Method, computer program product, computer-readable medium and system for scanning partial regions of a sample using a scanning microscope

16:

• A method for capturing partial regions of a sample by a microscope the method comprising:
• determining the partial regions of the sample which are captured by the microscope by a machine learning system, wherein the machine learning system is trained by supervised learning, unsupervised learning, and/or reinforcement learning for improved determination of the partial regions of the sample which are captured by the microscope;
• and
• capturing the determined partial regions of the sample by the microscope.

18:

• The method as claimed in claim 16, wherein the method further comprises training the machine learning system by reinforcement learning by:
• inputting one or more figures of merit into the machine learning system;
• and
• varying the partial regions of the sample which are captured and/or the order in which the partial regions of the sample are captured in order to achieve a figure of merit that is as optimal as possible or figures of merit that are as optimal as possible.

22:

• The method as claimed in claim 16, wherein the method further comprises training the machine learning system by unsupervised learning by:
• inputting non-reconstructed overall images of samples into the machine learning system;
• and
• determining the partial regions of the sample which have a particularly high information density on the basis of the respective non-reconstructed overall image by an autoencoder for improved determination of the partial regions of the sample which are captured by the microscope.

23:

• A method for capturing partial regions of a sample by a microscope and for reconstructing an overall image of the sample from data of the captured partial regions of the sample,
• the method comprising:
• determining the partial regions of the sample which are captured by the microscope;
• capturing the partial regions of the sample by the microscope;
• inputting the data of the captured partial regions into a machine learning system;
• and
• reconstructing the overall image of the sample from the data of the captured partial regions of the sample by the machine learning system,
• wherein non-captured partial regions of the sample are estimated of the data of the captured partial regions of the sample by the machine learning system,
• wherein the machine learning system is trained by supervised learning, unsupervised learning, and/or a reinforcement learning for improved reconstruction of the reconstructed overall image of the sample from the data of the captured partial regions of the sample.

24:

• The method as claimed in claim 23, wherein the method further comprises training the machine learning system for improved reconstruction of the reconstructed overall image from the data of the captured partial regions of the sample by supervised learning by:
• inputting data of captured partial regions, of the sample as training data into the machine learning system;
• and
• comparing the overall image of the sample that is reconstructed from the training data by the machine learning system with anon reconstructed overall image of the sample for training the machine learning system for improved reconstruction of the overall image from the data of the captured partial regions of the sample.

30:

• A system for capturing partial regions of a sample by a microscope, and for reconstructing an overall image of the sample from data of the captured partial regions of the sample, wherein the system comprises:
• a machine learning system which, by supervised learning, unsupervised learning, and/or reinforcement learning, is trained to carry out the following:
• reconstructing the overall image of the sample from the data of the captured partial regions of the sample by the machine learning system,
• wherein non-captured partial regions of the sample are estimated by the data of the captured partial regions of the sample by the machine learning system.