JP7483994B2 - Information processing device, information processing method, and program - Google Patents

Description

The present invention relates to an information processing device, an information processing method, and a program.

Various adjustment methods have been proposed for electron microscopes in the past. For example, the following Patent Document 2 describes the provision of a display device that displays an image 15a for adjusting the optical axis obtained based on secondary electrons detected by a detector when adjusting the optical axis.

In addition, the following Patent Document 3 aims to quantify an artificial judgment criterion and to judge whether or not the axis of the charged particle beam needs to be adjusted based on the quantified judgment criterion. Patent Document 3 describes that the adjustment conditions of the optical element that adjusts the charged particle beam are changed, multiple images are acquired under the different adjustment conditions, and from the multiple acquired images, images whose image quality is acceptable or unacceptable are selected, a first image quality evaluation value is acquired based on the selected images, and the acquired first image quality evaluation value is compared with a second image quality evaluation value obtained from an image obtained by scanning the charged particle beam.

JP 2020-74294 A JP 2016-35800 A JP 2010-182424 A JP 2008-84565 A JP 2015-2059 A Japanese Patent Application Laid-Open No. 6-325714

However, the above-mentioned conventional technology does not provide sufficient support for users when adjusting or maintaining an electron microscope. This means that there is a risk of variation in the adjustment or maintenance of an electron microscope due to individual skill. As a result, the device may not be used in a desirable condition.

In one aspect, the present invention assists a user in adjusting or maintaining an electron microscope.

An embodiment of the present invention is exemplified by an information processing device including a control unit, which executes the following: acquiring a first confirmation image taken by the electron microscope from a confirmation sample for confirming the state of the electron microscope after a predetermined adjustment is performed, acquiring a first judgment result as to whether the adjustment result of the electron microscope after the predetermined adjustment is good or not by inputting the confirmation image into a trained model trained using as teacher data a first image taken by the electron microscope from the confirmation sample when the adjustment result of the predetermined adjustment is good and a second image taken by the electron microscope from the confirmation sample when the adjustment result of the predetermined adjustment is not good, and outputting the first judgment result to an output device.
The embodiment of the present invention can also be exemplified by the following information processing device. The information processing device executes the following operations: acquiring a confirmation image taken by the electron microscope from a confirmation sample for confirming the state of the electron microscope after the optical axis of the electron beam has been adjusted; acquiring a judgment result as to whether the contamination state of the aperture is equal to or lower than a standard by inputting the confirmation image into a trained model trained using, as teacher data, a first image taken by the electron microscope from the confirmation sample when the contamination state of all apertures provided in the electron microscope is equal to or lower than a standard and a second image taken by the electron microscope from the confirmation sample when the contamination state of any aperture provided in the electron microscope has deteriorated beyond the standard; and outputting, to an output device, an instruction to correct the astigmatism of the electron beam or an instruction to replace at least one aperture provided in the electron microscope when the judgment result is not good.

This information processing device can assist users in adjusting or maintaining their electron microscopes.

FIG. 1 is a diagram illustrating a configuration of an information system including an electron microscope according to an embodiment. FIG. 2 is a diagram illustrating the state of the optical axis. FIG. 3 is a diagram illustrating an example of a knob for adjusting the optical axis of the electron beam. FIG. 4 is a diagram illustrating electron microscope images of a sample taken when the optical axis of the electron beam is placed at a proper position and when the optical axis of the electron beam is deviated from the proper position. FIG. 5 is a diagram illustrating an example of a hardware configuration of an information processing device. FIG. 6 is a diagram illustrating an example of a program configuration and data flow in an information system. FIG. 7 is a diagram illustrating an example of the number of pieces of training data and the determination results of the confirmation image using the trained model. FIG. 8 is a diagram illustrating a processing flow of the information processing device in this embodiment. FIG. 9 is a diagram illustrating a process performed by the information processing device according to the first modification of the present embodiment. FIG. 10 is a diagram illustrating a process performed by an information processing device according to the second modification of the present embodiment.

Below, an electron microscope according to one embodiment, an information system including the electron microscope, an information processing method, and a program will be described with reference to the drawings.

(System configuration)
Fig. 1 is a diagram illustrating an example of the configuration of an information system including an electron microscope 3 according to the present embodiment. This information system has an electron microscope 3 and an information processing device 1. There is no limitation on the type of electron microscope 3, and the electron microscope 3 may be, for example, a scanning electron microscope, a transmission electron microscope, or the like. Fig. 1 illustrates an example of the configuration of a scanning electron microscope.

As shown in FIG. 1, the electron microscope 3 has an electron gun 32, an electron optical system 3EO that controls the electron beam emitted from the electron gun 32, a stage 39 on which a sample 4 is mounted, a detector 37, and a control circuit 31.

The electron gun 32 extracts electrons from a filament or the like at a predetermined acceleration voltage, and supplies an electron beam to the electron optical system 3EO. The electron optical system 3EO has functions such as shaping, converging, and deflecting the electron beam. The electron optical system 3EO also has a converging lens 33, a converging aperture 34, an astigmatism correction coil 3A, an objective lens 35, and an objective aperture 36. However,
The electron optical system 3EO may have a plurality of convergent lenses 33 and a plurality of convergent apertures 34 .

The converging lens 33 controls the beam diameter of the electron beam passing through the converging aperture 34 and makes it enter the objective lens 35 and the objective aperture 36. The astigmatism correction coil 3A controls the trajectory of the electron beam so that the cross-sectional shape of the electron beam becomes close to circular. The objective lens 35 converges the electron beam controlled by the converging lens 33 to the minimum beam size at the surface of the sample 4 to form an electron probe. The objective aperture 36 reduces aberration by restricting the electron beam converged by the converging lens 33 to the vicinity of the central axis and converging it with the objective lens 35.

