​​(e.g., sets, vectors, or other multidimensional values). The method includes adjusting parameters of the machine learning model to reduce discrepancies between the values ​​of the one or more transformation parameters of the training data pairs and values ​​of the one or more transformation parameters output by the machine learning model in response to the respective values ​​of the one or more condition parameters of the training data pairs. For example, the one or more transformation parameters may be lateral (X-direction) and vertical (Y-direction) transformations, and the model may be trained to minimize the mean squared error between the values ​​of the lateral (X-direction) and vertical (Y-direction) transformations of the training data pairs and the values ​​of the lateral (X-direction) and vertical (Y-direction) transformations output by the model. In one embodiment, the model trained to minimize mean squared error may use the XGBoost architecture.

[0016] In some embodiments of the present disclosure, a method for obtaining training data for training a machine learning model includes recording a plurality of optical emission spectra of a reference analyte using an optical emission spectrometer for each operating condition and storing input data for the machine learning model. The input data includes, for each recorded optical emission spectrum, one or more parameters indicative of the respective operating conditions at the time the spectrum was recorded and / or at a preceding time. For each recorded optical emission spectrum, one or more transformation parameters for a transformation to align the optical emission spectrum to a baseline optical emission spectrum using a transformation are generated by adjusting the parameters to align the optical emission spectrum to the baseline optical emission spectrum. The baseline optical emission spectrum is recorded for a baseline set of operating conditions, for example, operating conditions recommended by the spectrometer manufacturer for capturing the data. The method includes storing output data for the machine learning model including, for each optical emission spectrum, the respective adjusted one or more transformation parameters as training targets, in association with the respective input data for each optical emission spectrum as a training data pair. It will be understood that the training data may be generated using the same or a different spectrometer than the one used to record the baseline spectrum. Similarly, the training data can include spectra recorded using different respective spectrometers, which can facilitate generalization of the model across different spectrometers. Preferably, the spectrometer used to generate the training data is of the same type, e.g., the same make and / or model, as the spectrometer whose spectra will be adjusted by the trained model; however, in some embodiments, the spectrometer used to generate the training data is of a different make and / or model than the spectrometer whose spectra will be adjusted by the trained model. In some embodiments, the same spectrometer is used to generate the training data. This can facilitate generating a model specific to that spectrometer, which can result in greater accuracy for that spectrometer.

[0017] The scientific instrumentation support embodiments disclosed herein may achieve improved performance compared to conventional approaches. For example, distortion of optical spectra due to temperature fluctuations is a common problem, but conventional methods use peaks, e.g., carbon peaks, that appear in both the reference spectrum and the sample spectrum to calculate the distortion from the peak's expected location. The identified location of an unknown peak in the same sample spectrum can then be shifted using the determined drift. In some embodiments of the present disclosure, the spectrum is an image, and adjusting one or more transformation parameters to align the spectrum may include comparing intensities between images. Such embodiments do not require peak identification. Because peaks do not need to be precisely identified, the resulting method may be more flexible and accurate by considering the entire image, as opposed to only identified peaks. Therefore, the embodiments disclosed herein provide improvements to scientific instrumentation technology (e.g., improvements to the computer technology supporting such scientific instruments, among other things).

[0018] Aspects and embodiments disclosed herein can achieve more flexible transformations by capturing a wider range of distortions than conventional approaches. For example, conventional approaches rely on identifying peaks that appear in both the reference spectrum and the sample spectrum, and therefore become difficult to implement when none of the peaks have clearly identifiable locations. In addition, these approaches suffer from several technical problems and limitations that arise due to the offsets that are applied linearly across the spectrum, which is particularly relevant for echelon or full-frame spectra.

[0019] Various aspects and embodiments disclosed herein, such as an embodiment in which adjusting transformation parameters in a method for providing training data for a machine learning model may include comparing intensities between two images, can improve upon conventional approaches to achieve the technical advantage of reducing distortion without the need to find peaks. Such technical advantages are unattainable with conventional stylized approaches, and all users of systems incorporating such embodiments can benefit from these advantages (e.g., by assisting users in performing technical tasks, such as broadening the temperature range over which spectra can be acquired because no discernible peaks are required). Thus, the technical features of the embodiments disclosed herein, as well as combinations of features of the embodiments disclosed herein, are clearly unconventional in the field of spectroscopy. The computational and user interface features disclosed herein not only involve the collection and comparison of information, but also apply novel analytical and technical techniques to modify the operation of machine learning models used in this field. Thus, the present disclosure introduces functionality that neither conventional computing devices nor humans could perform.

[0020] Accordingly, embodiments of the present disclosure may achieve any of several technical objectives. These technical objectives include controlling a particular technical system or process, determining how to control machinery from measurements, enhancing analysis, reducing the amount of sensor data to be processed, or providing faster processing of sensor data, the latter two of which result at least in part from the elimination of the requirement for peak identification to mitigate spectral distortion. In particular, the present disclosure provides technical solutions to technical problems, including, but not limited to, providing a more flexible transform that captures a wide range of local distortions, and removing the requirement for identifying spectral peaks when analyzing a spectrum, thereby reducing computational requirements during inference. Further technical solutions include providing an instrument that can be used under virtually any operating conditions (e.g., temperature), thus reducing the amount of expensive argon gas spent on plasmas because useful spectra can be acquired before the operating conditions stabilize at baseline operating conditions. Thus, embodiments disclosed herein provide improvements to analytical instrument technology (e.g., improvements in computer technology supporting chemical analysis, among other improvements).

[0021] In the following detailed description, reference is made to the accompanying drawings that form a part hereof, where like numerals refer to like parts throughout and which show, by way of illustration, embodiments that may be practiced. It is to be understood that other embodiments may be utilized and structural or logical changes may be made without departing from the scope of the present disclosure. Therefore, the following detailed description is not to be taken in a limiting sense.

[0022] Various operations may be described as multiple separate actions or operations, in the order most helpful in understanding the subject matter disclosed herein. However, the order of description should not be construed as implying that these operations are necessarily order dependent. In particular, these operations may not be performed in the order presented. The operations described may be performed in a different order than in the described embodiment. Various additional operations may be performed and / or described operations may be omitted in additional embodiments.

[0023] For purposes of this disclosure, the phrases “A and / or B” and “A or B” mean (A), (B), or (A and B). For purposes of this disclosure, the phrases “A, B, and / or C” and “A, B, or C” mean (A), (B), (C), (A and B), (A and C), (B and C), or (A, B, and C). Although some elements may be referred to in the singular (e.g., “processing device”), any suitable element may be represented by multiple instances of that element, and vice versa. For example, a set of operations described as being performed by a processing device may be implemented with different ones of the operations performed by different processing devices.

[0024] This description uses the phrases "one embodiment," "various embodiments," and "some embodiments," each of which may refer to one or more of the same or different embodiments. Furthermore, terms such as "comprising," "including," and "having," when used with respect to embodiments of the present disclosure, are synonymous. When used to describe a range of dimensions, the phrase "between X and Y" represents a range that includes X and Y. As used herein, an "apparatus" may refer to any individual device, a collection of devices, a portion of a device, or a collection of portions of devices. The drawings are not necessarily drawn to scale.

[0025] FIG. 1 is a diagram of an example of an inductively coupled plasma optical emission spectrometer (ICP-OES) 1000 for acquiring optical emission spectra, according to various embodiments. The ICP-OES 1000 may be a component in a scientific support system 1200, for example, as described below with reference to FIG. 12. Emission spectrometers operate by exciting atoms and ions to emit electromagnetic (EM) radiation at wavelengths characteristic of particular elements. Electron transitions from higher to lower energy states emit a frequency spectrum of EM radiation. The ICP-OES 1000 may include a plasma chamber 1002. The plasma chamber 1002 may be connected to a radio frequency (RF) source and a gas source, such as argon gas. The argon gas may be ionized within an oscillating RF field generated by the RF source to create and sustain a plasma within the plasma chamber 1002. The ICP-OES 1000 may be of an echelle-based optical design for generating two-dimensional optical emission spectra.

