Automatic structured analysis method for bull's eye images of echocardiography
By using segment feature information and grading scoring standards in echocardiography, the dynamic correlation between the bull eye diagram and the structured report content is solved, and the problem of inconsistency between the bull eye diagram and the report content in the existing technology is improved, and the analysis results are improved. Accuracy and consistency.
Patent Information
- Application Number
- CN202111551269.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-12-17
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2041-12-17
AI Technical Summary
In existing structured echocardiography reports, the content of the bull eye chart and report are usually produced by different people, resulting in inconsistent graphics and texts, and the synchronous modification cannot be achieved, making it difficult to guarantee the accuracy and consistency of the analysis results.
An echocardiography structured automatic auxiliary analysis method is designed. By synchronously outputting the structured report data content when generating the bull eye diagram, segment feature information, grading scoring standards and regional encoding are used to realize bull eye diagram and Dynamic correlation of report content ensures synchronous modification and accurate correspondence of pictures and texts.
It realizes the dynamic correlation between the bull eye diagram and the content of the structured report, which is consistent with the user's understanding habits to the greatest extent, ensures synchronous modification of the graphics and text, improves the accuracy and consistency of the analysis results, and meets the synchronous modification needs of medical reports.
Smart Images

Figure CN114420259B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to a medical information processing method, in particular to an ultrasonic cardiogram bull's eye diagram structured automatic auxiliary analysis method. Background Art
[0002] In order to intuitively show the state of the heart's ventricular wall movement in echocardiograms, existing medical reports use a bull's eye chart combined with a structured report. The bull's eye chart divides the heart's ventricular wall into several segments, and uses scores and colors on the bull's eye chart to mark the segment states. However, in the currently provided echocardiogram structured reports, the bull's eye chart and the report content are produced separately, which sometimes causes inconsistencies or even contradictions between the graphics and text. This is because the bull's eye chart content and the report content may be produced by different people, and the bull's eye chart or the report content cannot be modified synchronously. Summary of the invention
[0003] The purpose of the present invention is to solve the above-mentioned deficiencies of the prior art and to provide a structured automatic auxiliary analysis method of an echocardiogram bull's eye image that realizes dynamic association between the bull's eye image and report content.
[0004] In order to achieve the above-mentioned purpose, the present invention is a method for automatically assisting in the structured analysis of the bull's eye image of an echocardiogram, which can synchronously output structured report data content when generating the bull's eye image, and comprises the following steps:
[0005] a. Assign corresponding segment feature information to each segment on the bull's eye diagram, and the segment feature information includes segment name, segment number, and segment code;
[0006] b. According to the actual structure of the heart and the medical standards for ultrasound examination, the segments are combined to form a local area with a range larger than a single segment, and the local area is given a regional name based on the segment name and a regional code composed of a combination of segment codes;
[0007] c. Establish a grading standard, specify the score value and the score name that describes its meaning;
[0008] d. The operator assigns corresponding scoring values to each segment in step a on the computer according to the grading criteria in step c;
[0009] e. The computer divides the segments with the same score value into a group according to the score value provided by the operator in step d, and groups all the segments according to the score value, and then performs iterative matching, wherein the iterative matching includes:
[0010] e.1 Take a group of segment codes of segments with the same score value and combine them into a code group, match the code group with the area code in step b, if the code group and the area code are consistent, then at least obtain the area name corresponding to the matched area code and the score value of the segment corresponding to the code group, and output them in the form of a line of records;
[0011] e.2 If no consistent result is obtained after matching the coding group with the regional code, then according to the segment code in the coding group, at least the corresponding segment name and the score value of the segment corresponding to the coding group are obtained and output in the form of a line of records;
[0012] e.3 After completing the matching of the coding groups under a scoring value group, take the next group of segments with the same scoring value and repeat steps e.1 to e.3 until all segments are matched;
[0013] f. Integrate the records output after iterative matching as structured report data content.
[0014] To further facilitate the generation of structured reports, after completing step f, you can also directly obtain the structured report text through the following steps:
[0015] g. Establish a vocabulary containing descriptive words, and correspond the descriptive words to the area name or segment name and the scoring value. After receiving the structured report data content, the computer fills in the corresponding descriptive words at the specified position of the structured report text; when the area name, segment name and scoring name are all medical terms, the descriptive words can be directly a combination of the area name or segment name and the scoring name that expresses the meaning of the scoring value, and better expression effects can be achieved by adding punctuation, slashes or spaces.