A sample 4 is mounted on the stage 39. In this embodiment, the sample 4 mounted on the stage 39 when the electron microscope 3 is maintained, inspected, or adjusted is called a standard sample. The electron microscope 3 irradiates the sample 4 with an electron beam focused by the objective lens 35, detects secondary electrons or reflected electrons with the detector 37, and obtains an electron microscope image of the surface of the sample 4.

The control circuit 31 controls each part of the lens barrel of the electron microscope 3 and the stage 39 according to the command from the information processing device 1, or obtains detection values from the sensors of each part. For example, the control circuit 31 controls the emission of electrons from the electron gun 32 by a control signal C1 to the electron gun 32. For example, the control circuit 31 controls the extraction voltage of electrons from the electron gun 32. For example, the control circuit 31 controls the convergence of the electron beam by a control signal C2 to the convergence lens 33. For example, the control circuit 31 corrects the astigmatism of the electron beam by a control signal CA to the astigmatism correction coil 3A. Furthermore, for example, the control circuit 31 controls the objective lens 35 by a control signal C3 to converge the electron beam on the surface of the sample 4 to form a focus. Furthermore, the control circuit 31 deflects the electron beam by a deflection coil or electrostatic deflection plate included in the electron optical system 3EO to scan the sample 4. Furthermore, the control circuit 31 generates an electron microscope image of the sample 4 based on the detection signal from the detector 37 and information on the deflection position of the electron beam. Furthermore, the control circuit 31 controls the movement of the stage 39 using a control signal C4 sent to the stage 39.

The information processing device 1 sends various commands to the control circuit 31 to control or monitor each part of the electron microscope 3. The information processing device 1 also displays an electron microscope image of the sample 4 on a display (for example, the display device 14 in FIG. 5). The information processing device 1 communicates with various systems, such as external computers, servers, and database systems, via a network N1. The network N1 may be a Local Area Network (LAN) within the premises where the electron microscope 3 is installed, or a communication line connecting remote locations. The network N1 may be a dedicated line such as a Virtual Private Network (VPN), or a public network such as the Internet.

(Example of maintenance and adjustment of an electron microscope)
Examples of maintenance and adjustment items for electron microscopes include the following:
(1) Replacement of the filament when the filament breaks. (2) Cleaning of the Wehnelt electrode, etc. when replacing the filament. (3) Cleaning of the anode. (4) Replacement of the convergent aperture 34. (5) Replacement of the objective aperture 36. (6) Cleaning of the cooling fan intake. (7) Adjustment of the optical axis of the electron beam. After carrying out (1) to (5), the optical axis adjustment of (7) is carried out.

2A shows an example of the state of the optical axis. Fig. 2A shows an example of the state in which the optical axis of the electron beam is adjusted to a proper position. When the optical axis is in a proper state, the central axis of the electron beam extracted from the electron gun 32 passes through the center of the circular convergence aperture 34 and the center of the objective aperture 36, and is irradiated onto the sample 4 on the stage 39.

(B) of FIG. 2 illustrates a state in which the optical axis of the electron beam is deviated from the proper position. In a state in which the optical axis of the electron beam is deviated from the proper position, for example, the central axis of the electron beam extracted from the electron gun 32 does not pass through the center of the circular convergence aperture 34, but passes through the convergence aperture 34 with a cross-sectional shape that is asymmetric with respect to the center of the convergence aperture 34. In this case, the convergence of the electron beam by the objective lens 35 is likely to be insufficient, and as a result, it is difficult to form an electron probe with the minimum beam size on the surface of the sample 4. Therefore, it becomes difficult to focus the electron microscope image obtained from the sample 4. In this case, the current of the electron beam is reduced compared to the state in which the optical axis of (A) is placed in the proper position. In addition, the S/N (signal/noise) ratio of the detection signal obtained from the detector 37 is reduced compared to the state in which the optical axis of (A) is placed in the proper position. Furthermore, the brightness of the electron microscope image of the sample 4 obtained from the detector 37 is reduced compared to the state in which the optical axis of (A) is placed in the proper position.

Figure 3 illustrates an example of a knob for adjusting the optical axis of the electron beam. In this embodiment, the optical axis of the electron beam is adjusted by moving the electron source of the electron gun 32 on a horizontal plane perpendicular to the central axis of the tube of the electron microscope 3. The electron source of the electron gun 32 is, for example, a part including the cathode (filament), Wehnelt electrode, and anode of the electron gun 32. Figure 3 illustrates knobs 32A to 32D provided on the outside of the housing of the electron gun 32. In this example, the knobs 32A and 32C adjust the position in the direction of a first axis (e.g., the X-axis) on a horizontal plane perpendicular to the central axis of the tube. In addition, the knobs 32B and 32D adjust the position in the direction of a second axis (e.g., the Y-axis) perpendicular to the first axis on a horizontal plane perpendicular to the central axis of the tube. A user or maintenance person of the electron microscope 3 adjusts the pair of knobs 32A and 32C, or the pair of knobs 32B and 32D, so that the position of the optical axis of the electron beam coincides with the central axis of the lens barrel of the electron microscope 3. For example, the user or maintenance person operates knobs 32A and 32C to adjust so that the brightness of the electron microscope image of the sample 4 is relatively the brightest.