[0026] Specifically, the ICP-OES can include an echelle diffraction grating 1014, a prism 1012, and multiple focusing mirrors 1004, 1006, 1010, and 1016. Collectively, these components provide an optical system 1024. Light from the plasma chamber 1002 enters the ICP-OES 1000 and is selectively focused using multiple focusing mirrors, e.g., a first mirror 1004 and a second mirror 1006. The focused light can pass through an entrance slit 1008 using mirror 1010 and enter a prism 1012. The prism 1012 can separate the light by wavelength. The echelle diffraction grating 1014 can diffract the separated light from the prism 1012 into multiple diffraction orders to produce a high-resolution 2D spectrum known as an echelogram, also called a full-frame or echelle spectrum. After passing through these optical elements, a mirror 1016 can collect and focus the spectrum onto a detector, such as a camera 1018. The optical system is contained within a housing 1026. One or more heating pads 1022 can be located on the exterior of the housing 1026 to heat the optical system 1024 and / or its environment within the housing 1026, thereby controlling the temperature of the optical system 1024 and keeping it stable at a baseline operating temperature. The temperature sensor 1020 can indicate, for example, the temperature of the optical emission spectrometer's environment. In one embodiment, the temperature sensor 1020 can indicate the temperature of the optical emission spectrometer or the optical system of an optical emission spectrometer that includes optical components. Advantageously, the temperature sensor 1020 can be disposed on a mechanical structure that supports one or more of the optical components to better capture temperature fluctuations that affect the configuration of the optical components and, therefore, the matching of the spectrum. A second temperature sensor can be located away from the optical system 1024, but in some embodiments, can still be located inside the device. The second temperature sensor can provide an ambient temperature measurement. It will be understood that the number, identities, and locations of components described above with reference to FIG. 1 are used by way of example, and other configurations and / or optical components may be used.Although embodiments are described with reference to an inductively coupled plasma optical emission spectrometer, they are equally applicable to other types of spectra acquired using other types of spectrometers. For example, the methods described herein can be used for similar spectroscopic techniques such as inductively coupled plasma mass spectrometry (ICP-MS) and atomic absorption spectrometry (AAS), and indeed can be applied to any method of acquiring a spectrum in which the spectrum depends on the operating conditions of the spectrometer. Various embodiments include the use of spectroscopic techniques such as Rayleigh, Raman, or mass spectrometry.

[0027] FIG. 2 is a block diagram of a scientific instrument support module 2000 for performing support operations for a spectrometer, such as those described with reference to FIG. 2, in accordance with various embodiments. The scientific instrument support module 2000 may be implemented by a circuit (e.g., including electrical and / or optical components) such as a programmed computing device. The logic of the scientific instrument support module 2000 may be contained in a single computing device or may be distributed across multiple computing devices that communicate with each other as needed. An example of a computing device that may implement the scientific instrument support module 2000, alone or in combination, is discussed herein with reference to the computing device 11000 of FIG. 11, and an example of a system of interconnected computing devices in which the scientific instrument support module 2000 may be implemented across one or more of the computing devices is discussed herein with reference to the scientific instrument support system 12000 of FIG. 12.

[0028] The scientific instrument support module 2000 may include data acquisition logic 2002, image registration logic 2004, training logic 2006, and inference logic 2008. As used herein, the term "logic" may include an apparatus that performs a set of operations associated with the logic. For example, any of the logic elements included in the support module 2000 may be implemented by one or more computing devices programmed with instructions that cause one or more processing devices of the computing devices to perform the associated set of operations. In particular embodiments, a logic element may include one or more non-transitory computer-readable media having instructions that, when executed by one or more processing devices of the one or more computing devices, cause the one or more computing devices to perform the associated set of operations. As used herein, the term "module" may refer to a collection of one or more logic elements that together perform a function associated with the module. Different logic elements within a module may take the same form or different forms. For example, some logic in a module may be implemented by a programmed general-purpose processing device, while other logic in a module may be implemented by an application-specific integrated circuit (ASIC). In another example, different ones of the logic elements in a module may be associated with different sets of instructions executed by one or more processing devices. A module may not include all of the logic elements depicted in an associated figure; for example, a module may include a subset of the logic elements depicted in an associated figure when that module performs a subset of the operations discussed herein with reference to that module. While particular configurations of logic elements and modules are described, it will be understood that the corresponding functionality may be implemented and divided in many other ways.

[0029] Data acquisition logic 2002 may be configured to control ICP-OES 1000 to acquire data for use by, for example, training logic 2006 or inference logic 2008. Data acquisition logic 2002 may be configured to control ICP-OES 1000 to change the operating conditions of ICP-OES 1000, for example, by controlling heating pad 1022.

[0030] For example, the data acquisition logic 2002 may control the operating conditions of the ICP-OES to correspond to specific predetermined baseline operating conditions for acquiring a baseline spectrum. For example, to acquire a baseline spectrum, the data acquisition logic 2002 may be configured to maintain the operating conditions of the optics at a stable temperature of 38 degrees Celsius for a predetermined period of time. For example, in one embodiment, the data acquisition logic 2002 may be configured to maintain the operating conditions of the optics at a stable temperature for 30 minutes before acquiring the baseline spectrum. In some embodiments, multiple spectra may be acquired within a predetermined period of time, for example, at a rate of one spectrum per second. The data acquisition logic may compare one or more of the spectra acquired at the baseline operating conditions to confirm the absence of drift or a negligible amount of drift (e.g., 0.5 pixels in 30 minutes) in the baseline operating conditions. If either the absence of drift or the negligible amount of drift is confirmed, the data acquisition logic 2002 may use any one of the spectra acquired at the baseline operating conditions as the baseline spectrum. The data acquisition logic 2002 can be configured to change the operating conditions to a number of specific predetermined operating conditions to acquire training data, as described in more detail below with reference to FIG. 6 . For example, the data acquisition logic 2002 may acquire spectra periodically, e.g., one spectrum per second, while changing the operating conditions, e.g., by cycling through a number of temperature settings using the heating pad 1022. The data acquisition logic can simultaneously or sequentially change the RF power supplied to the plasma chamber 1002 to change the operating conditions of the optical system, which can include setting the RF power to alternate between a high setting, e.g., 1500 W, and a low setting, e.g., 1150 W, over a predetermined period of time, e.g., two minutes. Sequential cycling of both the heating pad 1022 and the temperature supplied by the RF generator can enhance thermal regulation of the optical system of the ICP-OES 1000, since plasma heat is one or the primary perturber of the thermal balance of the optical system.

[0031] The data acquisition logic 2002 can be configured to acquire and store spectral data and corresponding operating condition data. Figure 3 shows an exemplary full-frame spectrum or eschelogram of an intensity distribution captured by the photodetector 1018 on the ICP-OES 1000, according to some embodiments. As shown in Figure 3, multiple peaks are seen as white dots of varying intensity. Each peak of the multiple peaks may have a characteristic intensity and wavelength indicative of a particular atom or molecule. The intensity distribution captured by the photodetector 1018 may be captured and stored by the data acquisition logic 2002 in the form of a digital image including a collection of pixels with pixel values ​​reflecting the captured intensity distribution. The pixel values ​​may be grayscale scalar or vector values ​​representing color information.

[0032] The data acquisition logic 2002 can be configured to perform checks upon initialization of the ICP-OES 1000 and can be further configured to apply known background correction techniques before acquiring a spectrum. The data acquisition logic 2002 can perform standard sample preparation related operations, standard data quality control operations, and any standard data acquisition operations, and can further control sample introduction to the spectrometer and exposure time of the analyte to the plasma according to the concentration of the analyte in accordance with standard operating procedures.

