CN109785940B

CN109785940B - Method for typesetting medical image film

Info

Publication number: CN109785940B
Application number: CN201811625185.XA
Authority: CN
Inventors: 毛青; 蔡伟华; 李丹; 彭笳鑫; 熊博
Original assignee: Suzhou Chengze Medical Technology Co ltd
Current assignee: Suzhou Chengze Medical Technology Co ltd
Priority date: 2018-12-28
Filing date: 2018-12-28
Publication date: 2023-02-03
Anticipated expiration: 2038-12-28
Also published as: CN109785940A

Abstract

The invention discloses a method for typesetting a medical image film, which comprises the following steps: acquiring image sequence information required to be printed by images of various inspection types; acquiring row and column parameters of a film to be printed; inputting an image and reading sequence information of the image; calculating the number of films to be printed; screening the images to be printed in each sequence, thereby picking out the images on each line of the film; film content is generated for printing. The invention can automatically select and typeset the medical images for doctors or technicians to use for film printing, and can quickly and effectively process even if the number of the images is large, thereby greatly reducing the workload of the workers.

Description

Method for typesetting medical image film

Technical Field

The invention relates to a medical image film typesetting method.

Background

At present, the traditional film typesetting mode of medical service institutions is to manually select images by doctors or technicians and then perform typesetting for printing. However, the number of images for examinations such as CT and nuclear magnetic examinations is large, and therefore, a large workload is imposed on the doctor.

Disclosure of Invention

The invention provides a method for typesetting a medical image film.

The technical scheme of the invention is as follows: a method for film typesetting of medical images comprises the following steps: 1) Acquiring image sequence information required to be printed by images of various inspection types; 2) Acquiring row and column parameters of a film to be printed, wherein the row and column parameters are M rows and N columns respectively; 3) Inputting an image and reading sequence information of the image; 4) Calculating the number of films to be printed, and dividing into a step 4 a) and a step 4 b): 4a) Setting configuration conditions, and screening out the number m of image sequences meeting the printing conditions; 4b) Dividing M by M, if the remainder is 0, the quotient is the number of the films, otherwise, the quotient plus 1 is the number of the films; 5) Screening the images to be printed for each sequence, thereby picking out the images on each line of film, comprising steps 5 a) to 5 g): 5a) Cycling all the sequences of the images in sequence; 5b) Calculating whether the number of images in the sequence is larger than the number of columns of the film, if so, turning to the step 5 c), otherwise, turning to the step 5 d); 5c) Calculating the similarity of different pictures through an algorithm to select N images with the maximum difference among the images, and correspondingly filling the N images into N columns of the film; 5d) Selecting all images of the sequence and filling the images into corresponding columns of the film; 5e) Confirming whether all sequences are circularly completed or not, if so, turning to the step 5 f), and otherwise, turning to the step 5 g); 5f) Completing the film typesetting layout; 5g) Turning to step 5 a), continuously circulating all the image sequences in sequence; 6) Film content is generated for printing.

Further, in the present invention, the image acquired in step 1) is selected from CT examination, MR examination, DR examination, CR examination, and MG examination.

Further, in the present invention, the configuration condition in step 4 a) is a white list, and the white list lists names of image sequences to be printed.

Further, in the invention, the white list supports fuzzy matching.

Further, in the present invention, the algorithm adopted in step 5 c) is based on an unsupervised machine learning algorithm, and the algorithm includes: A. inputting medical images, reading data of each image, and performing normalization processing; B. setting the size of an initial K value, and dividing the K value into N types; C. and respectively training the pictures of different categories, and respectively taking out the top picture of different categories.

Compared with the prior art, the invention has the advantages that: the method for typesetting the medical image film can automatically select and typeset the medical image for doctors or technicians to use for film printing, can quickly and effectively process even if the number of the images is large, and greatly reduces the workload of workers.

Drawings

The invention is further described with reference to the following figures and examples:

FIG. 1 is a schematic flow diagram of the process of the present invention.

Detailed Description

Example (b):

the invention discloses a method for typesetting medical image films, which comprises the following steps:

step 1): and acquiring image sequence information required to be printed by images of various inspection types, wherein the acquired images are from CT inspection, MR inspection, DR inspection, CR inspection and MG inspection.

Step 2): acquiring row and column parameters of films to be printed for various inspection type images, wherein the row parameters are M lines and N lines;

and step 3): inputting an image and reading sequence information of the image;

and step 4): calculating the number of films to be printed, and dividing into a step 4 a) and a step 4 b):

step 4 a): setting configuration conditions, and screening out the number m of image sequences meeting the printing conditions, wherein the configuration conditions are a white list, the white list lists names of the image sequences needing to be printed, and the white list supports fuzzy matching;

the sequence name is as follows: t2WI _ TRA, T1WI _ TRA, FSFLAIR _ TRA, lung, stan. . .

The white list is to support fuzzy matching: for example, T1WI, image sequence information of a picture to be printed that meets the requirements can be screened out.