Figure 4 illustrates electron microscope images of the sample 4 taken when the optical axis of the electron beam is placed at the correct position and when the optical axis of the electron beam is shifted from the correct position. (A) of Figure 4 is an electron microscope image of the sample 4 when the optical axis is adjusted to the correct position. (B), (C), and (D) of Figure 4 are electron microscope images of the sample 4 when the knobs 32A and 32C (or knobs 32B and 32D) in the state of (A) of Figure 4 are rotated 11.25 degrees, 22.5 degrees, and 33.75 degrees, respectively, with the position of the knobs 32A and 32C (or knobs 32B and 32D) being 0 degrees. As already explained in Figure 3, the electron source of the electron gun 32 moves in the horizontal plane by a distance corresponding to the amount of rotation of the knobs 32A and 32C (or knobs 32B and 32D). As a result, (B), (C), and (D) of Figure 4 are electron microscope images of the sample 4 when the optical axis of the electron beam is shifted from the correct position. In Figure 4, as shown in (A), (B), (C), and (D), the contrast and brightness of the electron microscope image decrease as the knob is rotated, i.e., the amount of deviation of the optical axis increases.

In the present embodiment, a correct label is set for the image in FIG. 4 (A), and an incorrect label is set for the images in FIG. 4 (B), (C), and (D), and a learning system using deep learning is trained to create a trained model. In the present embodiment, the information system acquires an electron microscope image taken from the sample 4 after the optical axis adjustment, inputs the image into the trained model, and determines whether the optical axis of the electron beam is in the correct position or in an incorrect position.

(composition)
5 illustrates an example of the hardware configuration of the information processing device 1. The configuration of the information processing device 1 is the same as that of a normal computer. The information processing device 1 has a CPU 11, a main storage device 12, and external devices connected through an external interface (I/F), and executes information processing by a computer program. Examples of the external devices include an external storage device 13, a display device 14, an operation device 15, and a communication device 16. The CPU 11, the main storage device 12, and the external interface (I/F) can be referred to as a control unit.

The CPU 11 executes a computer program that is executable and deployed in the main storage device 12, and provides the functions of the information processing device 1. The CPU 11 is also called a processor. However, the CPU 11 is not limited to a single processor, and may have a multi-processor configuration. The CPU 11 may be a single processor connected to a single socket and have a multi-core configuration. Furthermore, at least a part of the processing of the information processing device 1 may be provided by a dedicated processor such as a Digital Signal Processor (DSP), a Graphics Processing Unit (GPU), a numerical calculation processor, a vector processor, an image processing processor, or an Application Specific Integrated Circuit (ASIC). Furthermore, at least a part of the information processing device 1 may be a dedicated large scale integration (LSI) such as a Field-Programmable Gate Array (FPGA), or other digital circuit. Furthermore, at least a part of the information processing device 1 may include an analog circuit.

The main storage device 12 is simply called memory, and stores the computer program executed by the CPU 11, the data processed by the CPU 11, etc. The main storage device 12 is Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Read Only Memory (ROM), etc. Furthermore, the external storage device 13 is used, for example, as a storage area that assists the main storage device 12, and stores the computer program executed by the CPU 11, the data processed by the CPU 11, etc. The external storage device 13 is a hard disk drive, a solid state drive (SSD), etc. Furthermore, the information processing device 1 may be provided with a drive device for a removable storage medium. The removable storage medium is, for example, a Blu-ray disc, a Digital Versatile Disc (DVD), a Compact Disc (CD), a flash memory card, etc.

The information processing device 1 also has a display device 14, an operation device 15, and a communication device 16. The display device 14 is, for example, a liquid crystal display or an electroluminescence panel. The operation device 15 is, for example, a keyboard or a pointing device. In this embodiment, a mouse is exemplified as a pointing device. The communication device 16 exchanges data with other devices on the network N1 (see FIG. 1).

Figure 6 is a diagram illustrating the program configuration and data flow in this information system. This information system has a device status acquisition program 101, a maintenance management support program 102, and a learning system 111. The device status acquisition program 101 and the maintenance management support program 102 are programs executed by the CPU 11 of the information processing device 1. The learning system 111 may be a program executed by a computer. In that case, the learning system 111 may be a system provided by the CPU 11 of the information processing device 1 executing a program in the main memory device 12. The learning system 111 may also be a system provided by another computer that can communicate with the information processing device 1 through the network N1 (see Figure 1) executing a program.

Furthermore, the learning system 111 may be a system including a neural network as hardware. Furthermore, when the learning system 111 is a system including a neural network as hardware, the learning system 111 may be provided in the premises where the information processing device 1 is installed. Furthermore, the learning system 111 may operate in the same housing as the information processing device 1. Furthermore, when the learning system 111 is a system including a neural network as hardware, the learning system 111 may be a system capable of communicating with the information processing device 1 through a network N1 (see FIG. 1 ).

The learning system 111 creates a learned model A 112A as a first learned model, a learned model B 112B as a second learned model, and the like by learning the teacher data. The teacher data for creating the learned model A 112A includes, for example, as described in FIG. 4, an electron microscope image of the sample 4 taken with the optical axis of the electron beam adjusted to an appropriate position, to which a correct answer label has been added. The teacher data also includes data in which an incorrect answer label has been added to an electron microscope image of the sample 4 taken with the optical axis of the electron beam shifted from the appropriate position. The teacher data is a collection of data to which these labels have been added. Note that, when creating such teacher data, adjustments other than the optical axis of the electron beam, such as adjustments to a state in which there is no contamination of the convergence aperture 34 or the objective aperture 36, are made. The electron microscope image of the sample 4 taken with the optical axis of the electron beam adjusted to an appropriate position is an example of a first image taken from a confirmation sample with the electron microscope 3 when the adjustment result by the predetermined adjustment is good. In addition, an electron microscope image of the sample 4 captured when the optical axis of the electron beam is shifted from the correct position is an example of a second image captured from the confirmation sample by the electron microscope 3 when the adjustment results from the specified adjustment are not satisfactory.