[0033] The image registration logic 2004 can be configured to register the acquired spectral image, also referred to as the spectral image, to generate transformation parameters for a transformation that aligns the acquired spectral image to the baseline spectral image. The image registration logic 2004 can be configured to retrieve the baseline spectral image stored by the data acquisition logic 2002 and the acquired spectral image stored by the data acquisition logic 2002 and perform the image registration. In some embodiments, the image registration logic 2004 can be configured to perform the image registration based on identified local points in the spectrum as landmarks, such as spectral peaks. In these embodiments, the transformation parameters are adjusted to minimize positional errors between landmarks in the acquired spectrum and the baseline spectrum when the corresponding transformation is applied to the acquired image. In other embodiments, the image registration logic 2004 can be configured to perform image registration of an entire spectral image, such as the full-frame spectrum shown in FIG. 3, for example, based on comparing image intensities for every pixel or a set of control points between the acquired spectrum and the baseline spectrum.

[0034] Thus, the image registration logic 2004 can determine transformation parameters for a transformation to align the acquired optical emission spectrum to the baseline optical emission spectrum. This transformation can be an affine transformation, such as a combination of translation and rotation operations applied globally to the acquired image, or the transformation can be non-affine, applying a local transformation, for example, by distortion or warp field, as is well known in the art of image registration. The image registration logic 2004 can perform the image registration using any suitable known method. For example, in one embodiment, the image alignment logic 2004 may include an adversarial similarity network [Adversarial Similarity Network for Evaluating Image Alignment in Deep Learning Based Registration Jingfan Fan, Xiaohuan Cao, Zhong Xue, Pew-Thian Yap, and Dinggang Shen, Med Image Comput Comput Assist Interv. 2018 Sep;11070:739-746. doi:10.1007 / 978-3-030-00928-1_83. Epub 2018 Sep 26. PMID:30627709; PMCID:PMC6322551, incorporated herein by reference], which is known for use in estimating image warping / shifting for medical purposes.

[0035] An example warp field for aligning an acquired spectrum to a baseline spectrum is shown in Figure 4. This warp field defines a local transformation, indicated by the arrows, that aligns the local intensity distributions in each image, and thus illustrates not only the specific type of transformation used in some embodiments, but also distortions that may arise due to varying operating conditions. Typically, such distortions arise due to various operating conditions that may be indicative of or correlated with the temperature of the optical system.

[0036] The training logic 2006 can be configured to perform a training process 7000 to use the training data acquired by the data acquisition logic 2004 to generate a machine learning model that generates transformation parameters in response to input of parameters describing the operating conditions of the optical emission spectrometer 1002, as described in more detail below with reference to FIG. 7.

[0037] The inference logic 2008 may be configured to perform an inference process 5000, as described in detail below with reference to FIG. 5, to generate transformation parameters and apply the corresponding transformation to the acquired spectrum to mitigate operating condition-related distortions in the acquired spectrum.

[0038] FIG. 5 is a flow diagram of an inference method 5000 for mitigating distortions in an optical emission spectrum, according to various embodiments. The steps of this process may be performed, for example, by inference logic 2008. Although the operations of method 5000 may be illustrated with reference to particular embodiments disclosed herein (e.g., scientific instrument support module 2000 discussed herein with reference to FIG. 2 , graphical user interface (GUI) 10000 discussed herein with reference to FIG. 10 , computing device 11000 discussed herein with reference to FIG. 11 , and / or scientific instrument support system 12000 discussed herein with reference to FIG. 12 ), method 5000 may be used in any suitable configuration to perform any suitable support operations. Although the operations are illustrated in FIG. 5 once each and in a particular order, the operations may be reordered and / or repeated as necessary or appropriate (e.g., operations may be performed in parallel, if suitable).

[0039] In 5002, a first operation can be performed. The first operation can include acquiring an optical emission spectrum using a spectrometer including an optical system for forming the optical emission spectrum. The spectrometer can be the ICP-OES 1000 described above with reference to FIG. 1. The spectrum can be, for example, a full frame as shown in FIG. 3. The spectrum can be acquired while the spectrometer is operating at stable operating conditions, e.g., baseline operating conditions, or at other times. For example, the disclosed method allows a spectrum to be usefully acquired while the spectrometer is warming up to reach baseline operating conditions. This is because the disclosed method allows for correction or at least mitigation of operating condition-related distortions, such as thermal distortions, while the spectrometer is reaching baseline operating conditions following switch-on. As a result, the spectrum can be used more quickly, reducing consumption of electricity and other consumables.

[0040] In 5004, a second operation may be performed. The second operation may include acquiring one or more condition parameters indicative of the operating conditions of the optical emission spectrometer or the environment of the optical emission spectrometer at and / or prior to the time of recording the optical emission spectrum. For example, in some embodiments, the one or more condition parameters may be indicative of the operating conditions of the optical system, which may be indicative of or correlated with the temperature of the optical system. The condition parameters may include one or more of an optical element temperature, an RF voltage, an RF current, an optical element heater voltage, and an ambient temperature substrate readback. In some embodiments, the one or more condition parameters may be a time series of the one or more condition parameters at each of a plurality of time points, for example, a 15-minute period leading up to and including the time of acquiring the optical emission spectrum. In some embodiments, when two or more condition parameters are acquired, the condition parameters may be acquired over different time periods. For example, the input data may include a time series of one or more condition parameters at each of a plurality of time points in a 15-minute period leading up to and including the time of acquiring the optical emission spectrum, and a time series of one or more other condition parameters at each of a plurality of time points in a 5-minute period leading up to and including the time of acquiring the optical emission spectrum.

[0041] At 5006, a third operation may be performed. The third operation may include providing the obtained one or more condition parameters as input to a machine learning model. For example, in some embodiments, a machine learning model using the XGBoost algorithm may be used. In some embodiments, the machine learning model includes a feedforward neural network. In embodiments in which the input data includes a time series, the one or more condition parameters at each of the multiple time points may be applied together as input to the feedforward neural network. In some embodiments, the machine learning model may include, for example, a recurrent neural network including an LSTM unit. In some embodiments, the machine learning model may include a Transformer model. The machine learning model may have been trained, for example, according to the process described below with reference to FIG. 7.

[0042] A fourth operation may be performed at 5008. The fourth operation may be obtaining one or more transformation parameters as an output of the machine learning model.

[0043] A fifth operation may be performed at 5010. The fifth operation may be applying a transform to the acquired optical emission spectrum according to one or more transform parameters to reduce distortion in the optical emission spectrum. As described above, reducing distortion due to varying operating conditions may allow useful spectra to be collected more quickly without having to wait for operating conditions to stabilize.

[0044] FIG. 6 is a flow diagram of an exemplary data collection method 6000 for acquiring training data for training a machine learning model used in the inference method 5000, according to various embodiments. Unlike traditional operating modes in which a system is left within a constrained range of stable operating conditions before data collection, the training data is acquired over a wide range of operating conditions. The steps of this process may be performed, for example, by data acquisition logic 2002 and image registration logic 2004. While the operation of data collection method 6000 may be illustrated with reference to particular embodiments disclosed herein (e.g., scientific instrument support module 2000 discussed herein with reference to FIG. 2 , GUI 10000 discussed herein with reference to FIG. 10 , computing device 11000 discussed herein with reference to FIG. 11 , and / or scientific instrument support system 12000 discussed herein with reference to FIG. 12 ), method 6000 may be used in any suitable setting to perform any suitable support operations. Although the operations are illustrated in FIG. 6 once each and in a particular order, the operations may be reordered and / or repeated as necessary or appropriate (e.g., operations may be performed in parallel where suitable).

[0045] In 6002, a first operation may be performed. The first operation may include providing a spectrometer with optics for forming an optical emission spectrum of a reference analyte. The spectrometer may be an ICP-OES 1000 as described above with reference to FIG. 1.