[0016] In order to improve data processing efficiency, the hierarchical scoring standard of step c at least includes three types of scoring values and corresponding scoring names, namely normal motion state, no score displayed, and abnormal motion state. In the iterative matching of step e, only the group whose scoring value means abnormal motion state is iteratively matched. It should be noted that there can be multiple scoring values for the abnormal motion state category. This is because in actual medical scenarios, the abnormal cardiac motion state is diverse. Therefore, it is necessary to give multiple scoring values indicating abnormal motion state in the hierarchical scoring standard based on medical practice.
[0017] In order to facilitate the use of operators and adapt to the habits of different operators, the present invention also provides a method for synchronously drawing a bull's eye chart when modifying the structured report content, comprising the following steps:
[0018] h. In step b, the conclusion codes corresponding to each segment and local area are assigned in the computer, and the conclusion codes are also included in the structured report data content. In step c, each score value of the graded scoring standard also has a corresponding color parameter;
[0019] i. In the structured report editing interface, the operator gives a score value for the segment or local area according to the grading criteria in step c;
[0020] j. After the data content of the structured report is edited and changed in step i, the computer automatically obtains the changed multiple lines of records and performs reverse matching line by line. The reverse matching includes:
[0021] j.1. Obtain the conclusion code in the current record and find the region code or segment code corresponding to the conclusion code;
[0022] j.2. Find the corresponding segment number by finding the area code or segment code;
[0023] j.3. Associate the score value in the current record with the segment number;
[0024] j.4 Repeat the above reverse matching operation for the next row of records until the traversal of multiple rows of records is completed, and finally the corresponding relationship between the score value given in step i and the segment number of each segment on the bull's eye chart is obtained
[0025] k. The computer obtains the correspondence between the segment numbers and the score values and the correspondence between the score values and the color parameters, and colors the bull's eye chart accordingly.
[0026] In order to make the scoring operation more intuitive, the grading standard also includes a bull's eye chart color parameter corresponding to each score value; when the operator scores each segment of the bull's eye chart in step d, the score value can be given by providing color information.
[0027] The structured automatic auxiliary analysis method of the bull's eye image of the echocardiogram obtained by the present invention associates the graphic content of the bull's eye image with the structured report content by encoding the segmental motion evaluation index. At the same time, by setting the local area and the area code, the correct expression of the local ventricular wall range expressed by the bull's eye image is achieved, which conforms to the user's understanding habits to the greatest extent and achieves the effect of synchronous modification of image and text and accurate correspondence between image and text. BRIEF DESCRIPTION OF THE DRAWINGS
[0028] Figure 1 This is a flow chart of a method for synchronously outputting structured report data content when generating a bull's eye image in Example 1 of the structured automatic auxiliary analysis method for an echocardiogram bull's eye image of the present invention;
[0029] Figure 2is a schematic diagram of the structure of the bull's eye diagram in Example 1;
[0030] Figure 3 This is a flow chart of synchronously drawing a bull's eye chart when modifying the structured newspaper content in Example 1. DETAILED DESCRIPTION
[0031] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation mode, structure, characteristics and effects of the present invention are described in detail below in combination with the accompanying drawings and preferred embodiments.
[0032] Embodiment 1:
[0033] The method for automatically assisting the structured analysis of the bull's eye image of an echocardiogram described in this embodiment is executed on a computer software system. The operator may be a medical worker or a trained neural network model, such as Figure 1 , Figure 2 As shown in the figure, it can output structured report data content synchronously when generating a bull's eye chart, and includes the following steps:
[0034] a. Assign corresponding segment characteristic information to each segment on the bull's eye chart, which includes segment name, segment number, and segment code; the segment name and segment number need to comply with existing medical standards, such as Figure 2 , this embodiment uses a 17-segment bull's eye chart. In the current medical standard, the segment number of the bull's eye chart is ( Figure 2 The numbers on the bull's eye chart) and the segment name and position on the bull's eye chart are defined, while the segment code can be assigned by the operator or computer software.
[0035] b. According to the actual structure of the heart and medical standards for ultrasound examination, the segments are combined to form a local area with a range larger than a single segment. The local area is given a regional name based on the segment name and a regional code composed of a combination of segment codes. The local area can be designated by the operator through software. For example, the areas where the anterior wall of the heart base with segment number 1, the middle anterior wall with segment number 7 and the apical anterior wall with segment number 13 are located are combined to form a local area called the anterior wall. It is not difficult to see from the segment names that the formation of the local area is actually based on the construction principle of the bull's eye diagram and the actual structure of the heart.
