Medical ultrasound imaging report structured generation method, device and storage medium

Abstract

The present application discloses a method, device and storage medium for structured generation of medical ultrasound image reports, and relates to the technical field of medical report generation. The method for structured generation of medical ultrasound image reports includes: obtaining a doctor's oral voice and multiple historical ultrasound image reports; dividing multiple historical ultrasound image reports into multiple clusters based on the location of the ultrasound image and the classification standard of the diagnostic opinion; obtaining a first cluster based on the multiple clusters; obtaining the importance data of each historical ultrasound image report in the first cluster based on the first cluster; and sorting each historical ultrasound image report in the first cluster in order of importance data from large to small. When the present application performs structured generation of medical ultrasound image reports based on the doctor's real-time voice data, templates with higher importance can be retrieved preferentially, thereby improving the matching speed while maintaining the matching accuracy of the ultrasound image template.

Description

Technical Field

[0001] The present application relates to the technical field of medical report generation, and specifically to a structured generation method, device and storage medium for medical ultrasound imaging reports. Background Art

[0002] Ultrasound testing currently includes clinicians issuing ultrasound examination orders, patients making appointments, ultrasound doctors examining patients, communicating with clinicians about the findings, and clinicians filling out ultrasound reports based on the communication content and exploration screenshots. Currently, most ultrasound reports are filled out manually by clinicians. In order to reduce the time doctors spend writing medical records and save overall time and labor costs, some hospitals are trying to use artificial intelligence (AIGC) technology to assist clinicians in writing medical records, so as to reduce the time clinicians spend writing and increase the time clinicians spend communicating about treatment.

[0003] Specifically, when the existing technology generates a medical ultrasound image report through AIGC technology, it first uses speech recognition technology to convert the clinician's oral statement into text data, and then matches the text data to the corresponding ultrasound image report template to generate the corresponding ultrasound report. Since ultrasound image report templates usually have multiple forms (the form here means that the template is different for different human organs for ultrasound examination and different conditions, respectively), the existing method is to directly match the corresponding ultrasound impact template based on the doctor's complete description, and the corresponding matching time is relatively long. In actual application, if too many ultrasound image reports need to be generated in real time, the accuracy and generation efficiency of the ultrasound image report will be affected. Summary of the invention

[0004] The purpose of this application is to provide a structured generation method, device and storage medium for medical ultrasound image reports to solve the technical problems of low accuracy and low generation efficiency of ultrasound image reports generated based on AIGC technology in the prior art.

[0005] To achieve the above objectives, this application provides the following technical solutions:

[0006] In a first aspect, the present application proposes a method for generating a structured medical ultrasound image report, the method comprising:

[0007] Obtain the physician's spoken voice and multiple historical ultrasound imaging reports;

[0008] Based on the ultrasound image location and diagnostic opinion classification criteria, multiple historical ultrasound image reports are divided into multiple clusters;

[0009] Based on the multiple clusters, a first cluster is obtained; the first cluster is any one of the multiple clusters;

[0010] Based on the first cluster, obtaining importance data of each historical ultrasound imaging report in the first cluster;

[0011] Sorting each historical ultrasound imaging report in the first cluster according to the order of importance data from large to small;

[0012] Based on the spoken speech, obtaining text data;

[0013] Matching the ultrasound image parts and diagnostic opinions in the text data with each cluster to obtain a target cluster;

[0014] Sequentially obtaining the target matching degree between the text data and each historical ultrasound image report in the target cluster;

[0015] If a target matching degree meets the preset conditions, the historical ultrasound image report corresponding to the target matching degree is obtained as a template to generate the current ultrasound image report.

[0016] As a specific solution in the technical solution of the present application, the classification standard based on the location and diagnostic opinion of the ultrasound image divides multiple historical ultrasound image reports into multiple clusters, including:

[0017] Based on the location of the ultrasound image, the multiple historical ultrasound image reports are divided into multiple second clusters;

[0018] Based on the multiple second clusters, a third cluster is obtained; the third cluster is any one of the multiple second clusters;

[0019] Based on the word segmentation tool, a first word segmentation array corresponding to each historical ultrasound imaging report in the third cluster is obtained, and a second word segmentation array corresponding to each diagnostic opinion in the diagnostic opinion classification standard is extracted;

[0020] Match each first word segmentation array with each second word segmentation array to obtain a first matching degree corresponding to each second word segmentation array;

[0021] According to the first matching degree, each historical ultrasound imaging report is divided into a plurality of fourth clusters according to the diagnosis opinions;

[0022] Based on the plurality of fourth clustering clusters, a plurality of clustering clusters are obtained.

[0023] As a specific solution in the technical solution of the present application, obtaining the importance data of each historical ultrasound imaging report in the first cluster based on the first cluster includes:

[0024] Based on the word segmentation tool, the word segmentations in all the historical ultrasound imaging reports in the first cluster are obtained to form a third word segmentation array;

[0025] Acquire a fourth word segmentation array corresponding to each historical ultrasound imaging report in the first cluster based on a word segmentation tool;

[0026] Based on the third word segmentation array and the fourth word segmentation array, importance data of the historical ultrasound image report corresponding to the fourth word segmentation array is obtained.

[0027] As a specific solution in the technical solution of the present application, the obtaining, based on the third word segmentation array and the fourth word segmentation array, the importance data of the historical ultrasound image report corresponding to the fourth word segmentation array includes:

[0028] Based on the fourth word segmentation array, obtain a first word segmentation; the first word segmentation is any word segmentation in the fourth word segmentation array;

[0029] Based on the first word segmentation, obtain the number of the first word segmentation in the third word segmentation array;

[0030] Based on the number of segmented words and the number of historical ultrasound image reports in the first cluster, importance data of the historical ultrasound image reports corresponding to the fourth segmented word array is obtained.