The learning system 111 has, for example, a convolutional layer network for deep learning. The convolutional layer network may be a hardware network or a virtual one constructed by a computer program in the CPU 11. The learning system 111 generates a trained model A 112A or the like in which the filtering coefficients of the multiple convolutional layers included in the convolutional layer network are adjusted by inputting the above-mentioned teacher data. Then, for example, when a confirmation electron microscope image is input from the maintenance management support program 102, the trained model A 112A determines whether the optical axis of the electron microscope 3 from which the confirmation electron microscope image was obtained has been adjusted (correct) or has not been adjusted (incorrect). Note that, as described with respect to FIG. 4, the contrast and brightness of an electron microscope image whose optical axis has not been adjusted is lower than that of an electron microscope image whose optical axis has been adjusted (correct).

Figure 7 illustrates the number of learning data items and the judgment results of the confirmation electron microscope image by the trained model A 112A. Figure 7 (A) illustrates the configuration of the training data and the number of electron microscope images. In the example of Figure 7, the number of electron microscope images is 60 correct data (TRUE) and 60 incorrect data (FALSE). In addition, among the incorrect data, there are 15 pieces of data with knob rotation angles (see Figure 3) of 11.25 degrees, 22.5 degrees, 33.75 degrees, and 45 degrees. In this case, when the correct data and the incorrect data are photographed, all adjustment items other than the optical axis have been adjusted. Therefore, for example, these training data are acquired without contamination of the convergence aperture 34, the objective aperture 36, etc., and with astigmatism corrected.

Figure 7 (B) shows an example of the judgment result for the confirmation electron microscope image by the trained model A 112A. In the judgment by the trained model A 112A, the results of a recall rate of 75%, a precision rate of 88.2%, and an F-value of 81.1% were obtained for the electron microscope image that was taken when the optical axis adjustment was misaligned and should be judged as incorrect data (FALSE). Furthermore, the results of a recall rate of 90%, a precision rate of 78.3%, and an F-value of 83.7% were obtained for the electron microscope image that was taken when the optical axis was adjusted and should be judged as correct data (TRUE). Note that when the confirmation electron microscope image in this case was taken, all adjustment items other than the optical axis had been adjusted.

Here, recall is the proportion of electron microscope images that the learning system judges to be "correct" compared to the electron microscope images whose true value is "correct." Precision is generally an index that indicates how accurate the output results are in a learning system that makes judgments using machine learning. For example, it is the proportion of electron microscope images that the learning system judges to be "correct" that actually have the correct true value. The F-value is the harmonic mean of recall and precision.

For example, the teacher data for creating the trained model B 112B may include images taken corresponding to the degree of contamination of one or more of the convergence aperture 34 and the objective aperture 36. For example, the teacher data may include data in which a correct answer label is attached to an electron microscope image taken of the sample 4 when the contamination level of the convergence aperture 34 and the objective aperture 36 is sufficiently low. The teacher data may also include data in which an incorrect answer label is attached to an electron microscope image taken of the sample 4 in a state in which the contamination level of one or more of the convergence aperture 34 and the objective aperture 36 has deteriorated beyond a predetermined limit. The teacher data may be a collection of data with such labels. Note that when the confirmation electron microscope image is taken in this case, all adjustment items other than the convergence aperture 34 and the objective aperture 36 have been adjusted. The electron microscope image taken of the sample 4 when the contamination level of the convergence aperture 34 and the objective aperture 36 is sufficiently low is an example of a third image. The electron microscope image taken of the sample 4 in a state in which the contamination level of one of the convergence aperture 34 and the objective aperture 36 has deteriorated beyond a predetermined limit is an example of a fourth image.

The learning system 111 generates the trained model B 112B by inputting such teaching data. Then, as illustrated in FIG. 6, the trained model B 112B judges the state of the convergence aperture 34 and the objective aperture 36 of the electron microscope 3 when a confirmation electron microscope image is input from the maintenance management support program 102. For example, the trained model B 112B judges that the contamination of the convergence aperture 34 and the objective aperture 36 is sufficiently low (correct answer) in the electron microscope 3 from which the confirmation electron microscope image was acquired. The trained model B 112B also judges that the level of contamination of at least one of the convergence aperture 34 and the objective aperture 36 exceeds a predetermined limit (incorrect answer). Note that when at least one of the convergence aperture 34 and the objective aperture 36 is contaminated, astigmatism becomes large, and the pattern in the electron microscope image of the surface of the sample 4 is distorted in a certain direction to a certain degree uniformly, and often appears to be stretched in that direction.

The device status acquisition program 101 acquires various information via the control circuit 31 of the electron microscope 3. The information acquired by the device status acquisition program 101 is, for example, an electron microscope image of the sample 4, various physical quantities, etc. Examples of the various physical quantities include the brightness of the electron microscope image, the change in the electron microscope image when adjusting the focus, the amount of astigmatism correction, the contrast of the image, etc.

The brightness of the electron microscope image can be calculated, for example, as the average value of each pixel in the pixel array in the electron microscope image of the sample 4. The brightness of the electron microscope image may be the result of a user's judgment of relative brightness. For example, when the user operates the optical axis adjustment knobs 32A and 32C illustrated in FIG. 3, the user may determine that the electron microscope image displayed on the display device 14 becomes relatively brighter or darker, and adjust the optical axis. The information processing device 1 may receive such an adjustment result from the user.

The manner in which the electron microscope image changes during focus adjustment is, for example, whether or not the electron microscope image displayed on the display device 14 changes with a translation on the screen of the display device 14 when the current of the objective lens 35 is changed. If the optical axis of the electron beam is deviated from the center of the lens barrel (for example, the center of the aperture of the convergence aperture 34), the irradiation position of the electron beam on the surface of the sample 4 will shift when the current of the objective lens 35 is changed to change the focus of the electron beam. As a result, the electron microscope image displayed on the display device 14 moves in a translation on the screen. In this way, the deviation of the optical axis of the electron beam from the center of the lens barrel can be determined by the translation on the screen of the electron microscope image corresponding to the change in the current value of the objective lens 35.