Step 4 b): dividing M by M, if the remainder is 0, the quotient is the number of the films, otherwise, the quotient plus 1 is the number of the films;

step 5): screening the images to be printed for each sequence, thereby picking out the images on each line of film, comprising steps 5 a) to 5 g):

step 5 a): cycling all the sequences of the images in sequence;

step 5 b): calculating whether the number of images in the sequence is larger than the number of columns of the film, if so, turning to the step 5 c), otherwise, turning to the step 5 d);

step 5 c): calculating the similarity of different pictures by adopting an unsupervised machine learning algorithm, selecting N images with the largest difference between the images, and correspondingly filling the N images into N columns of the film, wherein the algorithm comprises the following steps:

A. inputting medical images, reading data of each image, and performing normalization processing;

B. setting the size of an initial K value, and dividing the K value into N types;

C. and respectively training the pictures of different categories, and respectively taking out the top picture of different categories.

Step 5 d): selecting all images of the sequence and filling the images into corresponding columns of the film;

step 5 e): confirming whether all sequences are circularly completed or not, if so, turning to the step 5 f), and otherwise, turning to the step 5 g);

step 5 f): completing the film typesetting layout;

step 5 g): turning to step 5 a), continuously circulating all the image sequences in sequence;

step 6): film content is generated for printing.

In the specific operation of this embodiment, taking a 4-row and 3-column film as an example, if the number of the sequence of the input images is 5, such as (1-1, \8230;, 1-a), (2-1, \8230; 2-b), (3-1, \8230; 3-c), (4-1, \8230; 4-d), (5-1, \8230; 5-e), the remainder of dividing 5 by 4 is 1, and the quotient is 1, the number of the film is 2, and if the number of the sequence of the input images is 12, the remainder of dividing 12 by 4 is 0, and the quotient is 3, the number of the film is 3; still taking the number of the sequences of the input images as 5 as an example, (1-1, \8230;,/1-a) - (4-1, \8230; 4-d) are sequentially filled in the first to fourth rows of a first film, (5-1, \8230; 5-e) are filled in the first row of a second film, when selecting the images which need to be printed in each sequence, whether the number of the images of the sequence is greater than the number of the columns of the films is calculated, taking the first sequence as an example, if a is 2, the number is less than the number of the columns of the films 3, and all 2 images are filled in the corresponding column, if a is 6, the number is greater than the number of the columns of the films 3, and 3 images with the largest difference are selected from 6 images by adopting a unsupervised machine learning algorithm and are filled in the 3 columns of the films; and after all the sequences are circularly completed, the film typesetting layout is completed, and the film content is generated for printing.

By using the method for typesetting the medical image film, the medical image can be automatically selected and typeset for doctors or technicians to use for film printing, and even if the number of images is large, the images can be quickly and effectively processed, thereby greatly reducing the workload of workers.

It should be understood that the above-mentioned embodiments are only illustrative of the technical concepts and features of the present invention, and are intended to enable those skilled in the art to understand the contents of the present invention and implement the present invention, and not to limit the scope of the present invention. All modifications made in accordance with the spirit of the main technical scheme of the invention are intended to be covered by the scope of the invention.

Claims

1. A method for typesetting medical image films is characterized by comprising the following steps:

1) Acquiring image sequence information required to be printed by images of various inspection types;

2) Acquiring row and column parameters of a film to be printed, wherein the row and column parameters are M rows and N columns respectively;

3) Inputting an image and reading sequence information of the image;

4) Calculating the number of films to be printed, and dividing into a step 4 a) and a step 4 b):

4a) Setting configuration conditions, and screening out the number m of image sequences meeting the printing conditions;

4b) Dividing M by M, if the remainder is 0, the quotient is the number of the films, otherwise, the quotient plus 1 is the number of the films;

5) Screening the images to be printed for each sequence, thereby picking out the images on each line of film, comprising steps 5 a) to 5 g):

5a) Sequentially circulating all the sequences of the images;

5b) Calculating whether the number of the images of the sequence is larger than the number of the columns of the film, if so, turning to the step 5 c), otherwise, turning to the step 5 d);

5c) Calculating the similarity of different pictures through an algorithm to select N images with the maximum difference among the images, and correspondingly filling the N images into N columns of the film;

5d) Selecting all images of the sequence and filling the images into corresponding columns of the film;

5e) Confirming whether all sequences are circularly completed, if so, turning to the step 5 f), and otherwise, turning to the step 5 g);

5f) Completing the film typesetting layout;

5g) Turning to step 5 a), continuously circulating all the image sequences in sequence;

6) Film content is generated for printing.

2. The method for film layout of medical images according to claim 1, wherein the images obtained in step 1) are from CT examination, MR examination, DR examination, CR examination, MG examination.

3. The method for film layout of medical images according to claim 1, wherein the configuration condition in step 4 a) is a white list listing names of image sequences to be printed.

4. The method of claim 3, wherein said whitelist supports fuzzy matching.

5. The method for film layout of medical images according to claim 1, wherein the algorithm used in step 5 c) is based on unsupervised machine learning algorithm, and the algorithm steps comprise:

C. and respectively training the pictures of different classes, and respectively taking out the top picture of different classes.