[0046] At 6004, a second operation may be performed. The second operation may include recording an optical emission spectrum of a reference analyte for the operating conditions of the optical emission spectrometer or the environment of the optical emission spectrometer. For example, in one embodiment, the second operation may include recording the operating conditions of the optical system of the optical emission spectrometer.

[0047] A third operation may be performed at 6006. The third operation may include storing one or more parameters indicative of the operating conditions of the optical system under which the optical emission spectrum was recorded as input data for the machine learning model.

[0048] At 6008, a fourth operation may be performed. The fourth operation may include adjusting one or more transformation parameters of the transformation to align the optical emission spectrum to a baseline optical emission spectrum using the transformation, as described above with reference to the image alignment logic 2004. The baseline optical emission spectrum may be recorded for a baseline operating condition, e.g., by the data acquisition logic 2002, when the spectrometer is in a stable baseline operating condition. For example, in some embodiments, the baseline optical emission spectrum may be recorded on the same spectrometer as the optical emission spectrum to be aligned. In some embodiments, the baseline optical emission spectrum may be an industry-standard baseline spectrum obtained from a third party.

[0049] A fifth operation may be performed at 6010. The fifth operation may include storing the adjusted one or more transformation parameters in association with respective input data for the optical emission spectra as training data pairs as output data for the machine learning model.

[0050] After 6010, method 6000 can cycle back to 6004, performing operations 6004-6010 under different respective operating conditions for each cycle, thereby recording training data pairs for the different operating conditions. For example, data acquisition logic 2002 can be used to acquire multiple spectra having multiple corresponding operating conditions. Alternatively, operations 6004-6008 can be performed as a series of cycles, each time under different respective operating conditions, and step 6010 can be subsequently performed on all spectra in a batch to form training data pairs. In either case, the cycles can all be performed on the same spectrometer, or the cycles can be performed using different spectrometers, e.g., acquiring training data pairs for multiple operating conditions from each of multiple spectrometers of the same type, e.g., the same make and model.

[0051] FIG. 7 is a flow diagram of an exemplary training method for training a machine learning model for use in inference method 5000, according to various embodiments. Any suitable machine learning model may be used and trained in accordance with various embodiments, for example, a machine learning model using the XGBoost algorithm. The steps of this process may be performed, for example, by training logic 2006. While the operations of method 5000 may be illustrated with reference to particular embodiments disclosed herein (e.g., scientific instrument support module 2000 discussed herein with reference to FIG. 2, GUI 10000 discussed herein with reference to FIG. 10, computing device 11000 discussed herein with reference to FIG. 11, and / or scientific instrument support system 12000 discussed herein with reference to FIG. 12), method 5000 may be used in any suitable setting to perform any suitable support operations. Although the operations are illustrated in FIG. 7 once each and in a particular order, the operations may be reordered and / or repeated as necessary or appropriate (e.g., operations may be performed in parallel, if suitable).

[0052] A first operation may be performed at 7002. The first operation may include, for example, acquiring a training data set having training data pairs of input data and output data, as described above with reference to FIG.

[0053] A second operation may be performed at 7004. The second operation may include applying the training inputs to the machine learning model to generate an output for the machine learning model.

[0054] At 7006, a third operation may be performed. The third operation may include adjusting parameters of the machine learning model to reduce discrepancies between the generated outputs and the training outputs of the training pairs. Operations 7002-7006 may be performed for all training data pairs in the training dataset, and the parameters may be adjusted over multiple passes or epochs of the training dataset, for example, until a stopping criterion indicating satisfactory outputs is met. For example, the stopping criterion may be a test error between the actual outputs and the target outputs, evaluated on a test dataset separate from the training dataset. In some embodiments, the test error is evaluated using n-fold, e.g., 5-fold, cross-validation.

[0055] 8 illustrates example input data 8002 provided to a machine learning model 8004 and example output data 8006 that the model is trained to provide, according to various embodiments. The input data 8002 may be acquired using, for example, data acquisition logic 2002 in some embodiments according to data acquisition method 6000. The output data 8004 may be acquired using, for example, image registration logic 2004. The number and types of parameters illustrated in FIG. 8 are purely exemplary, and any suitable number or types of parameters may be included as input data 8002 and output data 8006.

[0056] The input data 8002 indicates one or more parameters of the operating conditions of the spectrometer's optics used to record the spectrum. For example, in some embodiments, the one or more parameters may indicate or be correlated to the temperature of the optics. The input data 8002 may include, for example, a sample index indicating the corresponding full frame, an optics temperature readback indicating the temperature of the optics at the moment of acquisition, and an ambient temperature measurement acquired by a sensor located on an instrument control board separate from the optics enclosure but still within the instrument. The sample index may be used solely for data management and, in some embodiments, may not be input into the model. The input data may further include RF power (voltage), which, together with RF power current, provides a direct estimate of the power delivered to the plasma and its equivalent temperature. The plasma temperature strongly influences the temperature of the optics and, therefore, can provide further useful information for the machine learning model. The input data may further include an optics heater voltage, indicating the voltage value set by a thermal stabilization routine, which estimates the absolute value and rate of change of the optics temperature readback to control the temperature of the optics, thus providing an indication of the rate of change of the temperature of the optics. The input data can further include an acquisition time, which includes information about the rate of change of temperature along with the temperature information. In some embodiments, the acquisition time and temperature information can be used to calculate the rate of change of temperature of the optical system as a further input to the machine learning model. In some embodiments, the input data includes a time series of condition parameters, which are several samples of the condition parameters at each different acquisition time, as described below, that are applied to the input.

[0057] The machine learning model 8004 may be any suitable model trained to provide output data 8006 in response to input data 8002. To handle the input data as a time series, the machine learning model, in some embodiments, may include a recurrent neural network or a Transformer model. In some embodiments, the machine learning model may include an XGBoost algorithm. In such embodiments, the model may be trained to provide one or more transformation parameters for a transformation applied to an acquired spectrum in response to a time series of one or more condition parameters. In some embodiments, the machine learning model may include a feedforward neural network. In some embodiments, the feedforward neural network is adapted to handle time-series input data by providing an input unit for the condition parameter at each time step of the time series. In such embodiments, the sample number and / or timestamp may not be included in the input data, as the identification of each input unit indicates the corresponding time step (and parameter).

[0058] The output data 8006 indicates one or more transformation parameters of a transformation defined in terms of the transformation parameters for aligning the optical emission spectrum to the baseline optical emission spectrum, as described above. In embodiments in which the input to the machine learning model includes a time series of condition parameters, the output includes, for each spectrum at each time step, a corresponding time series of transformation parameters of the corresponding transformation.

[0059] The scientific instrument assistance methods disclosed herein can include interactions with a human user (e.g., via a user local computing device 12020, discussed below with reference to FIG. 12 ). These interactions can include providing the user with information (e.g., information about the operation of the scientific instrument, such as the scientific instrument 12010 of FIG. 12 , information about a sample being analyzed or other tests or measurements performed by the scientific instrument, information retrieved from a local or remote database, or other information) or providing options for the user to enter commands (e.g., to control the operation of the scientific instrument, such as the scientific instrument 12010 of FIG. 12 , or to control the analysis of data generated by the scientific instrument), queries (e.g., to a local or remote database), or other information. In some embodiments, these interactions may be implemented through a graphical user interface (GUI) that includes a visual display on a display device (e.g., display device 11010 discussed herein with reference to FIG. 11 ) that provides output to the user and / or prompts the user to provide input (e.g., via one or more input devices, such as a keyboard, mouse, trackpad, or touchscreen included in other I / O devices 11012 discussed herein with reference to FIG. 11 ). The scientific instrument support systems disclosed herein may include any suitable GUI for interaction with a user.