The amount of astigmatism correction can be determined as the current value of the astigmatism correction coil 3A. The presence or absence of astigmatism can be determined by the deformation of the shape in the electron microscope image of the sample 4. For example, if the sample 4 has a known surface structure, it is sufficient to detect that the pattern in the electron microscope image of the sample 4 is uniformly distorted in a specific direction with respect to the known surface structure. The presence or absence of astigmatism may be determined by the user. Furthermore, if the user cannot correct the astigmatism even when the current value of the astigmatism correction coil 3A is maximized, and the electron microscope image is uniformly distorted in a specific direction, there is a high possibility that the contamination of the convergence aperture 34 or the objective aperture 36 has exceeded the limit. The information system determines the state of the electron microscope based on the judgment of the electron microscope image using the learned teacher data and the input from the user. The information system also supports maintenance and adjustment work for the electron microscope 3.

(Processing flow)
FIG. 8 illustrates a processing flow of the information processing device 1 in this embodiment. This processing is executed by the CPU 11 of the information processing device 1 according to a computer program executablely deployed in the main storage device 12. In this processing, the information processing device 1 receives an input indicating that the optical axis adjustment of the electron beam has been adjusted (hereinafter, the input of the optical axis adjustment completed) as an example of a predetermined adjustment (S1). Instead of the input indicating that the optical axis adjustment has been completed, for example, a command from a user to determine the state of the electron microscope 3 after the optical axis adjustment may be input to the information processing device 1. When the information processing device 1 receives the input indicating that the optical axis adjustment has been completed, it acquires an electron microscope image of the sample 4 for confirmation (S2). The electron microscope image acquired in S2 is an example of a first confirmation image, and the processing in S2 is an example of acquiring the first confirmation image.

Then, the information processing device 1 inputs the electron microscope image of the sample 4 to the trained model A 12A and obtains the judgment result (S3). As described above, the trained model A 12A has learned the electron microscope image of the sample 4 when the optical axis adjustment is good and the electron microscope image of the sample 4 when the optical axis adjustment is not good. Moreover, the judgment result in S3 is an example of the first judgment result.

If the judgment result of the trained model A 12A indicates that the optical axis adjustment is good (OK), the information processing device 1 outputs to the display device 14 that the optical axis adjustment is good (S5). On the other hand, if the judgment result of the trained model A 12A indicates that the optical axis adjustment is not good (OK), the information processing device 1 outputs to the display device 14 an instruction to re-perform the optical axis adjustment (S6). The processes of S5 and S6 are an example of outputting a first judgment result.

(Effects of the embodiment)
As described above, the information system of this embodiment has the trained model A 12A that has trained the electron microscope image of the sample 4 when the optical axis adjustment is good and the electron microscope image of the sample 4 when the optical axis adjustment is not good. Then, this information system judges the electron microscope image of the sample 4 after the optical axis adjustment by the trained model A 12A. Conventionally, the quality of the optical axis adjustment is judged by the user, for example, by the change in relative brightness of the electron microscope image displayed on the display device 14, or the presence or absence of parallel movement of the electron microscope image on the screen of the display device 14 when adjusting the objective lens 35. As in this embodiment, by acquiring the judgment result by the trained model A 12A, a judgment that is not based on the subjectivity of the user can be obtained in a short time. Therefore, this information system can improve the functionality and usability for the electron microscope user or the service personnel in charge of maintenance management, especially in terms of maintenance management.

(Modification)
Fig. 9 illustrates the process of the information processing device 1 according to the first modified example of the present embodiment. In the process of Fig. 8, if the optical axis adjustment is not good (OK) in the judgment result of the trained model A 12A, the information processing device 1 outputs an instruction to re-execute the optical axis adjustment to the display device 14 and ends the process. However, the information processing device 1 may perform a further judgment in response to a user operation.

In FIG. 9, the processes from S1 to S5 are the same as those in FIG. 8, and therefore the description thereof will be omitted. If the judgment result of the trained model A 12A indicates that the optical axis adjustment is not good (OK), the information processing device 1 outputs to the display device 14 an instruction to the user to check the optical axis adjustment, and waits for the user's confirmation (S11). As described in FIG. 8, the judgment result of the trained model A 12A is an example of a first judgment result. Moreover, the process of S11 is an example of outputting an instruction to readjust the optical axis of the electron beam to the display device 14, which is an output device, when the first judgment result is not good.

The user then checks the optical axis adjustment again. For example, the user operates knobs 32A and 32C (or knobs 32B and 32D) as illustrated in FIG. 3 to check whether the electron microscope image is in the relatively brightest state. For example, the user also checks whether there is any parallel movement of the electron microscope image on the screen of the display device 14 when the current value of the objective lens 35 is adjusted, or that the movement is minimal. Through these procedures, the user can check that the optical axis has been adjusted to the appropriate position. If the user realizes that the optical axis has not been adjusted to the appropriate position, he or she can adjust the optical axis again using a specified procedure. In other words, checking the optical axis adjustment includes readjusting the optical axis.

Then, the information processing device 1 accepts the confirmation result of the optical axis adjustment. If the user's confirmation result indicates that the optical axis adjustment is not good (OK) (N in S12), the information processing device 1 outputs an error (S13). In this case, for example, the maintenance manager is notified of the error information and a request for maintenance work. This is because there is a possibility that a problem has occurred that cannot be solved by the user's adjustment.

