Method for detecting the presence of pneumonia area in medical images of patients, detecting system, and electronic device employing method

1. A pneumonia area detecting method comprising:

obtaining medical images;

processing the medical images to obtain a standard format image;

obtaining a frame size of an initial pneumonia area;

processing the frame size of the initial pneumonia area through a preset algorithm and the standard format image, to obtain a frame size of a standard format pneumonia area; and

inputting the standard format image and the frame size of the standard format pneumonia area to a pneumonia area detecting model, extracting features of the standard format image to obtain pneumonia area features of the standard format image, and obtaining an output pneumonia area through the frame size of the standard format pneumonia area and the pneumonia area features of the standard format image;

wherein the pneumonia area detecting model detects the pneumonia area through a bidirectional feature pyramid network, the bidirectional feature pyramid network comprises a AdamW learning efficiency optimizer, and the AdamW learning efficiency optimizer learns features of the pneumonia area through weight decay.

2. The pneumonia area detecting method of claim 1 , further comprising:

calculating a size average value and a standard deviation of the medical images;

outputting the standard format image according to the size average value and the standard deviation of the medical images.

3. The pneumonia area detecting method of claim 1 , further comprising:

optimizing the pneumonia area detecting model through a category weight coefficient and a loss function.

4. The pneumonia area detecting method of claim 1 , further comprising:

obtaining a standard pneumonia area; and

verifying the output pneumonia area and the standard pneumonia area to obtain an intersection over union between the output pneumonia area and the standard pneumonia area.

5. The pneumonia area detecting method of claim 4 , further comprising:

outputting the output pneumonia area when the intersection over union is greater than a preset value.

6. A pneumonia area detecting system comprising:

an information obtaining module obtaining medical images and a frame size of an initial pneumonia area;

image preprocessing module processing the medical images to obtain a standard format image;

a calculating module processing the frame size of the initial pneumonia area through a preset algorithm and the standard format image, to obtain a frame size of a standard format pneumonia area; and

a pneumonia area detecting model obtaining an output pneumonia area according to the standard format image and the frame size of the standard format pneumonia area;

wherein the pneumonia area detecting model is configurated to extract features of the standard format image to obtain pneumonia area features of the standard format image, and obtain the output pneumonia area through the frame size of the standard format pneumonia area and the pneumonia area features of the standard format image;

wherein the pneumonia area detecting model detects the pneumonia area through a bidirectional feature pyramid network, the bidirectional feature pyramid network comprises a AdamW learning efficiency optimizer, and the AdamW learning efficiency optimizer learns features of the pneumonia area through weight decay.

7. The pneumonia area detecting system of claim 6 , wherein the image preprocessing module is configured to calculate a size average value and a standard deviation of the medical images, and output the standard format image according to the size average value and the standard deviation of the medical images.

8. The pneumonia area detecting system of claim 6 , wherein the calculating module is configurated to optimize the pneumonia area detecting model through a category weight coefficient and a loss function.

9. The pneumonia area detecting system of claim 6 , further comprising a model verifying module, the model verifying module obtaining a standard pneumonia area, verifying the output pneumonia area and the standard pneumonia area to obtain an intersection over union between the output pneumonia area and the standard pneumonia area.

10. The pneumonia area detecting system of claim 9 , wherein the pneumonia area detecting model outputs the output pneumonia area when the intersection over union is greater than a preset value.

11. An electronic device, comprising:

a storage device; and

at least one processor, wherein the storage device stores one or more programs, when executed by the at least one processor, the one or more programs cause the at least one processor to:

obtain medical images and a frame size of an initial pneumonia area;

process the medical images to obtain a standard format image;

process the frame size of the initial pneumonia area through a preset algorithm and the standard format image, to obtain a frame size of a standard format pneumonia area; and

obtain an output pneumonia area according to the standard format image and the frame size of the standard format pneumonia area;

extract features of the standard format image to obtain pneumonia area features of the standard format image, and obtain the output pneumonia area through the frame size of the standard format pneumonia area and the pneumonia area features of the standard format image;

detect the pneumonia area through a bidirectional feature pyramid network, wherein the bidirectional feature pyramid network comprises a AdamW learning efficiency optimizer, and the AdamW learning efficiency optimizer learns features of the pneumonia area through weight decay.

12. The electronic device according to claim 11 , wherein the at least one processor is further caused to:

calculate a size average value and a standard deviation of the medical images;

output the standard format image according to the size average value and the standard deviation of the medical images.

13. The electronic device according to claim 11 , wherein the at least one processor is further caused to:

optimize a pneumonia area detecting model through a category weight coefficient and a loss function.

14. The electronic device according to claim 11 , wherein the at least one processor is further caused to:

obtain a standard pneumonia area; and

verify the output pneumonia area and the standard pneumonia area to obtain an intersection over union between the output pneumonia area and the standard pneumonia area.

15. The electronic device according to claim 14 , wherein the at least one processor is further caused to:

output the output pneumonia area when the intersection over union is greater than a preset value.