[0060] 10 illustrates an exemplary GUI 10000 that may be used in implementing some or all of the assistance methods disclosed herein, according to various embodiments. As described above, the GUI 10000 may be provided on a display device (e.g., the display device 11010 discussed herein with reference to FIG. 11 ) of a computing device (e.g., the computing device 11000 discussed herein with reference to FIG. 11 ) of a scientific instrument assistance system (e.g., the scientific instrument assistance system 12000 discussed herein with reference to FIG. 12 ), and a user may interact with the GUI 10000 using any suitable input device (e.g., any of the input devices included in the other I / O devices 11012 discussed herein with reference to FIG. 11 ) and input technique (e.g., cursor movement, motion capture, facial recognition, gesture detection, voice recognition, button activation, etc.). The GUI 10000 may include a data display area 10002, a data analysis area 10004, a scientific instrument control area 10006, and a settings area 10008. 10 is merely exemplary, and any number and arrangement of regions containing any desired characteristics may be included in GUI 10000. Data display area 10002 may display data generated by a scientific instrument (e.g., scientific instrument 12010 discussed herein with reference to FIG. 12). For example, data display area 10002 may display one or more optical emission spectra obtained from an optical emission spectrometer, e.g., as shown in FIG. 3. Data display area 10002 may further display one or more operating conditions of the optical system of spectrometer 1000. Data analysis area 10004 may display the results of data analysis (e.g., the results of analyzing the data shown in data display area 10002 and / or other data).

[0061] For example, the data analysis area 10004 can display one or more transformation parameters of a transformation to align the optical emission spectrum to a baseline optical emission spectrum using the transformation. In some embodiments, the data display area 10002 and the data analysis area 10004 can be combined in the GUI 10000 (e.g., to include data output from the scientific instrument and some analysis of the data in a common graph or area). The scientific instrument control area 10006 can include options that allow a user to control the scientific instrument (e.g., the scientific instrument 12010 discussed herein with reference to FIG. 12). For example, the scientific instrument control area 10006 can include temperature controls and RF controls. Settings area 10008 may include options that allow a user to control features and functions of GUI 10000 (and / or other GUIs) and / or perform common computing operations with respect to data display area 10002 and data analysis area 10004 (e.g., saving data on a storage device such as storage device 11004 discussed herein with reference to FIG. 11 , sending data to another user, labeling data, etc.). For example, settings area 10008 may include options for saving a spectrum as it is acquired, or options for setting a particular temperature or RF power voltage.

[0062] Exemplary Implementations In the exemplary embodiment described below, training data was acquired from two different types of optical emission spectrometers: a Thermo Scientific™ iCAP™ PRO ICP-OES model and an iCAP™ PRO X ICP-OES. For each optical emission spectrometer, a baseline spectrum of a reference analyte with the following composition was acquired at baseline operating conditions: [Al: 1 mg / L, Ba: 0.2 mg / L, Ca: 0.2 mg / L, Cu: 1 mg / L, K: 5 mg / L, Mn: 1 mg / L, Ni: 5 mg / L, Pm: 10 mg / L, Zn: 0.2 mg / L, Mg: 0.2 mg / L, others: 0.2% HNO 3 mg / L]. The baseline operating conditions were maintained for 30 minutes before acquiring the baseline spectrum. The baseline operating conditions under which the baseline spectra were acquired included an optics temperature of 38° C. and an RF power of 1150 W.

[0063] Condition parameters indicative of the operating conditions were collected at a sampling frequency of 0.1 Hz.

[0064] The condition parameters were as follows: A parameter (0, 1) that indicates whether the user pressed the button to start the plasma being generated. A parameter indicative of the temperature of an optical system (in degrees Celsius) obtained from a temperature sensor attached to the mechanical structure supporting one or more optical components of the optical system. A parameter indicating the temperature change of the optical system (in degrees Celsius). · A parameter indicating the temperature (in degrees Celsius) surrounding a scientific instrument, obtained from a temperature sensor attached to a mechanical structure separate from the optical system. A parameter indicating the temperature of the power supply of the RF generator connected to the plasma chamber 1002 (in degrees Celsius). A parameter indicating the temperature (in degrees Celsius) of the printed circuit board that controls the heating pad of the optical system. A parameter indicating the voltage of the power supply of the RF generator connected to the plasma chamber 1002 (V). A parameter indicating the current on the power supply of the RF generator connected to the plasma chamber 1002 (A). A parameter indicating the voltage applied to the optical heating pad (V). A parameter indicating the pressure in the exhaust pipe used to draw in the exhaust plasma gas (mBar). A parameter indicating the temperature of the exhaust pipe used to draw the exhaust plasma gases (in degrees Celsius). · () Parameter indicating the drain behavior (1 / s). A parameter indicating the current in the spectrometer control board (A). A parameter indicating the current in the camera's PCB (A). A parameter (A) that indicates the current in the circuit board of the optical system (including the heating pad, shutter, and slit).

[0065] Training data was obtained by modifying the operating conditions of the optical emission spectrometers while acquiring spectra. Training data was obtained from 15 optical emission spectrometers within 10 hours of each optical emission spectrometer being turned on. Test and validation data were obtained from different optical emission spectrometers to provide an unbiased assessment of model performance. Validation data was obtained from three additional optical emission spectrometers, and test data was obtained from four additional optical emission spectrometers. Spectra were obtained at high and low power conditions by setting the RF power source connected to the plasma chamber 1002 to either high or low power. This was achieved using two methods. First, by running the entire experiment at either the high power (1500 W) or low power (1150 W) setting. Second, by alternating between the 1500 W and 1150 W settings of the RF power source connected to the plasma chamber 1002. The cycling rate between RF powers varied between training samples and included, for example, alternating between high and low power every two minutes. Spectra were acquired at a sampling frequency of 1 spectrum / min. For each acquired spectrum, the lateral (X) or vertical (Y) drift from the baseline spectrum was calculated by applying a classical peak detection algorithm and recorded as one or more transformation parameters of the transformation.

[0066] The one or more transformation parameters were provided to a machine learning model. In the exemplary embodiment, an XGBoost model was used, but it is understood that any suitable machine learning model, such as a recurrent neural network or a Transformer model, may be used. The one or more transformation parameters for each transformation were provided to the XGBoost model as target variables for training the XGBoost model. For each transformation, the input variables included the corresponding time series of the condition parameters set above at each of multiple time points during the 15-minute period up to and including the time when the spectrum corresponding to the transformation was acquired. The XGBoost model was fitted to the training data using the input variables and output variables by training the model to minimize the mean squared error.

[0067] The hyperparameters of the XGBoost model were adjusted to maximize model performance on the validation dataset. During hyperparameter tuning, models were selected based on their performance on the validation dataset. Specifically, the model with the smallest mean absolute error in predictions between 0.5 and 10 hours after instrument start time was selected. A detailed description of the XGBoost model hyperparameters is provided in the algorithm documentation (https: / / xgboost.readthedocs.io / en / stable / parameter.html). Figures 9A and 9B show the performance of the trained model on the test data from three optical emission spectrometers: iCAPPRO60465 (test), iCAPPRO60474 (test), and iCAPPRO60468 (test). Figure 9A shows the prediction of lateral drift by the XGBoost model for the three instruments on which the test dataset was acquired. Left: actual drift (gray) and predicted drift (black). Right: prediction error (diff). Figure 9B shows the prediction of vertical drift by the XGBoost model for the three instruments on which the test dataset was acquired. Left: Actual drift (gray) and predicted drift (black). Right: Prediction error (difference). The low-frequency waves, which show the largest changes in drift, are the result of temperature cycling of the heating pad, which is consistent with the understanding that drift is caused by thermal deformation of optical components. The high-frequency waves, which show smaller changes in drift, are the result of rapid (e.g., every 1-2 minutes) changes in the RF power supplied to the plasma.