On the other hand, if the user checks the optical axis adjustment and finds that the optical axis adjustment is good (OK) (Y in S12), the information processing device 1 again acquires an electron microscope image of the confirmation sample 4. Then, the information processing device 1 inputs the acquired electron microscope image of the sample 4 to the trained model B 12B and acquires a judgment result (S14). Here, acquiring an electron microscope image of the confirmation sample 4 again is an example of acquiring a second confirmation image taken from the sample 4, which is the confirmation sample, with the electron microscope 3 after the optical axis has been readjusted. Also, the process of S14 is an example of acquiring a second judgment result as to whether the contamination state of the aperture is below the standard.

Here, the trained model B 12B has learned an electron microscope image of the sample 4 when either the convergence aperture 34 or the objective aperture 36 is contaminated, and an electron microscope image of the sample 4 when neither the convergence aperture 34 nor the objective aperture 36 is contaminated. If the judgment result of the trained model B 12B is good (OK) (Y in S15), the information processing device 1 ends the process.

On the other hand, if the judgment result of the trained model B 12B is not good (OK) (N in S15), the information processing device 1 outputs to the display device 14 an instruction to have the user perform astigmatism correction. Then, the information processing device 1 waits for the input of the correction result from the user (S16). The astigmatism correction is performed by adjusting the current of the astigmatism correction coil 3A. Then, if it is input that the result of the astigmatism correction is good (OK) (Y in S17), the information processing device 1 ends the processing. Note that if the judgment result of the trained model B 12B is not good (OK) (N in S15), the information processing device 1 may immediately perform the processing of S18 without performing the processing of S16 and S17. In this case, the information processing device 1 outputs to the display device 14 an instruction to the user to replace at least one of the convergence aperture 34 and the objective aperture 36 and to correct the astigmatism after the replacement.

On the other hand, if it is input that the result of the astigmatism correction is not good (OK) (N in S17), the information processing device 1 outputs to the display device 14 an instruction to the user to replace at least one of the convergence aperture 34 and the objective aperture 36 and to correct the astigmatism after the replacement. Then, the information processing device 1 waits for the user to input the result of the astigmatism correction thereafter (S18). The result of the astigmatism correction is not good (OK) means that the astigmatism of the electron beam cannot be adjusted to the reference range by performing the astigmatism correction. In this case, at least one of the convergence aperture 34 and the objective aperture 36 is replaced. The user may replace all of the convergence aperture 34 and the objective aperture 36. Then, the user performs the astigmatism correction after the aperture replacement. If it is input that the result of the astigmatism correction after the aperture replacement is good (OK) (Y in S19), the information processing device 1 ends the process. At this time, the information processing device 1 may again input the electron microscope image of the sample 4 into the trained model B 12B to obtain the judgment result.

On the other hand, if it is input that the result of astigmatism correction after replacing all of the convergence aperture 34 and the objective aperture 36 is not good (OK) (N in S19), the information processing device 1 outputs an error (S20). In this case, for example, the maintenance manager is notified of the error information and a request for maintenance work.

According to the processing of the above modified example, if the judgment result of the trained model A 12A after the optical axis adjustment is not good (OK), the information processing device 1 outputs to the display device 14 an instruction to the user to check the optical axis adjustment again. As a result, the information processing device 1 judges the electron microscope image of the confirmation sample 4 by the trained model B 12B in a state where it is confirmed that the optical axis adjustment has been performed. Therefore, the information processing device 1 can more reliably judge the contamination of the convergence aperture 34 and the objective aperture 36, etc., after excluding the state where the optical axis is not adjusted. Furthermore, if the judgment result of the trained model B 12B is not good (OK) (N in S15), the information processing device 1 instructs the user to correct the astigmatism. Therefore, even if it is determined that there is a possibility of contamination of the convergence aperture 34 and the objective aperture 36, etc., the information processing device 1 first instructs the user to correct the astigmatism. As a result, even if there is a possibility of contamination of the convergence aperture 34 and the objective aperture 36, etc., if the problem can be dealt with by astigmatism correction, replacement of the convergence aperture 34 and the objective aperture 36, etc. is not necessary. On the other hand, if the result of the astigmatism correction is not good (OK), the information processing device 1 instructs replacement of at least one of the convergence aperture 34 and the objective aperture 36. In other words, when it is determined that the problem cannot be dealt with by astigmatism correction, the information processing device 1 instructs replacement of at least one of the convergence aperture 34 and the objective aperture 36. In this way, the information processing device 1 can support the user's maintenance management work by combining the judgments made by the trained model A 12A and the trained model B 12B for the electron microscope image of the confirmation sample 4 with the results of the user's maintenance management work for the electron microscope 3.

FIG. 10 illustrates the processing of the information processing device 1 according to the second modification of the present embodiment. In FIG. 8 (first modification) above, after the user reconfirms the optical axis adjustment, the information processing device 1 inputs the electron microscope image of the sample 4 to the trained model B 12B and determines whether or not astigmatism correction is required. However, the information processing device 1 may prompt the user to confirm instead of having the trained model B 12B make such a determination. That is, if the result of the reconfirmation of the optical axis adjustment by the user is good (OK) (Y in S12), the information processing device 1 again acquires an electron microscope image of the sample 4 for confirmation. Then, the information processing device 1 instructs the user to confirm the acquired electron microscope image of the sample 4 for confirmation (S14A). As described in FIG. 9, the electron microscope image of the sample 4 for confirmation acquired again is the second confirmation image. Therefore, the processing of S14A can be said to be an example of accepting a judgment on the image quality of the second confirmation image from the user.