[0031] As a specific solution in the technical solution of the present application, the calculation formula for obtaining the importance data of the historical ultrasound image reports corresponding to the fourth word segmentation array based on the number of word segments and the number of historical ultrasound image reports in the first cluster is as follows:

[0032]

[0033] in, represents the importance data of the k-th historical ultrasound image report in the first cluster; V represents the number of segmented words in the fourth segmented word array corresponding to the k-th historical ultrasound image report; represents the number of words in the third word array where the i-th word in the fourth word array is located; Represents the number of historical ultrasound imaging reports in the first cluster.

[0034] As a specific solution in the technical solution of the present application, matching the ultrasound image parts and diagnostic opinions in the text data with each cluster to obtain the target cluster includes:

[0035] Based on the word segmentation tool, a fifth word segmentation array of ultrasound image parts and diagnostic opinions in the text data and a sixth word segmentation array corresponding to each cluster are obtained;

[0036] Based on the fifth word segmentation array and each sixth word segmentation array, obtaining a second matching degree between the text data and each clustering cluster;

[0037] Based on each second matching degree, a target cluster is obtained.

[0038] As a specific solution in the technical solution of the present application, the step of sequentially obtaining the target matching degree between the text data and each historical ultrasound image report in the target cluster includes:

[0039] Based on the word segmentation tool, a seventh word segmentation array of the text data is obtained, and an eighth word segmentation array of each historical ultrasound imaging report in the target cluster is obtained in sequence;

[0040] Based on the seventh word segmentation array and the eighth word segmentation array, obtaining a third matching degree between the text data and each historical ultrasound image report in each target cluster;

[0041] Based on each of the third matching degrees, the target matching degree is obtained.

[0042] As a specific solution in the technical solution of this application, the preset conditions are met, including:

[0043] If a target matching degree is greater than or equal to a preset threshold, it is considered that the target matching degree meets the preset condition.

[0044] In a second aspect, the present application proposes a medical ultrasound image report structured generation device, the medical ultrasound image report structured generation device comprising:

[0045] An acquisition device for acquiring a doctor's spoken voice and multiple historical ultrasound image reports;

[0046] A server, used for dividing a plurality of historical ultrasound image reports into a plurality of clusters based on the location of the ultrasound image and the classification standard of the diagnosis opinion;

[0047] And, based on the multiple clusters, a first cluster is obtained; the first cluster is any one of the multiple clusters;

[0048] And, based on the first cluster, obtaining importance data of each historical ultrasound imaging report in the first cluster;

[0049] And, sorting each historical ultrasound imaging report in the first cluster according to the order of importance data from large to small;

[0050] and, based on the spoken speech, obtaining text data;

[0051] Matching the ultrasound image parts and diagnostic opinions in the text data with each cluster to obtain a target cluster;

[0052] and, sequentially obtaining the target matching degree between the text data and each historical ultrasound image report in the target cluster;

[0053] Furthermore, if a target matching degree meets a preset condition, a historical ultrasound image report corresponding to the target matching degree is obtained as a template to generate a current ultrasound image report.

[0054] As a specific solution in the technical solution of the present application, the server is further used to divide the multiple historical ultrasound image reports into multiple second clusters based on the location of the ultrasound image;

[0055] And, based on the multiple second clusters, a third cluster is obtained; the third cluster is any one of the multiple second clusters;

[0056] And, based on the word segmentation tool, a first word segmentation array corresponding to each historical ultrasound imaging report in the third cluster is obtained, and a second word segmentation array corresponding to each diagnostic opinion in the diagnostic opinion classification standard is extracted;

[0057] And, matching each first word segmentation array with each second word segmentation array to obtain a first matching degree corresponding to each second word segmentation array;

[0058] and, dividing each historical ultrasound imaging report into a plurality of fourth clusters according to the diagnosis opinions according to the first matching degree;

[0059] And, based on the plurality of fourth clustering clusters, a plurality of clustering clusters are obtained.

[0060] As a specific solution in the technical solution of the present application, the server is further used to obtain the segmented words in all historical ultrasound imaging reports in the first cluster based on the segmentation tool to form a third segmented word array;

[0061] And, obtaining a fourth word segmentation array corresponding to each historical ultrasound imaging report in the first cluster based on the word segmentation tool;

[0062] And, based on the third word segmentation array and the fourth word segmentation array, the importance data of the historical ultrasound image report corresponding to the fourth word segmentation array is obtained.

[0063] As a specific solution in the technical solution of the present application, the server is further used to obtain a first word segmentation based on the fourth word segmentation array; the first word segmentation is any word segmentation in the fourth word segmentation array;

[0064] And, based on the first word segmentation, obtaining the number of the first word segmentation located in the third word segmentation array;

[0065] And, based on the number of segmented words and the number of historical ultrasound image reports in the first cluster, importance data of the historical ultrasound image reports corresponding to the fourth segmented word array is obtained.

[0066] As a specific solution in the technical solution of the present application, the server is further used to obtain the importance data of the historical ultrasound image report corresponding to the fourth word segmentation array using the following calculation formula:

[0067]

[0068] in, represents the importance data of the k-th historical ultrasound image report in the first cluster; V represents the number of segmented words in the fourth segmented word array corresponding to the k-th historical ultrasound image report; represents the number of words in the third word array where the i-th word in the fourth word array is located; Represents the number of historical ultrasound imaging reports in the first cluster.