[0036] c. Establish a graded scoring standard, specify the scoring values and the scoring names that describe their meanings; in actual operation, a graded scoring standard can be established, and the graded scoring standard includes the scoring values for the segments and the scoring names for the meanings of the scoring values; in this embodiment, a total of 9 scoring values from -1 to 5 are used for the cardiac motion state (the scoring values are not necessarily integers), where 0 represents no score displayed, and 1 represents normal motion. Scores other than 0 and 1 represent different abnormal motion states of the heart.
[0037] In this embodiment, steps a to c can be implemented by establishing a table in a computer system.
[0038] d. The operator assigns corresponding scoring values to each segment in step a on the computer according to the grading criteria in step c;
[0039] In this embodiment, in order to facilitate the operator to use and facilitate the computer to process data, before the operator scores, the computer constructs a segment scoring array according to the hierarchical scoring standard and segment feature information provided in steps ac. The segment scoring array is an N×M array, wherein N is the total number of segments and M is the number of scoring values. Therefore, the segment scoring array includes M one-dimensional hierarchical scoring arrays corresponding to the scoring values, and each one-dimensional hierarchical scoring array includes N marking parameters corresponding to each segment, and the marking parameters of each segment correspond to the segment number corresponding to it; in this embodiment, 17 segments and 9 scoring values are used, and the segment scoring array corresponds to a 17×9 multidimensional array, that is, there are 9 one-dimensional hierarchical scoring arrays containing 17 elements, each element containing a marking parameter, and the positions of the 17 elements correspond to the 17 segment numbers in sequence respectively;
[0040] When the operator assigns a score value to a segment, this can be accomplished by modifying the marking parameters in the segment scoring array, specifically: finding a one-dimensional hierarchical scoring array corresponding to the score value given by the operator in the segment scoring array, finding the marking parameter corresponding to the segment number of the segment scored by the operator in the one-dimensional hierarchical scoring array, and modifying the initial value to the marking value. The segment whose marking parameter is modified is the marked segment, which includes its own segment number (position parameter of the array element) and marking value.
[0041] e. The computer divides the segments with the same score value into a group according to the score value provided by the operator in step d, and groups all the segments accordingly, and then performs iterative matching, wherein the iterative matching includes:
[0042] e.1 The segment codes of the segments with the same score value are combined into a code group, and the code group is matched with the regional code described in step b. If the code group and the regional code are consistent, at least the regional name corresponding to the matched regional code and the score value of the segment corresponding to the code group are obtained and output in the form of a row of records. In this embodiment, a local region comparison table containing at least regional names and regional codes is established, and the code group is traversed and matched with the regional codes in the table to achieve the maximum complete match between the code group and the regional code. Taking the local region named "anterior wall" in step b as an example, when the code group contains and only contains the segment codes of the three segments of the anterior wall of the heart base with segment number 1, the middle anterior wall with segment number 7, and the apical anterior wall with segment number 13, the code group and the regional code are consistent. At this time, the regional name contained in the output record is "anterior wall".
[0043] e.2 If no consistent result is obtained after matching the coding group with the regional coding, then according to the segment coding in the coding group, at least the corresponding segment name and the score value of the segment corresponding to the coding group are obtained, and output in the form of a line of records; the significance of step e.2 is that, since the local area is established according to the actual structure of the heart and the medical standards for ultrasound examination, there must be segments represented by the coding group scattered in the heart structure that does not meet the definition of the local area. At this time, step e.2 is executed to indicate the specific segment and its motion state.
[0044] e.3 After completing the matching of the coding group under one scoring value grouping, take the next group of segments with the same scoring value and repeat steps e.1 to e.3 until all segments are matched; for this embodiment, since a 17×9 segment scoring array is used, the grouping of the segments according to the scoring values in step e is equivalent to the aforementioned one-dimensional hierarchical scoring array, and the iterative matching is equivalent to taking out each one-dimensional hierarchical scoring array in turn, and performing segment coding extraction and matching and comparison work on the segments whose internal marking parameters have been modified to the marking values; of course, the segment scoring array can also be in the form of 9×17.
[0045] f. Integrate the records output after iterative matching as structured report data content.