If the user inputs that the confirmation result is good (OK) (Y in S15A), the information processing device 1 ends the process. On the other hand, if the user inputs that the confirmation result is not good (OK) (N in S15A), the information processing device 1 may execute the process following the instruction to correct astigmatism (S16), as in FIG. 9.

As described above, similar to variant example 1, the information processing device 1 combines the judgment made by the trained model A 12A for the electron microscope image of the confirmation sample 4 with the status of the maintenance and management work of the electron microscope 3 by the user. By combining these, the information processing device 1 can determine the status of the electron microscope 3 and assist the user in the maintenance and management work.

In the above embodiment, a scanning electron microscope is used as an example to illustrate the judgment of an image of the sample 4 captured by the trained model A 12A, the trained model B 12B, etc., and the process of assisting in optical axis adjustment, astigmatism correction, and aperture replacement. However, the implementation of the present invention is not limited to the case where the electron microscope 3 is a scanning electron microscope. That is, even if the electron microscope 3 is a transmission electron microscope, the information processing device 1 can perform the process of assisting the user, as in the above embodiment. In addition, when the electron microscope 3 is a transmission electron microscope, the image of the sample 4 may be, for example, an image from a CCD camera or CMOS camera that captures a fluorescent screen irradiated with transmitted electrons.

(Computer-readable recording medium)
A program for causing a computer or other machine or device (hereinafter, referred to as a computer, etc.) to realize any of the above functions can be recorded on a recording medium readable by the computer, etc. Then, the computer, etc. can provide the function by reading and executing the program from the recording medium.

Here, a computer-readable recording medium refers to a recording medium that stores information such as data and programs electrically, magnetically, optically, mechanically, or chemically and can be read by a computer. Among such recording media, those that can be removed from a computer include, for example, flexible disks, magneto-optical disks, CD-ROMs, CD-R/Ws, DVDs, Blu-ray disks, DAT, 8 mm tapes, and memory cards such as flash memory. In addition, examples of recording media that are fixed to a computer include hard disks and read only memories (ROMs). Furthermore, solid state drives (SDs) are also known.
(SSD) can be used as a recording medium that can be removed from a computer or the like, or as a recording medium that is fixed to a computer or the like.

(Other) The aspects of this embodiment are summarized below.
(Appendix 1)
acquiring a first confirmation image taken by the electron microscope from a confirmation sample for confirming a state of the electron microscope after a predetermined adjustment has been performed;
acquiring a first judgment result as to whether or not the adjustment result of the electron microscope after the predetermined adjustment is good by inputting the first confirmation image into a trained model that has been trained using, as teacher data, a first image taken of the confirmation sample by the electron microscope when the adjustment result by the predetermined adjustment is good and a second image taken of the confirmation sample by the electron microscope when the adjustment result by the predetermined adjustment is not good;
and outputting the first determination result to an output device.
(Appendix 2)
the predetermined adjustment is an adjustment of an optical axis of an electron beam,
the control unit outputs, when the first determination result is not good, an instruction to readjust the optical axis of the electron beam to the output device;
obtaining a second confirmation image taken of the confirmation sample with the electron microscope after the optical axis has been readjusted;
acquiring a second judgment result as to whether the contamination state of the aperture is below the standard by inputting the second confirmation image into a second trained model trained using as teacher data a third image taken from the confirmation sample by the electron microscope when the contamination states of all apertures provided in the electron microscope are below the standard and a fourth image taken from the confirmation sample by the electron microscope when the contamination state of any aperture provided in the electron microscope has deteriorated beyond the standard;
outputting, to the output device, an instruction to correct astigmatism of the electron beam when the second determination result is not good;
The information processing device described in Appendix 1 executes the following: when the astigmatism of the electron beam cannot be adjusted to a standard range by performing the astigmatism correction, outputting to the output device an instruction to replace at least one aperture provided in the electron microscope.
(Appendix 3)
the predetermined adjustment is an adjustment of an optical axis of an electron beam,
the control unit outputs, when the first determination result is not good, an instruction to readjust the optical axis of the electron beam to the output device;
obtaining a second confirmation image taken by the electron microscope from the confirmation sample after the optical axis has been readjusted;
receiving a judgment on the image quality of the second confirmation image from a user, and when the judgment that the image quality is insufficient is input, outputting an instruction to correct astigmatism of the electron beam to the output device;
The information processing device described in Appendix 1 executes the following: when the astigmatism of the electron beam cannot be adjusted to a standard range by performing the astigmatism correction, outputting to the output device an instruction to replace at least one aperture provided in the electron microscope.
(Appendix 4)
The computer acquires a first confirmation image taken by the electron microscope from a confirmation sample for confirming a state of the electron microscope after the predetermined adjustment is performed;
acquiring a first judgment result as to whether or not the adjustment result of the electron microscope after the predetermined adjustment is good by inputting the first confirmation image into a trained model that has been trained using, as teacher data, a first image taken of the confirmation sample by the electron microscope when the adjustment result by the predetermined adjustment is good and a second image taken of the confirmation sample by the electron microscope when the adjustment result by the predetermined adjustment is not good;
outputting the first determination result to an output device.
(Appendix 5)
the predetermined adjustment is an adjustment of an optical axis of an electron beam,
the computer outputs, when the first determination result is not good, an instruction to readjust the optical axis of the electron beam to the output device;
obtaining a second confirmation image taken of the confirmation sample with the electron microscope after the optical axis has been readjusted;
acquiring a second judgment result as to whether the contamination state of the aperture is below the standard by inputting the second confirmation image into a second trained model trained using as teacher data a third image taken from the confirmation sample by the electron microscope when the contamination states of all apertures provided in the electron microscope are below the standard and a fourth image taken from the confirmation sample by the electron microscope when the contamination state of any aperture provided in the electron microscope has deteriorated beyond the standard;
outputting, to the output device, an instruction to correct astigmatism of the electron beam when the second determination result is not good;
An information processing method as described in Appendix 4, which executes the following: when the astigmatism of the electron beam cannot be adjusted to a standard range by performing the astigmatism correction, outputting to the output device an instruction to replace at least one aperture provided in the electron microscope.
(Appendix 6)
the predetermined adjustment is an adjustment of an optical axis of an electron beam,
the computer outputs, when the first determination result is not good, an instruction to readjust the optical axis of the electron beam to the output device;
obtaining a second confirmation image taken by the electron microscope from the confirmation sample after the optical axis has been readjusted;
receiving a judgment on the image quality of the second confirmation image from a user, and when the judgment that the image quality is insufficient is input, outputting an instruction to correct astigmatism of the electron beam to the output device;
An information processing method as described in Appendix 4, which executes the following: when the astigmatism of the electron beam cannot be adjusted to a standard range by performing the astigmatism correction, outputting to the output device an instruction to replace at least one aperture provided in the electron microscope.
(Appendix 7)
acquiring, in a computer, a first confirmation image taken by the electron microscope from a confirmation sample for confirming a state of the electron microscope after a predetermined adjustment has been made;
acquiring a first judgment result as to whether or not the adjustment result of the electron microscope after the predetermined adjustment is good by inputting the first confirmation image into a trained model that has been trained using, as teacher data, a first image taken of the confirmation sample by the electron microscope when the adjustment result by the predetermined adjustment is good and a second image taken of the confirmation sample by the electron microscope when the adjustment result by the predetermined adjustment is not good;
outputting the first determination result to an output device.
(Appendix 8)
the predetermined adjustment is an adjustment of an optical axis of an electron beam,
outputting, to the output device, an instruction to the computer to readjust the optical axis of the electron beam when the first judgment result is not good;
obtaining a second confirmation image taken of the confirmation sample with the electron microscope after the optical axis has been readjusted;
acquiring a second judgment result as to whether the contamination state of the aperture is below the standard by inputting the second confirmation image into a second trained model trained using as teacher data a third image taken from the confirmation sample by the electron microscope when the contamination states of all apertures provided in the electron microscope are below the standard and a fourth image taken from the confirmation sample by the electron microscope when the contamination state of any aperture provided in the electron microscope has deteriorated beyond the standard;
outputting, to the output device, an instruction to correct astigmatism of the electron beam when the second determination result is not good;
The program described in Appendix 7 for executing the program is to output to the output device an instruction to replace at least one aperture provided in the electron microscope when the astigmatism of the electron beam cannot be adjusted to a standard range by performing the astigmatism correction.
(Appendix 9)
the predetermined adjustment is an adjustment of an optical axis of an electron beam,
outputting, to the output device, an instruction to the computer to readjust the optical axis of the electron beam when the first judgment result is not good;
obtaining a second confirmation image taken by the electron microscope from the confirmation sample after the optical axis has been readjusted;
receiving a judgment on the image quality of the second confirmation image from a user, and when the judgment that the image quality is insufficient is input, outputting an instruction to correct astigmatism of the electron beam to the output device;
The program described in Appendix 7 for executing the program is to output to the output device an instruction to replace at least one aperture provided in the electron microscope when the astigmatism of the electron beam cannot be adjusted to a standard range by performing the astigmatism correction.