[0069] As a specific solution in the technical solution of the present application, the server is further used to obtain, based on a word segmentation tool, a fifth word segmentation array of ultrasound image parts and diagnostic opinions in the text data and a sixth word segmentation array corresponding to each cluster;

[0070] And, based on the fifth word segmentation array and each sixth word segmentation array, obtaining a second matching degree between the text data and each clustering cluster;

[0071] And, based on each second matching degree, a target cluster is obtained.

[0072] As a specific solution in the technical solution of the present application, the server is further used to obtain the seventh word segmentation array of the text data based on the word segmentation tool, and sequentially obtain the eighth word segmentation array of each historical ultrasound imaging report in the target cluster;

[0073] And, based on the seventh word segmentation array and the eighth word segmentation array, obtaining a third matching degree between the text data and each historical ultrasound image report in each of the target clusters;

[0074] And, based on each of the third matching degrees, the target matching degree is obtained.

[0075] As a specific solution in the technical solution of this application, the preset conditions are met, including:

[0076] If a target matching degree is greater than or equal to a preset threshold, it is considered that the target matching degree meets the preset condition.

[0077] In a third aspect, the present application proposes a computer-readable storage medium storing a computer program, which, when executed by a processor, implements a structured generation method for a medical ultrasound imaging report as described in any one of the first aspects.

[0078] Compared with the prior art, the beneficial effects of this application are:

[0079] This application can obtain diagnostic opinions in multiple ultrasound image packages based on historical ultrasound image reports, and sort all ultrasound image report templates in each diagnostic opinion category in order of importance from large to small. When the medical ultrasound image report is subsequently structured and generated based on the doctor's real-time voice data, templates with higher importance can be retrieved first, which can improve the matching speed while maintaining the matching accuracy of ultrasound image templates. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] Figure 1 A schematic diagram of a flow chart of a method for generating a structured medical ultrasound image report according to an embodiment of the present application;

[0081] Figure 2 A schematic diagram of the structure of a medical ultrasound image report structured generation device proposed in an embodiment of the present application. DETAILED DESCRIPTION

[0082] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0083] The terms "first", "second", etc. in the description of the embodiment of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence, such as the first matching degree and the second matching degree proposed below, which belong to different matching degrees. It should be understood that the matching degrees used in this way can be interchanged where appropriate, so that the embodiments described here can be implemented in an order other than the content illustrated or described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device including a series of steps or modules is not necessarily limited to those steps or modules clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products or devices. The division of modules that appear in the embodiment of the present application is only a logical division. There may be other division methods when implemented in actual applications, such as multiple modules can be combined or integrated into another system, or some features can be ignored, or not executed. In addition, the coupling or direct coupling or communication connection between each other shown or discussed can be through some interfaces, and the indirect coupling or communication connection between modules can be electrical or other similar forms, which are not limited in the embodiment of the present application. Furthermore, the modules or sub-modules described as separate components may or may not be physically separated, may or may not be physical modules, or may be distributed in multiple circuit modules, and some or all of the modules may be selected according to actual needs to achieve the purpose of the embodiment of the present application.

[0084] In order to solve the technical problems of low accuracy and low generation efficiency of ultrasound image reports generated based on AIGC technology in the prior art mentioned in the background technology, this application proposes a structured generation method for medical ultrasound image reports, such as Figure 1 As shown, the method for generating a structured medical ultrasound image report includes steps S100 to S900.

[0085] Step S100: Acquire the doctor's oral speech and multiple historical ultrasound image reports.

[0086] It should be clear that, generally, there are a large number of historical ultrasound image reports in the database of the hospital. In this embodiment, the structure and content of the historical ultrasound image reports can be used as a reference for generating the current ultrasound image report. Therefore, in this embodiment, each historical ultrasound image report can be regarded as a template, and the historical ultrasound image reports can be retrieved according to the voice input to obtain the template that best matches the current ultrasound image report. Then, the template can be adaptively modified according to the voice input to generate a current ultrasound image report with a higher degree of adaptability for each patient.

[0087] In the actual ultrasound examination process, the doctor usually does not verbally describe the patient's entire ultrasound description at once (for example, sometimes the patient asks a question and the doctor says something related to the ultrasound description), so it is necessary to dynamically search and match the template of the current ultrasound image report based on the doctor's real-time oral content. In the embodiment of the present application, the oral voice can be obtained by using a microphone, which is a mature technology and will not be described in detail here.

[0088] Step S200: dividing a plurality of historical ultrasound image reports into a plurality of clusters based on the location of the ultrasound image and the diagnosis opinion classification standard.

[0089] It should be noted that the diagnostic opinion classification standard is an existing technical standard and will not be described in detail here. In this embodiment, multiple historical ultrasound image reports can be divided into multiple clusters based on the location of the ultrasound image and the diagnostic opinion classification standard in any manner. For example, each historical ultrasound image report can be divided into multiple clusters manually.

[0090] In order to automatically divide multiple historical ultrasound image reports into multiple clusters, in one embodiment of the present application, step S200 divides multiple historical ultrasound image reports into multiple clusters based on the location of the ultrasound image and the diagnosis opinion classification standard, including steps S210 to S260.

[0091] Step S210: Based on the location of the ultrasound image, multiple historical ultrasound image reports are divided into multiple second clusters.

[0092] It should be clear that there are many parts that need to be examined by ultrasound imaging, such as the neck, chest, abdomen, pelvic organs, and blood vessels of the limbs, etc. Therefore, multiple historical ultrasound imaging reports can be preliminarily divided into multiple second clusters based on these parts.

