Medical report generation method and device, storage medium and electronic equipment
By generating medical report generation method, the initial diagnostic report and user input information are used to adjust it, which solves the problem that doctors take too long to write medical reports and improves the efficiency of report generation.
Patent Information
- Application Number
- CN202411217057.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-30
- Publication Date
- 2024-11-26
AI Technical Summary
When writing medical reports, doctors need to combine the patient's oral statements and medical examination results, which takes a long time and reduces the efficiency of diagnosis and treatment.
Provide a method for generating medical reports, by determining and displaying the initial diagnostic report, obtaining user input information, predicting user subsequent input, adjusting initial diagnostic report, generating medical reports, reducing user input, and improving report generation efficiency.
By automatically generating part of the report content, users' workload in the process of generating medical reports is reduced and the efficiency of generating diagnostic reports is improved.
Smart Images

Figure CN119028510A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of computer technology, and in particular to a method, device, storage medium and electronic device for generating a medical report. Background Art
[0002] Usually, during the diagnosis and treatment process of a patient, a doctor needs to manually write a medical report based on the patient's oral statement, medical tests, medical images, etc. The writing of medical reports takes up a long time of the doctor's working time, so that the doctor cannot spend more time on the diagnosis and treatment process of the patient, which reduces the doctor's diagnostic efficiency.
[0003] Therefore, the present specification provides a method for generating a medical report. Summary of the invention
[0004] The present specification provides a medical report generation method, device, storage medium and electronic device to at least partially solve the above-mentioned problems existing in the prior art.
[0005] This manual adopts the following technical solutions:
[0006] This specification provides a method for generating a medical report, comprising:
[0007] Determine initial diagnostic report and present;
[0008] Acquire input information input by a user based on the initial diagnosis report, predict a subsequent input of the user according to the input information, and determine prediction information;
[0009] The initial diagnosis report is adjusted according to the input information and the prediction information to determine a medical report.
[0010] Optionally, determine and present an initial diagnostic report, including:
[0011] Determine report templates containing pre-set diagnostic text for each body part;
[0012] According to the report template, an initial diagnosis report is determined, and the initial diagnosis report is displayed.
[0013] Optionally, determining an initial diagnostic report according to the report template specifically includes:
[0014] Acquire a medical image and determine a region of interest in the medical image;
[0015] Among the human body parts, determining the human body part corresponding to the region of interest as the focus part;
[0016] Determining an initial diagnosis text of the part of interest according to the region of interest;
[0017] An initial diagnosis report is determined based on the initial diagnosis text of the concerned part, and the initial diagnosis report is displayed.
[0018] Optionally, predicting a subsequent input of the user according to the input information and determining the prediction information specifically includes:
[0019] Predicting the user's subsequent input according to the input information to determine each predicted candidate information;
[0020] Displaying each candidate information to the user;
[0021] In response to the user's selection operation on each piece of candidate information, the candidate information selected by the user is used as prediction information.
[0022] Optionally, presenting the candidate information to the user specifically includes:
[0023] respectively determining a matching probability between each candidate information and the input information;
[0024] According to the magnitude of each matching probability, the candidate information corresponding to each matching probability is sorted to determine the sorted candidate information;
[0025] The sorted candidate information is displayed to the user.
[0026] Optionally, adjusting the initial diagnosis report according to the input information and the prediction information to determine a medical report specifically includes:
[0027] Determining adjustment information according to the input information and the prediction information;
[0028] In the initial diagnosis report, determining a text segment corresponding to the adjustment information;
[0029] The text segment is adjusted according to the adjustment information.
[0030] Optionally, the input information is voice information or text information;
[0031] Determining adjustment information according to the input information and the prediction information specifically includes:
[0032] When the input information is voice information, determining text information corresponding to the voice information, integrating the text information with the prediction information, and determining adjustment information;
[0033] When the input information is text information, the text information is integrated with the prediction information to determine the adjustment information.
[0034] Optionally, adjusting the initial diagnosis report according to the input information and the prediction information to determine a medical report specifically includes:
[0035] Determining adjustment information according to the input information and the prediction information;
[0036] Determine, among the human body parts, a human body part corresponding to the adjustment information as a target part;
[0037] The initial diagnosis text of the target part in the initial diagnosis report is adjusted according to the adjustment information to determine the medical report.
[0038] Optionally, the number of the target sites is multiple;
[0039] According to the adjustment information, the initial diagnosis text of the target part in the initial diagnosis report is adjusted, specifically including:
[0040] In the adjustment information, the adjustment content of each target part is determined respectively;
[0041] For each target part, the initial diagnosis text of the target part is adjusted according to the adjustment content of the target part.
[0042] Optionally, according to the adjustment content of the target part, the initial diagnosis text of the target part is adjusted, specifically including:
[0043] When the adjustment content of the target part contains keywords representing deletion semantics, the initial diagnosis text of the target part is deleted;
[0044] When the adjustment content of the target part does not contain the keyword representing the deletion semantics, the adjustment content of the target part is integrated with the initial diagnosis text of the target part.
[0045] Optionally, the adjustment content of the target part is integrated with the initial diagnosis text of the target part, specifically including:
[0046] Performing semantic analysis on the adjustment content of the target part and the initial diagnosis text of the target part;
[0047] When it is determined that the adjusted content of the target part is inconsistent with the semantics of the initial diagnosis text of the target part, the initial diagnosis text of the target part is replaced with the adjusted content of the target part;
[0048] When it is determined that the semantics of the adjustment content of the target part and the initial diagnosis text of the target part are repeated, the adjustment content of the target part and the initial diagnosis text of the target part are merged.