REFERENCE SIGNS LIST 1 Information processing device 3 Electron microscope 4 Sample 31 Control circuit 32 Electron gun 33 Converging lens 34 Converging aperture 35 Objective lens 36 Objective aperture 39 Stage 3A Astigmatism correction coil

Claims (3)

acquiring a confirmation image taken by the electron microscope from a confirmation sample for confirming a state of the electron microscope after the optical axis of the electron beam has been adjusted;
acquiring a judgment result as to whether or not the contamination state of the aperture is below a standard by inputting the confirmation image into a trained model trained using, as training data, a first image taken of the confirmation sample by the electron microscope when the contamination state of all apertures provided in the electron microscope is below a standard and a second image taken of the confirmation sample by the electron microscope when the contamination state of any aperture provided in the electron microscope has deteriorated beyond the standard;
and outputting to an output device an instruction to correct the astigmatism of the electron beam or an instruction to replace at least one aperture provided in the electron microscope when the judgment result is not good. a computer acquires a confirmation image taken by the electron microscope from a confirmation sample for confirming a state of the electron microscope after the optical axis of the electron beam has been adjusted;
acquiring a judgment result as to whether or not the contamination state of the aperture is below a standard by inputting the confirmation image into a trained model trained using, as training data, a first image taken of the confirmation sample by the electron microscope when the contamination state of all apertures provided in the electron microscope is below a standard and a second image taken of the confirmation sample by the electron microscope when the contamination state of any aperture provided in the electron microscope has deteriorated beyond the standard;
and outputting to an output device an instruction to correct the astigmatism of the electron beam or an instruction to replace at least one aperture provided in the electron microscope when the judgment result is not good. acquiring, in a computer, a confirmation image taken by the electron microscope from a confirmation sample for confirming a state of the electron microscope after the optical axis of the electron beam has been adjusted;
acquiring a judgment result as to whether or not the contamination state of the aperture is below a standard by inputting the confirmation image into a trained model trained using, as training data, a first image taken of the confirmation sample by the electron microscope when the contamination state of all apertures provided in the electron microscope is below a standard and a second image taken of the confirmation sample by the electron microscope when the contamination state of any aperture provided in the electron microscope has deteriorated beyond the standard;
and outputting to an output device an instruction to correct the astigmatism of the electron beam or an instruction to replace at least one aperture provided in the electron microscope when the judgment result is not good.

Images