[0093] Step S220: obtaining a third cluster based on the plurality of second clusters.

[0094] It should be clear that the conditions caused by the same part of the human body are also diverse. For example, ultrasound examination can detect lesions of the biliary system such as gallstones, cholecystitis, and gallbladder polyps; kidney-related diseases such as kidney stones, renal cysts, hydronephrosis, and renal vascular lesions can also be detected, and liver diseases such as liver cysts, hepatic fatty degeneration, hepatic hemangiomas, and liver tumors can also be detected. That is to say, in an embodiment of the present application, each second cluster can be more accurately classified according to the diagnostic opinion classification criteria. That is to say, in an embodiment of the present application, each second cluster needs to be subdivided again. In other words, in an embodiment of the present application, the third cluster is any one of the multiple second clusters.

[0095] Step S230: obtaining a first word segmentation array corresponding to each historical ultrasound imaging report in the third cluster based on a word segmentation tool, and extracting a second word segmentation array corresponding to each diagnostic opinion in the diagnostic opinion classification standard.

[0096] It should be clear that word segmentation tools refer to tools that split a sequence of Chinese characters into separate words. These tools play an important role in Chinese information processing because Chinese, unlike English, has no obvious separators such as spaces between words. Therefore, word segmentation tools are needed to recombine continuous sequences of Chinese characters into meaningful vocabulary sequences. The application of word segmentation tools can help computers better understand and process Chinese texts, and is a key step in natural language processing. There are many word segmentation tools in the prior art, such as NLPIR, THULAC, LTP, HanLP, JieBa and SnowNLP, etc., which will not be listed here one by one.

[0097] Step S240: Match each first word segmentation array with each second word segmentation array to obtain a first matching degree corresponding to each second word segmentation array.

[0098] It should be clear that if a first word segmentation array has the highest matching degree with a second word segmentation array among the second word segmentation arrays, the historical ultrasound image report corresponding to the first word segmentation array can be classified into the classification cluster of the diagnostic opinion corresponding to the second word segmentation array.

[0099] Step S250: dividing each historical ultrasound image report into a plurality of fourth clusters according to the first matching degree and the diagnosis opinions.

[0100] As can be seen from the foregoing, in the embodiment of the present application, each historical ultrasound image report is divided into the diagnostic opinion with the highest matching degree therewith, so as to form a plurality of fourth clusters.

[0101] In order to explain the embodiment of the present application in detail, the present application takes the thyroid gland (i.e., the third cluster, hereinafter referred to as the thyroid cluster) as an example of the part of the ultrasound examination, and explains steps S230 to S250 of the present application. Specifically, in this embodiment, first, according to the professional thyroid report diagnostic opinion classification standard (which is an existing standard and will not be repeated), all diagnostic opinions are obtained, assuming that there are N in total. For all historical ultrasound imaging reports in the thyroid cluster, all prompt opinions are classified by analyzing the text descriptions of the ultrasound prompt column in all historical ultrasound imaging reports. In the ultrasound imaging report, the content before the brackets in the ultrasound prompt column represents the corresponding diagnostic opinion, and the content in the brackets represents the attribute characteristics corresponding to the diagnostic opinion. The same diagnostic opinion usually corresponds to different attribute characteristics. Specifically, the text data of the diagnostic opinion and attribute feature part in the ultrasound prompt column in each ultrasound imaging report in the thyroid cluster are obtained respectively, and then the JieBa word segmentation tool is used to perform word segmentation on the obtained text data. For example, the "thyroid bilobed cystic nodule" in the above report is segmented as ["thyroid", "bilobed", "cystic", "nodule"] (that is, the first word segmentation array is obtained). Similarly, all the N standard diagnostic opinions are also segmented (that is, each second segmentation array is obtained). Then the matching degree between each ultrasound image report in the thyroid cluster and the N professional diagnostic opinions is calculated. The specific mathematical formula is:

[0102]

[0103] In the formula, represents the best match between the jth historical ultrasound imaging report and N standard diagnostic opinions in the thyroid cluster; represents the Jaccard correlation coefficient; The word segmentation array representing the nth diagnostic opinion among N standard diagnostic opinions; The word segmentation array representing the diagnostic opinion of the j-th historical ultrasound imaging report in the thyroid cluster; represents the matching degree between the diagnosis opinion of the jth historical ultrasound imaging report in the thyroid cluster and the nth standard diagnosis opinion among N standard diagnosis opinions; Indicates taking the maximum value.

[0104] Then The value of and the corresponding standard diagnostic opinion are recorded as the matching diagnostic opinion of the jth ultrasound image report in the thyroid cluster. Therefore, according to the above method, the standard diagnostic opinion matched by each historical ultrasound image report in the standard diagnostic opinion can be calculated. Then, all ultrasound image reports in the standard diagnostic opinion can be divided into N clusters.

[0105] Step S260: Based on the plurality of fourth clusters, a plurality of clusters are obtained.

[0106] In one embodiment of the present application, the plurality of fourth clusters may be directly used as the plurality of clusters. Of course, the plurality of fourth clusters may also be further subdivided to obtain the plurality of clusters, which will not be described in detail here.