[0046] In this embodiment, the structured report data content needs to at least include the region name or segment name and the score value, and is output in the form of a description table. The table structure is shown in Table 1:
[0047]
[0048] Table 1 Description table
[0049] To further facilitate the generation of structured reports, after completing step f, you can also directly obtain the structured report text through the following steps:
[0050] g. Establish a vocabulary containing descriptive words, and correspond the descriptive words to the area name or segment name and the scoring value. After receiving the structured report data content, the computer fills in the corresponding descriptive words at the specified position of the structured report text; when the area name, segment name and scoring name are all medical terms, the descriptive words can be directly a combination of the area name or segment name and the scoring name that expresses the meaning of the scoring value, and better expression effects can be achieved by adding punctuation, slashes or spaces.
[0051] In order to improve data processing efficiency, the hierarchical scoring standard of step c at least includes three types of scoring values and corresponding scoring names, namely normal motion state, no score displayed, and abnormal motion state. In the iterative matching of step e, only the group whose scoring value means abnormal motion state is iteratively matched. It should be noted that there can be multiple scoring values for the abnormal motion state. This is because in actual medical scenarios, the abnormal cardiac motion state is diverse. Therefore, it is necessary to give multiple scoring values indicating abnormal motion state in the hierarchical scoring standard based on medical practice. The scoring value that meets this requirement has been given in step c of this embodiment.
[0052] In order to make the scoring operation more intuitive, the grading and scoring criteria also include color parameters of the bull's eye chart corresponding to each scoring value; when the operator scores each segment of the bull's eye chart in step d, the scoring value can be given by providing color information; in actual operation, the bull's eye chart is displayed on a computer operation interface, and the scoring of each segment of the bull's eye chart is achieved by selecting the segments and corresponding colors by the computer.
[0053] In order to facilitate the use of operators and adapt to the habits of different operators, such as Figure 3 As shown, this embodiment also provides a method for synchronously drawing a bull's eye chart when modifying structured newspaper content, including the following steps:
[0054] h. In step b, the conclusion codes corresponding to each segment and local area are assigned in the computer, and the conclusion codes are also included in the structured report data content, as shown in Table 1. In step c, each score value of the grading standard also has a corresponding color parameter; in actual operation, this step and steps a to c only need to be specified at the first use and can be reused;
[0055] i. In the structured report editing interface, the operator gives a score value for the segment or local area according to the grading criteria in step c;
[0056] j. After the content of the structured report is edited and changed in step i, the computer automatically obtains the changed multiple rows of records. In this embodiment, the computer automatically obtains the changed records of each row from the corresponding changed structured report data content, i.e., the description table, and performs reverse matching. The reverse matching includes:
[0057] j.1. Obtain the conclusion code corresponding to the area name or segment name in the current record, and find the area code or segment code corresponding to the conclusion code; in this example, although the conclusion code is used for reverse matching, this method is only to facilitate the forward iterative matching and reverse matching synchronization operations in step e. In fact, due to the rich information in the description table, the segment code or area code corresponding to the area name or segment name in the table can also be matched. The use of conclusion coding in this embodiment is only a preferred method. In fact, the conclusion code, area code and other segment feature information can be used as feature parameters of the record. The description table of this implementation uses the area name or segment name as the name of a row of records (i.e., the basis for branching). Therefore, as long as the corresponding feature parameters are taken out, the reverse matching process can be realized.
[0058] j.2. Find the corresponding segment number through the found area code or segment code; if the area code is found, decompose the area code into individual segment codes; if the segment code is found, directly find the corresponding segment number.
[0059] j.3. Associating the score value in the current record with the segment number; in this implementation, according to the score value in the record, find the corresponding one-dimensional hierarchical score array in the segment score array, and according to the found segment number, modify the tag parameter of the corresponding element in the one-dimensional hierarchical score array to the tag value;
[0060] j.4 Repeat the above reverse matching operation for the next row of records until the traversal of multiple rows of records is completed, and finally the corresponding relationship between the score value given in step i and the segment number of each segment on the bull's eye chart is obtained.
[0061] k. The computer obtains the correspondence between the segment numbers and the score values and the score values and the color parameters, and colors the bull's eye chart accordingly; the coloring is to find the corresponding segment on the bull's eye chart according to the segment number from the segment score array, and then find the corresponding bull's eye chart color parameter according to the score value corresponding to the mark corresponding to the segment number, so as to automatically color the bull's eye chart.