[0068] Exploratory testing on datasets The individual condition parameters used to train the model allow the model to predict drift to different degrees. To explore this, spectra were acquired at a sampling frequency of 1 spectrum / minute, and condition parameters indicative of operating conditions were collected at a sampling frequency of once every 6 seconds. Individual machine learning models were trained using each of the condition parameters alone. The mean squared error of each trained model was

[0069]

number

[0070] [Table 1]

[0071] Figure 9C shows a comparison between the actual lateral drift (first row) and the predicted lateral drift (subsequent rows) resulting from exploratory testing by training a machine learning model using only one condition parameter and then predicting drift in two different spectrometers (iCAPPRO60614 and iCAPPRO60485). Figure 9C shows that the drift predicted by models trained using only temperature or only the current in the optics circuit board is closer to the true drift than either the exhaust pressure (the pressure in the exhaust pipe used to draw in the exhaust plasma gas) or the ambient temperature (the temperature surrounding the scientific instrument). Compared to Figures 9A and 9B, Figure 9C also suggests that the current in the optics circuit board is better at predicting drift caused by fluctuations in the RF supplied to the plasma. This is consistent with the proximity of the plasma to the optics, which results in a correlation between the current supplied to the heating pad to maintain the optics at a stable temperature and the RF supplied to the plasma.

[0072] Computing Infrastructure As noted above, the scientific instrument support module 2000 can be implemented by one or more computing devices. Figure 11 is a block diagram of a computing device 11000 capable of implementing some or all of the scientific instrument support methods disclosed herein, according to various embodiments. In some embodiments, the scientific instrument support module 2000 can be implemented by a single computing device 11000 or by multiple computing devices 11000. Furthermore, as discussed below, the computing device 1100 (or multiple computing devices 11000) implementing the scientific instrument support module 1000 may be part of one or more of the scientific instrument 12010, user local computing device 12020, service local computing device 12010, or remote computing device 12040 of Figure 12.

[0073] 11 is illustrated as having several components, any one or more of these components may be omitted or duplicated if suitable for the application and configuration. In some embodiments, some or all of the components included in computing device 11000 may be mounted on one or more motherboards and enclosed in a housing (e.g., comprising plastic, metal, and / or other materials). In some embodiments, some of these components may be fabricated on a single system-on-a-chip (SoC) (e.g., an SoC may include one or more processing devices 11002 and one or more storage devices 11004). In some embodiments, one or more of these components may be remote from computing device 11000. 11 , but may include interface circuitry (not shown) for coupling to one or more components using any suitable interface (e.g., a Universal Serial Bus (USB) interface, a High-Definition Multimedia Interface (HDMI®) interface, a Controller Area Network (CAN) interface, a Serial Peripheral Interface (SPI) interface, an Ethernet interface, a wireless interface, or any other suitable interface). For example, computing device 11000 may not include display device 11010, but may include display device interface circuitry (e.g., connectors and driver circuits) to which display device 11010 can be coupled.

[0074] The computing device 11000 may include a processing device 11002 (e.g., one or more processing devices). As used herein, the term “processing device” may refer to any device or portion of a device that processes electronic data from registers and / or memory and converts the electronic data into other electronic data that may be stored in registers and / or memory. The processing device 11002 may include one or more digital signal processors (DSPs), application specific integrated circuits (ASICs), central processing units (CPUs), graphics processing units (GPUs), cryptographic processors (dedicated processors that execute cryptographic algorithms in hardware), server processors, or any other suitable processing devices.

[0075] The computing device 11000 may include a storage device 11004 (e.g., one or more storage devices). The storage device 11004 may include one or more memory devices, such as random access memory (RAM) (e.g., static RAM (SRAM) devices, magnetic RAM (MRAM) devices, dynamic RAM (DRAM) devices, resistive RAM (RRAM) devices, or conductive-bridging RAM (CBRAM) devices), hard drive-type memory devices, solid-state memory devices, networked drives, cloud drives, or any combination of memory devices. In some embodiments, the storage device 11004 may include memory that shares a die with the processing device 11002. In such embodiments, the memory may be used as cache memory and may include, for example, embedded dynamic random access memory (eDRAM) or spin transfer torque magnetic random access memory (STT-MRAM). In some embodiments, storage device 11004 may include a non-transitory computer-readable medium having instructions that, when executed by one or more processing devices (e.g., processing device 11002), cause computing device 11000 to perform any suitable of the methods, or portions of those methods, disclosed herein.

[0076] The computing device 11000 may include an interface device 11006 (e.g., one or more interface devices 11006). The interface device 11006 may include one or more communication chips, connectors, and / or other hardware and software for managing communications between the computing device 11000 and other computing devices. For example, the interface device 11006 may include circuitry for managing wireless communications for data transfer to and from the computing device 11000. The term "wireless" and its derivatives may be used to describe circuits, devices, systems, methods, techniques, communication channels, etc. that may communicate data through the use of modulated electromagnetic radiation over a non-solid medium. This term does not imply that the associated device does not include any wiring, although in some embodiments it may not. The circuitry included in the interface device 11006 for managing wireless communications may implement any of several wireless standards or protocols, including, but not limited to, Wi-Fi (IEEE 802.11 family), the IEEE 802.16 standard (e.g., the IEEE 802.16-2005 amendment), Institute for Electrical and Electronic Engineers (IEEE) standards including the Long-Term Evolution (LTE) project with any amendments, updates, and / or revisions (e.g., the Advanced LTE project, the Ultra Mobile Broadband (UMB) project (also referred to as "3GPP2"), etc.).In some embodiments, the circuitry included in the interface device 11006 for managing wireless communications may operate in accordance with a Global System for Mobile Communications (GSM), General Packet Radio Service (GPRS), Universal Mobile Telecommunications System (UMTS), High Speed ​​Packet Access (HSPA), Evolved HSPA (E-HSPA), or LTE network. In some embodiments, the circuitry included in the interface device 11006 for managing wireless communications may operate in accordance with an Enhanced Data for GSM Evolution (EDGE), GSM EDGE Radio Access Network (GERAN), Universal Terrestrial Radio Access Network (UTRAN), or Evolved UTRAN (E-UTRAN). In some embodiments, the circuitry included in interface device 11006 for managing wireless communications may operate in accordance with Code Division Multiple Access (CDMA), Time Division Multiple Access (TDMA), Digital Enhanced Cordless Telecommunications (DECT), Evolution-Data Optimized (EV-DO), and derivatives thereof, as well as any other wireless protocols designated as 3G, 4G, 5G, and beyond. In some embodiments, interface device 11006 may include one or more antennas (e.g., one or more antenna arrays) for receiving and / or transmitting wireless communications.

[0077] In some embodiments, the interface device 11006 may include circuitry for managing wired communications, such as electrical, optical, or any other suitable communications protocol. For example, the interface device 11006 may include circuitry to support communications according to Ethernet technology. In some embodiments, the interface device 11006 may support both wireless and wired communications, and / or may support multiple wired and / or wireless communications protocols. For example, a first set of circuits in the interface device 11006 may be dedicated to short-range wireless communications, such as Wi-Fi or Bluetooth, and a second set of circuits in the interface device 11006 may be dedicated to long-range wireless communications, such as global positioning system (GPS), EDGE, GPRS, CDMA, WiMAX, LTE, EV-DO, etc. In some embodiments, the first set of circuits in the interface device 11006 may be dedicated to wireless communications, and the second set of circuits in the interface device 11006 may be dedicated to wired communications.

[0078] Computing device 11000 may include battery / power circuitry 11008. Battery / power circuitry 11008 may include one or more energy storage devices (e.g., batteries or capacitors) and / or circuitry for coupling components of computing device 11000 to an energy source (e.g., AC line power) separate from computing device 11000.

[0079] The computing device 11000 may include a display device 11010 (e.g., multiple display devices). The display device 11010 may include any visual indicator, such as a heads-up display, a computer monitor, a projector, a touchscreen display, a liquid crystal display (LCD), a light-emitting diode display, or a flat panel display.