[0107] It should be clear that in this embodiment, the doctor's spoken voice can be converted into text data, and then the similarity calculation is performed based on the text data and the diagnostic opinions in each historical ultrasound image report in the database, and the historical ultrasound image report with the greatest similarity is used as the template of the current ultrasound image report. It is easy to understand that if the number of historical ultrasound image reports is large, it will seriously affect the efficiency of generating ultrasound image reports. In order to be able to quickly generate the current ultrasound image report. The historical ultrasound image reports in each cluster can be sorted according to their importance. When the similarity between the text data and each historical ultrasound image report is calculated later, the historical ultrasound image report that is more similar to the text data can be quickly obtained as a template. In this embodiment, the sorting method for each historical ultrasound image report in each cluster is as follows in steps S300 to S500.

[0108] Step S300: obtaining a first cluster based on multiple clusters.

[0109] It should be clear that, in the embodiment of the present application, the first cluster refers to a cluster in which each historical ultrasound image report needs to be sorted. In other words, the first cluster is any one of the multiple clusters.

[0110] Step S400: Based on the first cluster, obtaining importance data of each historical ultrasound image report in the first cluster.

[0111] It should be clear that, in the present embodiment, the importance data of each historical ultrasound image report is used to characterize the importance of the historical ultrasound image report in the first cluster. Any indicator can be used to characterize the importance of the historical ultrasound image report in the first cluster. For example, in one embodiment of the present application, the number of diagnostic opinion words in the historical ultrasound image report can be used to characterize the importance of the historical ultrasound image report in the first cluster. The more diagnostic opinion words in the historical ultrasound image report, the higher the importance of the historical ultrasound image report in the first cluster.

[0112] In a specific embodiment of the present application, the importance of the historical ultrasound image report in the first cluster can also be represented by the number of occurrences of the diagnostic opinion of the historical ultrasound image report in the first cluster. The more times the diagnostic opinion of the historical ultrasound image report appears in the first cluster, the higher the importance of the historical ultrasound image report in the first cluster. Specifically, in this embodiment, step S400, based on the first cluster, obtains the importance data of each historical ultrasound image report in the first cluster, including steps S410 to S430.

[0113] Step S410: obtaining the segmented words in all the historical ultrasound imaging reports in the first cluster based on the segmented word tool to form a third segmented word array.

[0114] It should be clear that the third word segmentation array includes the word segmentations in all historical ultrasound imaging reports in the first cluster, that is, the word segmentations in the third word segmentation array may be repeated, for example, the word "thyroid" or the word "nodule" may appear multiple times.

[0115] Step S420: obtaining a fourth word segmentation array corresponding to each historical ultrasound imaging report in the first cluster based on the word segmentation tool.

[0116] It should be clear that obtaining a word segmentation array in a historical ultrasound image report based on a word segmentation tool is a mature technology and will not be described in detail here.

[0117] Step S430: Based on the third word segmentation array and the fourth word segmentation array, obtain importance data of the historical ultrasound image report corresponding to the fourth word segmentation array.

[0118] It is easy to understand that for historical ultrasound imaging reports, if a certain word extracted from it appears more frequently in other historical ultrasound imaging reports in the same cluster, it means that the importance of the word and the cluster is higher, that is, the corresponding historical ultrasound imaging report is more important in the entire cluster.

[0119] In order to clearly characterize the importance of the historical ultrasound image report in the entire cluster, in one embodiment of the present application, step S430, based on the third word segmentation array and the fourth word segmentation array, obtains the importance data of the historical ultrasound image report corresponding to the fourth word segmentation array, including steps S431 to S433.

[0120] Step S431: Based on the fourth word segmentation array, obtain the first word segmentation.

[0121] It is easy to understand that in this embodiment, the first word segmentation is any word segmentation in the fourth word segmentation array. That is, it is necessary to calculate the importance of each word segmentation in the historical ultrasound image report, and then evaluate the importance of the entire historical ultrasound image report in the cluster.

[0122] Step S432: Based on the first word segmentation, obtain the number of the first word segmentations in the third word segmentation array.

[0123] It should be clear that obtaining the number of occurrences of a certain word in the word segmentation array is a mature technology and will not be described in detail here.

[0124] Step S433: Based on the number of segmented words and the number of historical ultrasound image reports in the first cluster, obtain importance data of the historical ultrasound image reports corresponding to the fourth segmented word array.

[0125] It should be clear that in an embodiment of the present application, the importance data of the historical ultrasound image report can be obtained based on the sum of the number of occurrences of each word segment in the historical ultrasound image report. In another embodiment of the present application, step S433, based on the number of word segments and the number of historical ultrasound image reports in the first cluster, the calculation formula for obtaining the importance data of the historical ultrasound image report corresponding to the fourth word segment array is as follows:

[0126]

[0127] in, represents the importance data of the k-th historical ultrasound image report in the first cluster; V represents the number of segmented words in the fourth segmented word array corresponding to the k-th historical ultrasound image report; represents the number of words in the third word array where the i-th word in the fourth word array is located; Represents the number of historical ultrasound imaging reports in the first cluster.

[0128] Step S500: sorting the historical ultrasound imaging reports in the first cluster according to the order of importance data from large to small.

[0129] That is to say, in the embodiment of the present application, each historical ultrasound image report in each cluster is sorted according to the importance, from large to small.

[0130] Step S600: obtaining text data based on the spoken voice;

[0131] It should be clear that obtaining text data based on oral speech is a mature technology. For example, the doctor's real-time oral speech can be recognized and obtained through iFlytek speech recognition software and converted into text data.

[0132] Step S700: Matching the ultrasound image parts and diagnostic opinions in the text data with each cluster to obtain a target cluster, including steps S710 to S730.