[0062] The structured automatic assisted analysis method of the bull's eye chart of echocardiography provided in this embodiment associates the graphic content of the bull's eye chart with the structured report content by encoding the segmental motion evaluation indicators. At the same time, by setting local areas and regional codes, the correct description of the local ventricular wall range expressed by the bull's eye chart is achieved, which maximizes the user's understanding habits and achieves the effect of synchronous modification of pictures and texts and accurate correspondence between pictures and texts.
[0063] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Although the present invention has been disclosed as a preferred embodiment as above, it is not used to limit the present invention. Any technical personnel in this field can make some changes or modify the technical contents disclosed above into equivalent embodiments without departing from the scope of the technical solution of the present invention. However, any brief modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are still within the scope of the technical solution of the present invention.
Claims
1. A structured automatic auxiliary analysis method for an echocardiogram bull's eye image, characterized in that: Ability to output structured report data content simultaneously when generating a bull's eye chart, including the following steps: a. Assign corresponding segment feature information to each segment on the bull's eye diagram, and the segment feature information includes segment name, segment number, and segment code; b. According to the actual structure of the heart and the medical standards for ultrasound examination, the segments are combined to form a local area with a range larger than a single segment, and the local area is given a regional name based on the segment name and a regional code composed of a combination of segment codes; c. Establish a grading standard, specify the score value and the score name that describes its meaning; d. The operator assigns corresponding scoring values to each segment in step a on the computer according to the grading criteria in step c; e. The computer divides the segments with the same score value into a group according to the score value provided by the operator in step d, and groups all the segments according to the score value, and then performs iterative matching, wherein the iterative matching includes: e.1 Take a group of segment codes of segments with the same score value and combine them into a code group, match the code group with the area code in step b, if the code group and the area code are consistent, then at least obtain the area name corresponding to the matched area code and the score value of the segment corresponding to the code group, and output them in the form of a line of records; e.2 If no consistent result is obtained after matching the coding group with the regional code, then according to the segment code in the coding group, at least the corresponding segment name and the score value of the segment corresponding to the coding group are obtained and output in the form of a line of records; e.3 After completing the matching of the coding groups under a scoring value group, take the next group of segments with the same scoring value and repeat steps e.1 to e.3 until all segments are matched; f. Integrate the records output after iterative matching as structured report data content; After completing step f, it also includes: g. Establish a vocabulary containing descriptive words, and correspond the descriptive words to the region name or segment name and the score value. When the structured report data content is received, the computer fills the corresponding descriptive words in the designated position of the structured report text; It is also possible to draw a bull's eye chart synchronously when modifying the structured report content, including the following steps: h. In step b, the conclusion codes corresponding to each segment and local area are assigned in the computer, and the conclusion codes are also included in the structured report data content. In step c, each score value of the graded scoring standard also has a corresponding color parameter; i. In the structured report editing interface, the operator gives a score value for the segment or local area according to the grading criteria in step c; j. After the data content of the structured report is edited and changed in step i, the computer automatically obtains the changed multiple lines of records and performs reverse matching line by line. The reverse matching includes: j.
1. Obtain the conclusion code in the current record and find the region code or segment code corresponding to the conclusion code; j.
2. Find the corresponding segment number by finding the area code or segment code; j.
3. Associate the score value in the current record with the segment number; j.4 Repeat the above reverse matching operation for the next row of records until the traversal of multiple rows of records is completed, and finally the corresponding relationship between the score value given in step i and the segment number of each segment on the bull's eye chart is obtained k. The computer obtains the correspondence between the segment numbers and the score values and the correspondence between the score values and the color parameters, and colors the bull's eye chart accordingly.
2. The method for structured automatic auxiliary analysis of the bull's eye image of echocardiography according to claim 1, characterized in that: The hierarchical scoring criteria of step c at least include three types of scoring values and corresponding scoring names, namely, normal motion status, no score displayed, and abnormal motion status. In the iterative matching of step e, only the group whose scoring value means abnormal motion status is iteratively matched.
3. The method for structured automatic auxiliary analysis of the bull's eye image of echocardiography according to claim 1 or 2, characterized in that: The grading and scoring criteria also include a bull's eye chart color parameter corresponding to each score value; when the operator scores each segment of the bull's eye chart in step d, the score value can be given by providing color information.