[0080] The computing device 11000 may include other input / output (I / O) devices 11012. The other I / O devices 11012 may include, for example, one or more audio output devices (e.g., speakers, headsets, earphones, alarms, etc.), one or more audio input devices (e.g., microphones or microphone arrays), a location device (e.g., a GPS device that communicates with a satellite-based system to receive the location of the computing device 11000, as known in the art), an audio codec, a video codec, a printer, sensors (e.g., thermocouples or other temperature sensors, humidity sensors, pressure sensors, vibration sensors, accelerometers, gyroscopes, etc.), an image capture device such as a camera, a keyboard, a cursor control device such as a mouse, stylus, trackball, or touchpad, a barcode reader, a Quick Response (QR) code reader, or a radio frequency identification (RFID) reader.

[0081] The computing device 11000 may have any form factor suitable for its application and configuration, such as a handheld or mobile computing device (e.g., a mobile phone, smartphone, mobile internet device, tablet computer, laptop computer, netbook computer, ultrabook computer, personal digital assistant (PDA), ultra-mobile personal computer, etc.), a desktop computing device, or a server computing device or other networked computing component.

[0082] One or more computing devices implementing any of the scientific instrument support modules or methods disclosed herein may be part of a scientific instrument support system. Figure 12 is a block diagram of an exemplary scientific instrument support system 12000 in which some or all of the scientific instrument support methods disclosed herein may be implemented, according to various embodiments. The scientific instrument support modules and methods disclosed herein (e.g., the scientific instrument support module 1000 of Figure 2 and the methods 5000, 6000, and 7000 of Figures 5, 6, and 7) may be implemented by one or more of the scientific instrument 12010, the user local computing device 12020, the service local computing device 12030, or the remote computing device 12040 of the scientific instrument support system 12000.

[0083] Any of the scientific instrument 12010, the user local computing device 12020, the service local computing device 12030, or the remote computing device 12040 may include any of the embodiments of the computing device 11000 discussed in this specification with reference to FIG. 11, and any of the scientific instrument 12010, the user local computing device 12020, the service local computing device 12030, or the remote computing device 12040 may take the form of any suitable of the embodiments of the computing device 11000 discussed in this specification with reference to FIG. 11.

[0084] The scientific instrument 12010, the user local computing device 12020, the service local computing device 12030, or the remote computing device 12040 may each include a processing device 12002, a storage device 12004, and an interface device 12006. The processing device 12002 may take any suitable form, including the form of any of the processing devices 11002 discussed herein with reference to Figure 11, and the processing devices 12002 included in different ones of the scientific instrument 12010, the user local computing device 12020, the service local computing device 12030, or the remote computing device 12040 may take the same or different forms. The storage device 12004 may take any suitable form, including any of the forms of the storage device 11004 discussed herein with reference to Figure 11, and the storage device 12004 included in different ones of the scientific instrument 12010, the user local computing device 12020, the service local computing device 12030, or the remote computing device 12040 may take the same form or different forms. The interface device 12006 may take any suitable form, including any of the forms of the interface device 11006 discussed herein with reference to Figure 11, and the interface device 12006 included in different ones of the scientific instrument 12010, the user local computing device 12020, the service local computing device 12030, or the remote computing device 12040 may take the same form or different forms.

[0085] The scientific instrument 12010, the user local computing device 12020, the service local computing device 12030, and the remote computing device 12040 may communicate with other elements of the scientific instrument support system 12000 via communication paths 12008. The communication paths 12008 may communicatively couple the interface devices 12006 of different elements of the scientific instrument support system 12000, as shown, and may be wired or wireless communication paths (e.g., according to any of the communication techniques discussed herein with reference to the interface device 11006 of the computing device 11000 of FIG. 11 ). While the particular scientific instrument support system 12000 depicted in FIG. 12 includes communication paths between each pair of the scientific instrument 12010, the user local computing device 12020, the service local computing device 12030, and the remote computing device 12040, this “fully connected” implementation is merely illustrative, and in various embodiments, various ones of the communication paths 12008 may not be present. For example, in some embodiments, the service local computing device 12030 may have a direct communication path 12008 between its interface device 12006 and the interface device 12006 of the scientific instrument 12010, or may alternatively communicate with the scientific instrument 12010 via a communication path 12008 between the service local computing device 12030 and the user local computing device 12020, and a communication path 12008 between the user local computing device 12020 and the scientific instrument 12010.

[0086] Scientific instrument 12010 may include any suitable scientific instrument, such as an optical emission spectrometer, for example, an inductively coupled plasma optical emission spectrometer 1000 as shown in FIG.

[0087] The user local computing device 12020 may be a computing device that is local to a user of the scientific instrument 12010 (e.g., according to any of the embodiments of the computing device 11000 discussed herein). In some embodiments, the user local computing device 12020 may also be local to the scientific instrument 12010, although this need not be the case; for example, a user local computing device 12020 in a user's home or office may be remote from the scientific instrument 12010 but may communicate with it, such that the user can control and / or access data from the scientific instrument 12010 using the user local computing device 12020. In some embodiments, the user local computing device 12020 may be a laptop, smartphone, or tablet device. In some embodiments, the user local computing device 12020 may be a portable computing device.

[0088] The service local computing device 12030 may be a computing device that is local to an entity that provides services to the scientific instrument 12010 (e.g., according to any of the embodiments of computing device 11000 discussed herein). For example, the service local computing device 12030 may be local to the manufacturer of the scientific instrument 12010 or a third-party service company. In some embodiments, the service local computing device 12030 may communicate with the scientific instrument 12010, the user local computing device 12020, and / or the remote computing device 12040 (e.g., via a direct communication path 12008 or via multiple "indirect" communication paths 12008, as discussed above) to receive data regarding the operation of the scientific instrument 12010, the user local computing device 12020, and / or the remote computing device 12040 (e.g., results of self-diagnostic tests of the scientific instrument 12010, calibration coefficients used by the scientific instrument 12010, measurements of sensors associated with the scientific instrument 12010, etc.). In some embodiments, the service local computing device 12030 can communicate with the scientific instrument 12010, the user local computing device 12020, and / or the remote computing device 12040 (e.g., via a direct communication path 12008 or multiple "indirect" communication paths 12008, as discussed above) to transmit data to the scientific instrument 12010, the user local computing device 12020, and / or the remote computing device 12040 (e.g., to update programmed instructions such as firmware in the scientific instrument 12010, to initiate the performance of a test or calibration sequence in the scientific instrument 12010, to update programmed instructions such as software in the user local computing device 12020 or the remote computing device 12040, etc.).A user of the scientific instrument 12010 can use the scientific instrument 12010 or the user local computing device 12020 to communicate with the service local computing device 12030 to report a problem with the scientific instrument 12010 or the user local computing device 12020, to request a technician visit to improve the operation of the scientific instrument 12010, to order consumable or replacement parts associated with the scientific instrument 12010, or for other purposes.

[0089] The remote computing device 12040 may be a computing device that is remote from the scientific instrument 12010 and / or remote from the user local computing device 12020 (e.g., according to any of the embodiments of computing device 11000 discussed herein). In some embodiments, the remote computing device 12040 may be included in a data center or other large-scale server environment. In some embodiments, the remote computing device 12040 may include network-attached storage (e.g., as part of the storage device 12004). The remote computing device 12040 may store data generated by the scientific instrument 12010, perform analysis of the data generated by the scientific instrument 12010 (e.g., according to programmed instructions), facilitate communications between the user local computing device 12020 and the scientific instrument 12010, and / or facilitate communications between the service local computing device 12030 and the scientific instrument 12010.