[0133] It should be clear that in the embodiments of the present application, the doctor's first sentence can be described as a preset fixed description. For example, if the examination site is the thyroid gland, the doctor's first sentence is described as a thyroid examination. The cluster corresponding to this ultrasound examination is matched according to the text data described in the first sentence. The matching method is to obtain the word segmentation array of the first sentence description, match the word segmentation in the word segmentation array with all the clusters, and if the word segmentation in the word segmentation array and the name of the cluster are the same, then the cluster is a matching cluster.

[0134] Step S710: Based on the word segmentation tool, a fifth word segmentation array of ultrasound image parts and diagnostic opinions in the text data and a sixth word segmentation array corresponding to each cluster are obtained.

[0135] It should be clear that obtaining word segmentation arrays from text data based on word segmentation tools is a mature technology and will not be elaborated on here.

[0136] Step S720: Based on the fifth word segmentation array and each sixth word segmentation array, obtain a second matching degree between the text data and each clustering cluster.

[0137] It should be noted that the step of obtaining the second matching degree here may be similar to step S250.

[0138] Step S730: Obtain target clusters based on each second matching degree.

[0139] In an embodiment of the present application, the cluster corresponding to the sixth word segmentation array having the highest matching degree with the fifth word segmentation array may be obtained as the target cluster.

[0140] Step S800: sequentially obtaining the target matching degree between the text data and each historical ultrasound image report in the target cluster.

[0141] In an embodiment of the present application, the historical ultrasound image report with the highest matching degree between the target cluster and the text data can be obtained as a template for the current ultrasound image report. In order to avoid too many historical ultrasound image reports, which leads to too much calculation, thereby affecting the generation efficiency of the current ultrasound image report. In one embodiment of the present application, step S800, sequentially obtaining the target matching degree between the text data and each historical ultrasound image report in the target cluster, includes steps S810 to S830.

[0142] Step S810: Based on the word segmentation tool, the seventh word segmentation array of the text data is obtained, and the eighth word segmentation array of each historical ultrasound imaging report in the target cluster is obtained in sequence.

[0143] It should be clear that obtaining word segmentation arrays from text data based on word segmentation tools is a mature technology and will not be elaborated on here.

[0144] Step S820: Based on the seventh word segmentation array and the eighth word segmentation array, a third matching degree between the text data and each historical ultrasound image report in each of the target clusters is obtained.

[0145] It should be noted that the step of obtaining the third matching degree here may be similar to step S250.

[0146] Step S830: Based on each of the third matching degrees, the target matching degree is obtained.

[0147] It should be clear that in this embodiment, since each historical ultrasound image report in the target cluster is arranged from most important to least important, it is not necessary to traverse each historical ultrasound image report in the target cluster, and it is only necessary to calculate the matching degree (that is, the target matching degree) of each historical ultrasound image report and the text data in turn. If the target matching degree of a historical ultrasound image report meets the preset conditions, the historical ultrasound image report corresponding to the target matching degree can be used as a template to generate the current ultrasound image report.

[0148] Step S900: If a target matching degree meets a preset condition, a historical ultrasound image report corresponding to the target matching degree is obtained as a template to generate a current ultrasound image report.

[0149] In an embodiment of the present application, the meeting of the preset condition includes: if a target matching degree is greater than or equal to a preset threshold, then the target matching degree is considered to meet the preset condition. In an embodiment of the present application, the preset threshold may be an empirical value, for example, the preset threshold may be 0.7 or 0.8.

[0150] It should be clear that the method for generating a structured medical ultrasound image report proposed in the embodiment of the present application can obtain diagnostic opinions in multiple ultrasound image reports based on historical ultrasound image reports, and sort all ultrasound image report templates in each diagnostic opinion class in order of importance from large to small. When the medical ultrasound image report is subsequently structured and generated based on the doctor's real-time voice data, templates with higher importance can be retrieved first, which can improve the matching speed while maintaining the matching accuracy of the ultrasound image template.

[0151] After introducing the method for generating a structured medical ultrasound image report according to the embodiment of the present application, the following describes a device for generating a structured medical ultrasound image report according to the embodiment of the present application. Figure 2 As shown, the medical ultrasound image report structured generation device 10 includes:

[0152] An acquirer 11, used for acquiring a doctor's oral speech and a plurality of historical ultrasound image reports;

[0153] The server 12 is used to divide the plurality of historical ultrasound image reports into a plurality of clusters based on the location of the ultrasound image and the classification standard of the diagnosis opinion;

[0154] And, based on the multiple clusters, a first cluster is obtained; the first cluster is any one of the multiple clusters;

[0155] And, based on the first cluster, obtaining importance data of each historical ultrasound imaging report in the first cluster;

[0156] And, sorting each historical ultrasound imaging report in the first cluster according to the order of importance data from large to small;

[0157] and, based on the spoken speech, obtaining text data;

[0158] Matching the ultrasound image parts and diagnostic opinions in the text data with each cluster to obtain a target cluster;

[0159] and, sequentially obtaining the target matching degree between the text data and each historical ultrasound image report in the target cluster;

[0160] Furthermore, if a target matching degree meets a preset condition, a historical ultrasound image report corresponding to the target matching degree is obtained as a template to generate a current ultrasound image report.

[0161] As a specific embodiment of the present application, the server 12 is further used to divide the plurality of historical ultrasound image reports into a plurality of second clusters based on the location of the ultrasound image;

[0162] And, based on the multiple second clusters, a third cluster is obtained; the third cluster is any one of the multiple second clusters;

[0163] And, based on the word segmentation tool, a first word segmentation array corresponding to each historical ultrasound imaging report in the third cluster is obtained, and a second word segmentation array corresponding to each diagnostic opinion in the diagnostic opinion classification standard is extracted;

[0164] And, matching each first word segmentation array with each second word segmentation array to obtain a first matching degree corresponding to each second word segmentation array;

[0165] and, dividing each historical ultrasound imaging report into a plurality of fourth clusters according to the diagnosis opinions according to the first matching degree;

[0166] And, based on the plurality of fourth clustering clusters, a plurality of clustering clusters are obtained.

