Control method for automated microscope system, microscope system and computer-readable storage medium

1. A method for controlling an autonomous microscope system, comprising

the steps of:

obtaining a low magnification image with a device;

inputting the low magnification image to a first neural network to select a region of interest, wherein the first neural network is trained in a reinforcement learning scheme;

magnifying the region of interest to generate a high magnification image;

inputting the high magnification image to a second neural network to analyze whether the high magnification image comprises a target feature and generate a statistical result relevant to the target feature; and

generating a feedback signal according to the statistical result and sending the feedback signal to the first neural network to train the first neural network in the reinforcement learning scheme.

2. The method for controlling an autonomous microscope system according to claim 1 , wherein the first neural network is a first convolutional neural network or a fully connected neural network.

3. The method for controlling an autonomous microscope system according to claim 1 , wherein the step of inputting the low magnification image to the first neural network to select the region of interest further comprises partitioning the low magnification image into a plurality of regions.

4. The method for controlling an autonomous microscope system according to claim 3 , wherein the first neural network is the first convolutional neural network or a fully connected neural network; wherein the plurality of regions are inputted to the first convolutional neural network to generate a probability distribution model; and

wherein the probability distribution model represents the probability of any one of the plurality of regions to be the region of interest.

5. The method for controlling an autonomous microscope system according to claim 3 , wherein when the first neural network finds that any one of the plurality of regions is the region of interest, a positive feedback signal is generated and sent to the first neural network to train the first neural network in the reinforcement learning scheme.

6. The method for controlling an autonomous microscope system according to claim 3 , wherein when the first neural network selects any one of the plurality of regions as the region of interest, a negative feedback signal is generated and sent to the first neural network every specific time period to train the first neural network in the reinforcement learning scheme.

7. The method for controlling an autonomous microscope system according to claim 1 , wherein the first neural network is further trained by a supervised learning algorithm, an unsupervised learning algorithm, an imitation learning algorithm, or combinations thereof.

8. The method for controlling an autonomous microscope system according to claim 1 , wherein the method further comprises: determining whether the statistical result satisfies an overall goal.

9. The method for controlling an autonomous microscope system according to claim 8 , wherein when the statistical result does not satisfy the overall goal, a negative feedback signal is generated and sent to the first neural network to train the first neural network in the reinforcement learning scheme.

10. The method for controlling an autonomous microscope system according to claim 1 , wherein when the high magnification image does not comprise the target feature, a negative feedback signal is generated and sent to the first neural network to train the first neural network in the reinforcement learning scheme.

11. The method for controlling an autonomous microscope system according to claim 1 , wherein the second neural network is a second convolutional neural network or a fully connected neural network.

12. The method for controlling an autonomous microscope system according to claim 1 , wherein the second neural network is arranged to be an instance segmentation model, a semantic segmentation model or an image classification model to analyze whether the high magnification image comprises the target feature.

13. A microscope system, characterized by comprising a processor performing a method for controlling an autonomous microscope system comprising:

obtaining a low magnification image with a device;

inputting the low magnification image to a first neural network to select a region of interest, wherein the first neural network is trained in a reinforcement learning scheme;

magnifying the region of interest to generate a high magnification image;

inputting the high magnification image to a second neural network to analyze whether the high magnification image comprises a target feature and generate a statistical result relevant to the target feature; and

generating a feedback signal according to the statistical result and sending the feedback signal to the first neural network to train the first neural network in the reinforcement learning scheme.

14. The microscope system according to claim 13 , further comprising an optical unit, a stage, an electronic control unit, a storage unit, or combinations thereof.

15. A computer readable storage medium with a program stored therein, wherein when the program is loaded to a computer, the computer performs a method for controlling an autonomous microscope system comprising:

obtaining a low magnification image with a device;

inputting the low magnification image to a first neural network to select a region of interest, wherein the first neural network is trained in a reinforcement learning scheme;

magnifying the region of interest to generate a high magnification image;

inputting the high magnification image to a second neural network to analyze whether the high magnification image comprises a target feature and generate a statistical result relevant to the target feature; and

generating a feedback signal according to the statistical result and sending the feedback signal to the first neural network to train the first neural network in the reinforcement learning scheme.