[0090] In some embodiments, one or more of the elements of the scientific instrument support system 12000 illustrated in Figure 12 may not be present. Further, in some embodiments, more than one of various of the elements of the scientific instrument support system 12000 of Figure 12 may be present. For example, the scientific instrument support system 12000 may include multiple user local computing devices 12020 (e.g., different user local computing devices 12020 associated with different users or at different locations). In another example, the scientific instrument support system 12000 may include multiple scientific instruments 12010 that all communicate with a service local computing device 12030 and / or a remote computing device 12040; in such an embodiment, the service local computing device 12030 can monitor these multiple scientific instruments 12010, and the service local computing device 12030 can generate updates or other information that can be "broadcast" to the multiple scientific instruments 12010 simultaneously. Different scientific instruments 12010 in the scientific instrument support system 12000 may be located near each other (e.g., in the same room) or far from each other (e.g., on different floors of a building, in different buildings, in different cities, etc.). In some embodiments, the scientific instruments 12010 may be connected to an Internet-of-Things (IoT) stack that enables command and control of the scientific instruments 12010 through web-based applications, virtual or augmented reality applications, mobile applications, and / or desktop applications. Any of these applications may be accessed by a user operating a user local computing device 12020 that communicates with the scientific instruments 12010 by way of an intervening remote computing device 12040. In some embodiments, the scientific instruments 12010 may be sold by a manufacturer as part of a local scientific instrument computing unit 12012, along with one or more associated user local computing devices 12020.

[0091] In some embodiments, different ones of the scientific instruments 12010 included in the scientific instrument support system 12000 may be different types of scientific instruments 12010. For example, one scientific instrument 12010 may be an optical emission spectrometer. In some such embodiments, the remote computing device 12040 and / or the user local computing device 12020 may combine data from different types of scientific instruments 12010 included in the scientific instrument support system 12000.

Claims

1. 1. A method for reducing distortion in an optical emission spectrum obtained from an optical emission spectrometer, the optical emission spectrometer comprising optics for forming a spectrum, the method comprising: obtaining input data including a spectrum recorded using the optical emission spectrometer and one or more condition parameters indicative of the operating conditions of the optical emission spectrometer or the environment of the optical emission spectrometer at the time the spectrum was recorded using the spectrometer; providing a machine learning model configured to, in response to the input data, output data including one or more transformation parameters for a transformation applied to the acquired spectrum to mitigate distortions of the spectrum due to a mismatch between the operating conditions and a baseline operating condition; applying the input data as an input to the machine learning model and obtaining one or more transformation parameters as an output of the machine learning model; applying the transformation to the acquired spectrum in accordance with the acquired one or more transformation parameters to reduce distortions in the spectrum due to mismatches between the operating conditions and the baseline operating conditions.

2. The method of claim 1 , wherein the spectrum comprises a set of intensity values ​​across each two-dimensional location, and the transformation comprises a motion that varies across locations.

3. The method of claim 2 , wherein the action comprises applying a deformation field to the optical emission spectrum.

4. The method of any one of claims 1 to 3, wherein the one or more condition parameters include a parameter indicative of temperature.

5. The method of any one of claims 1 to 4, wherein the one or more condition parameters indicate operating conditions of the optical system of the optical emission spectrometer.

6. The method of claim 4 or 5, wherein the one or more condition parameters include at least one temperature measurement obtained from a temperature sensor attached to a mechanical structure of the optical emission spectrometer.

7. The method of claim 6 , wherein the mechanical structure of the optical emission spectrometer supports one or more optical components of the optical system.

8. The method according to any one of claims 1 to 7, wherein the one or more condition parameters include one or more condition change parameters that indicate a change in the condition or a direction of the change in the condition.

9. The method of any one of claims 1 to 8, wherein the input data comprises a time series of one or more condition parameters at each of a plurality of time points.

10. 10. The method of claim 9, wherein the output data includes a corresponding time series of one or more transformation parameters of a transformation at each of the plurality of time points, and the method includes applying the transformation to the optical emission spectrum acquired at each time point according to the one or more transformation parameters at the time point of the acquired instance of the output data.

11. The method of any one of claims 1 to 10, wherein the machine learning model comprises a decision tree-based ensemble machine learning algorithm.

12. The method of any one of claims 1 to 11, wherein the transformation comprises at least one of a translation, a rotation, a scaling operation, a twisting operation, a stretching operation, and a deformation field.

13. The one or more condition parameters are: a parameter indicating a heating current applied to a heating device for heating the optical system and stabilizing the temperature of the optical system; a parameter indicative of the temperature of the environment of the spectrometer; a parameter indicating the temperature of the optical system; a parameter indicative of at least one temperature measurement obtained from a temperature sensor attached to a mechanical structure supporting one or more optical components of the optical system; a parameter indicative of RF power of an RF generator for generating a plasma for use in obtaining the emission spectrum; and and a parameter indicative of an exhaust pressure of a plasma chamber for containing a plasma for use in obtaining the optical emission spectrum.

14. The method according to any one of claims 1 to 13, wherein the optical emission spectrometer is a plasma emission spectrometer configured to record the emission spectrum of light emitted from a plasma.

15. The method of any one of claims 1 to 14, wherein the spectrum is an echelle spectrum.

16. One or more non-transitory computer readable media comprising instructions thereon that, when executed by one or more processing devices of a scientific instrument support apparatus, cause the scientific instrument support apparatus to perform the method of any one of claims 1 to 15.

17. 17. A scientific instrument system comprising an optical emission spectrometer for recording an optical emission spectrum and a scientific instrument support device configured to implement the method of any one of claims 1 to 16 so as to reduce distortions in said optical spectrum.

18. 20. The scientific instrument system of claim 17, wherein the optical emission spectrometer is an inductively coupled plasma optical emission spectrometer.

19. 1. A method of training a machine learning model to output one or more transformation parameters in response to one or more condition parameters, comprising: the one or more condition parameters are indicative of an operating condition of the optical emission spectrometer or an environment of the optical emission spectrometer at a time when the optical emission spectrometer is used to record an optical emission spectrum of a reference analyte; the one or more transformation parameters of a transformation are applied to each recorded optical emission spectrum to reduce distortion of the spectrum relative to a baseline optical emission spectrum recorded using the optical emission spectrometer having an optical system at baseline operating conditions; The method comprises: obtaining a training data set including training data pairs, each training data pair including a value of the one or more condition parameters and a respective value of the one or more transformation parameters for a respective operating condition of the optical system that differs from the baseline operating condition; in response to respective values ​​of the one or more condition parameters of the training data pairs; and adjusting parameters of the machine learning model to reduce discrepancies between values ​​of the one or more transformation parameters of the training data pairs and values ​​of the one or more transformation parameters output by the machine learning model.

20. 1. A method for obtaining training data for training a machine learning model, the method comprising: recording a plurality of optical emission spectra of a reference analyte using the optical emission spectrometer for each operating condition of the optical emission spectrometer; storing input data for the machine learning model, the input data including, for each recorded optical emission spectrum, one or more parameters indicative of the respective operating conditions; for each recorded optical emission spectrum, adjusting one or more transformation parameters of the transform using the transform to align the optical emission spectrum with a baseline optical emission spectrum of the reference analyte, the baseline optical emission spectrum being recorded for baseline operating conditions; and storing, for each optical emission spectrum, output data for the machine learning model including each adjusted one or more transformation parameters in association with each of the input data for each optical emission spectrum as a training data pair.

21. 21. The method of claim 20, wherein the optical emission spectrum and the baseline optical emission spectrum each comprise a set of intensity values ​​across a respective two-dimensional location, and the transformation comprises a motion that varies across multiple locations.

22. 22. The method of claim 21, wherein the action comprises applying a distorting field to the optical emission spectrum.

23. 22. The method of claim 21 , wherein the optical emission spectrum and the baseline optical emission spectrum are respective images, and adjusting the one or more transformation parameters comprises comparing respective image intensities between the respective images.

24. 24. The method of claim 23, wherein adjusting the one or more transformation parameters comprises using an adversarial similarity network.

25. 20. A machine learning model trained using the method of claim 19.

26. A training dataset for a machine learning model obtained using the method of any one of claims 20 to 24.

Images