[0167] As a specific embodiment of the present application, the server 12 is further used to obtain the segmented words in all the historical ultrasound imaging reports in the first cluster based on the segmentation tool to form a third segmented word array;

[0168] And, obtaining a fourth word segmentation array corresponding to each historical ultrasound imaging report in the first cluster based on the word segmentation tool;

[0169] And, based on the third word segmentation array and the fourth word segmentation array, the importance data of the historical ultrasound image report corresponding to the fourth word segmentation array is obtained.

[0170] As a specific embodiment of the present application, the server 12 is further configured to obtain a first word segmentation based on the fourth word segmentation array; the first word segmentation is any word segmentation in the fourth word segmentation array;

[0171] And, based on the first word segmentation, obtaining the number of the first word segmentation located in the third word segmentation array;

[0172] And, based on the number of segmented words and the number of historical ultrasound image reports in the first cluster, importance data of the historical ultrasound image reports corresponding to the fourth segmented word array is obtained.

[0173] As a specific embodiment of the present application, the server 12 is further used to obtain the importance data of the historical ultrasound image report corresponding to the fourth word segmentation array using the following calculation formula:

[0174]

[0175] in, represents the importance data of the k-th historical ultrasound image report in the first cluster; V represents the number of segmented words in the fourth segmented word array corresponding to the k-th historical ultrasound image report; represents the number of words in the third word array where the i-th word in the fourth word array is located; Represents the number of historical ultrasound image reports in the first cluster.

[0176] As a specific solution in the technical solution of the present application, the server 12 is further used to obtain, based on a word segmentation tool, a fifth word segmentation array of ultrasound image parts and diagnostic opinions in the text data and a sixth word segmentation array corresponding to each cluster;

[0177] And, based on the fifth word segmentation array and each sixth word segmentation array, obtaining a second matching degree between the text data and each clustering cluster;

[0178] And, based on each second matching degree, a target cluster is obtained.

[0179] As a specific embodiment of the present application, the server 12 is further used to obtain the seventh word segmentation array of the text data based on the word segmentation tool, and sequentially obtain the eighth word segmentation array of each historical ultrasound imaging report in the target cluster;

[0180] And, based on the seventh word segmentation array and the eighth word segmentation array, obtaining a third matching degree between the text data and each historical ultrasound image report in each of the target clusters;

[0181] And, based on each of the third matching degrees, the target matching degree is obtained.

[0182] As a specific embodiment of the present application, the meeting of the preset conditions includes:

[0183] If a target matching degree is greater than or equal to a preset threshold, it is considered that the target matching degree meets the preset condition.

[0184] It should be clear that the medical ultrasound image report structured generation device proposed in the embodiment of the present application can obtain the diagnosis opinions in multiple ultrasound image reports based on the historical ultrasound image reports, and sort all ultrasound image report templates in each diagnosis opinion class in order of importance from large to small. When the medical ultrasound image report is subsequently structured and generated based on the doctor's real-time voice data, templates with higher importance can be retrieved first, which can improve the matching speed while maintaining the matching accuracy of the ultrasound image templates.

[0185] After introducing the medical ultrasound image report structured generation device proposed in the embodiment of the present application, the computer-readable storage medium proposed in the embodiment of the present application is introduced below, on which a computer program is stored. When the computer program is executed by the processor, it implements the medical ultrasound image report structured generation method as described in any of the above embodiments.

[0186] It should be clear that the computer-readable storage media in this application include permanent and non-permanent, removable and non-removable media that can be used to store information by any method or technology. Information can be computer-readable instructions, data structures, modules of programs or other data. Examples of computer storage media include, but are not limited to, phase change memory, static random access memory, dynamic random access memory, other types of random access memory, read-only memory, electrically erasable programmable read-only memory, flash memory or other memory technology, read-only compact disc read-only memory, digital versatile disc or other optical storage, magnetic cassettes, disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include temporary computer-readable media, such as modulated data signals and carrier waves.

[0187] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0188] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the methods, devices and equipment described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0189] In the several embodiments provided in the embodiments of the present application, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the modules is only a logical function division. There may be other division methods in actual implementation, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or modules, which can be electrical, mechanical or other forms.

[0190] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, that is, they may be located in one place or distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0191] In addition, each functional module in each embodiment of the present application can be integrated into a processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The above-mentioned integrated module can be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium.

[0192] In the above embodiments, all or part of the embodiments may be implemented by software, hardware, firmware or any combination thereof. When implemented by software, all or part of the embodiments may be implemented in the form of a computer program product.

[0193] The computer program product includes one or more computer instructions. When the computer program is loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a server or data center that includes one or more available media integrated therein. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a digital video disk), or a semiconductor medium (e.g., a solid-state drive (SSD)).

[0194] Although the embodiments of the present application have been shown and described, it will be appreciated by those skilled in the art that many changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles of the present application.

Claims

1. A method for generating a structured medical ultrasound image report, characterized in that: include: Obtain the physician's spoken voice and multiple historical ultrasound imaging reports; Based on the ultrasound image location and diagnostic opinion classification criteria, multiple historical ultrasound image reports are divided into multiple clusters; Based on the multiple clusters, a first cluster is obtained; the first cluster is any one of the multiple clusters; Based on the first cluster, obtaining importance data of each historical ultrasound imaging report in the first cluster; Sorting each historical ultrasound imaging report in the first cluster according to the order of importance data from large to small; Based on the spoken voice, obtaining text data; Matching the ultrasound image parts and diagnostic opinions in the text data with each cluster to obtain a target cluster; Sequentially obtaining the target matching degree between the text data and each historical ultrasound image report in the target cluster; If a target matching degree meets the preset conditions, the historical ultrasound image report corresponding to the target matching degree is obtained as a template to generate the current ultrasound image report; The ultrasound image-based location and diagnostic opinion classification standard divides multiple historical ultrasound image reports into multiple clusters, including: Based on the location of the ultrasound image, the multiple historical ultrasound image reports are divided into multiple second clusters; Based on the multiple second clusters, a third cluster is obtained; the third cluster is any one of the multiple second clusters; Based on the word segmentation tool, a first word segmentation array corresponding to each historical ultrasound imaging report in the third cluster is obtained, and a second word segmentation array corresponding to each diagnostic opinion in the diagnostic opinion classification standard is extracted; Match each first word segmentation array with each second word segmentation array to obtain a first matching degree corresponding to each second word segmentation array; According to the first matching degree, each historical ultrasound imaging report is divided into a plurality of fourth clusters according to the diagnosis opinions; Based on the plurality of fourth clustering clusters, obtaining a plurality of clustering clusters; The acquiring, based on the first cluster, importance data of each historical ultrasound imaging report in the first cluster comprises: Based on the word segmentation tool, the word segmentations in all the historical ultrasound imaging reports in the first cluster are obtained to form a third word segmentation array; Acquire a fourth word segmentation array corresponding to each historical ultrasound imaging report in the first cluster based on a word segmentation tool; Based on the third word segmentation array and the fourth word segmentation array, acquiring importance data of the historical ultrasound image report corresponding to the fourth word segmentation array; The acquiring, based on the third word segmentation array and the fourth word segmentation array, importance data of the historical ultrasound imaging report corresponding to the fourth word segmentation array includes: Based on the fourth word segmentation array, obtain a first word segmentation; the first word segmentation is any word segmentation in the fourth word segmentation array; Based on the first word segmentation, obtain the number of the first word segmentation in the third word segmentation array; Based on the number of segmented words and the number of historical ultrasound image reports in the first cluster, acquiring importance data of the historical ultrasound image reports corresponding to the fourth segmented word array; The calculation formula for obtaining the importance data of the historical ultrasound image reports corresponding to the fourth word segmentation array based on the number of segmented words and the number of historical ultrasound image reports in the first cluster is as follows: in, represents the importance data of the k-th historical ultrasound image report in the first cluster; V represents the number of segmented words in the fourth segmented word array corresponding to the k-th historical ultrasound image report; represents the number of words in the third word array where the i-th word in the fourth word array is located; Represents the number of historical ultrasound imaging reports in the first cluster.

2. The method for generating a structured medical ultrasound image report according to claim 1, characterized in that: The step of matching the ultrasound image parts and diagnostic opinions in the text data with each cluster to obtain a target cluster includes: Based on the word segmentation tool, a fifth word segmentation array of ultrasound image parts and diagnostic opinions in the text data and a sixth word segmentation array corresponding to each cluster are obtained; Based on the fifth word segmentation array and each sixth word segmentation array, obtaining a second matching degree between the text data and each clustering cluster; Based on each second matching degree, a target cluster is obtained.

3. The method for generating a structured medical ultrasound image report according to claim 1, characterized in that: The step of sequentially obtaining the target matching degree between the text data and each historical ultrasound image report in the target cluster includes: Based on the word segmentation tool, a seventh word segmentation array of the text data is obtained, and an eighth word segmentation array of each historical ultrasound imaging report in the target cluster is obtained in sequence; Based on the seventh word segmentation array and the eighth word segmentation array, obtaining a third matching degree between the text data and each historical ultrasound image report in each target cluster; Based on each of the third matching degrees, the target matching degree is obtained.

4. The method for generating a structured medical ultrasound image report according to claim 1, characterized in that: The preset conditions include: If a target matching degree is greater than or equal to a preset threshold, it is considered that the target matching degree meets the preset condition.

5. A medical ultrasound image report structured generation device, the device being used to implement the medical ultrasound image report structured generation method as claimed in any one of claims 1 to 4, characterized in that: include: An acquisition device for acquiring a doctor's spoken voice and multiple historical ultrasound image reports; A server, used for dividing a plurality of historical ultrasound image reports into a plurality of clusters based on the location of the ultrasound image and the classification standard of the diagnosis opinion; And, based on the multiple clusters, a first cluster is obtained; the first cluster is any one of the multiple clusters; And, based on the first cluster, obtaining importance data of each historical ultrasound imaging report in the first cluster; And, sorting each historical ultrasound imaging report in the first cluster according to the order of importance data from large to small; and, based on the spoken speech, obtaining text data; Matching the ultrasound image parts and diagnostic opinions in the text data with each cluster to obtain a target cluster; and, sequentially obtaining the target matching degree between the text data and each historical ultrasound image report in the target cluster; Furthermore, if a target matching degree meets a preset condition, a historical ultrasound image report corresponding to the target matching degree is obtained as a template to generate a current ultrasound image report.

6. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed by a processor, the method for generating a structured medical ultrasound image report according to any one of claims 1 to 4 is implemented.

Images