[0049] Optionally, the initial diagnosis report is used to describe the imaging information of the medical image;
[0050] Adjusting the initial diagnosis report according to the input information and the prediction information to determine a medical report specifically includes:
[0051] Adjusting the initial diagnosis report according to the input information and the prediction information;
[0052] Inputting the adjusted initial diagnosis report into a third language model to generate an examination conclusion of the medical image;
[0053] A medical report is determined based on the adjusted initial diagnosis report and the examination conclusion of the medical image.
[0054] Optionally, after determining the medical report, the method further comprises:
[0055] Acquiring patient information, wherein the patient information includes at least one of basic information, chief complaint, relevant examination results, and medical history;
[0056] Determining whether there is any mismatch between the medical report and the patient information;
[0057] If yes, determining prompt information according to the sentence corresponding to the mismatching content in the medical report, wherein the prompt information is used to prompt the user to check the mismatching content again;
[0058] If not, the medical report is displayed to the user.
[0059] This specification provides a medical report generating device, comprising:
[0060] An initial diagnosis report determination module determines and displays the initial diagnosis report;
[0061] An input module, which obtains input information input by a user based on the initial diagnosis report, predicts a subsequent input of the user according to the input information, and determines prediction information;
[0062] The medical report determination module adjusts the initial diagnosis report according to the input information and the prediction information to determine the medical report.
[0063] The present specification provides a computer-readable storage medium, wherein the storage medium stores a computer program, and when the computer program is executed by a processor, the above-mentioned medical report generation method is implemented.
[0064] The present specification provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned medical report generating method when executing the program.
[0065] At least one of the above technical solutions adopted in this specification can achieve the following beneficial effects:
[0066] In the medical report generation method provided in the present specification, an initial diagnosis report is determined and displayed, input information input by a user based on the initial diagnosis report is obtained, and based on the input information, subsequent input by the user is predicted, and prediction information is determined. Based on the input information and the prediction information, the initial diagnosis report is adjusted to determine the medical report. Through this method, in the process of determining the medical report, based on the partial information input by the user for adjusting the initial diagnosis report, the prediction information that the user continues to input for adjusting the initial diagnosis report can be automatically generated, and the initial diagnosis report is adjusted based on the input information and the prediction information. It can be seen that when determining the medical report, this method only requires the user to make a small amount of input, which reduces the workload of the user in the process of generating the medical report and improves the efficiency of generating the medical report. BRIEF DESCRIPTION OF THE DRAWINGS
[0067] The drawings described herein are used to provide a further understanding of this specification and constitute a part of this specification. The illustrative embodiments and descriptions of this specification are used to explain this specification and do not constitute an improper limitation on this specification. In the drawings:
[0068] Figure 1 A schematic diagram of a process for generating a medical report in this specification;
[0069] Figure 2 A schematic diagram of a prediction information generation process provided in an embodiment of this specification;
[0070] Figure 3 A schematic diagram of a prediction information generation process provided in an embodiment of this specification;
[0071] Figure 4 This is a schematic diagram of an initial diagnostic report provided in an embodiment of this specification;
[0072] Figure 5 A schematic diagram of a medical report generation process provided in an embodiment of this specification;
[0073] Figure 6 A voice-triggered optimization instruction process diagram provided in an embodiment of this specification;
[0074] Figure 7 A medical report generation flow chart provided in an embodiment of this specification;
[0075] Figure 8 A schematic diagram of a medical report generating device provided in this specification;
[0076] Fig. 9 The corresponding Figure 1 Schematic diagram of electronic equipment. DETAILED DESCRIPTION
[0077] In order to make the purpose, technical solutions and advantages of this specification more clear, the technical solutions of this specification will be clearly and completely described below in combination with the specific embodiments of this specification and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.
[0078] The technical solutions provided by the embodiments of this specification are described in detail below in conjunction with the accompanying drawings.
[0079] Figure 1 The following is a flow chart of a method for generating a medical report in this specification, which specifically includes the following steps:
[0080] S100: Determine and display an initial diagnostic report.
[0081] All steps in the medical report generation method provided in this specification can be implemented by any electronic device with computing functions, such as a terminal, a server, etc. For ease of description, the medical report generation method provided in this specification is described below with the server as the execution subject.
[0082] Due to the need for standardization, there are usually many relatively fixed description sentences in medical reports, such as the diagnosis conclusion of normal human body parts, the commonly used diagnosis conclusions in specialist clinics, etc. Therefore, these relatively fixed description sentences can be used as preset diagnosis texts, and the preset report template can be determined based on the preset diagnosis texts of various human body parts.
[0083] This manual does not limit the specific content of each human body part included in the report template. It can be set to any human body position according to actual diagnostic needs, such as any human organ such as the chest, lungs, spine, or a component structure in the human organ, such as the left lung, lymph, etc.
[0084] The number of preset report templates is not limited in this specification. In this method, only one preset report template may be set, or multiple report templates may be set according to structural or functional similarities. When the user determines the medical report through the method provided in this specification, one or more appropriate report templates may be selected according to actual diagnostic needs. Alternatively, in one or more embodiments of this specification, the server may also support user-defined report templates to meet diagnostic needs, that is, which human body parts are included in the report template, and the preset diagnostic texts for each human body part can be set by the user.
[0085] The server may determine a preset report template for the medical report generation process based on a pre-stored report template or a user-defined report template, and determine an initial diagnosis report based on preset diagnosis texts for various body parts contained in the report template.
[0086] There are many ways to determine the initial diagnostic report. In one or more embodiments of the present specification, the server directly uses the report template as the initial diagnostic report and displays the initial diagnostic report to the user so that the user can view and adjust the initial diagnostic report.
[0087] S102: Obtain input information input by the user based on the initial diagnosis report, predict the user's subsequent input according to the input information, and determine prediction information.
[0088] In the medical report generation method provided in this specification, after the server displays the initial diagnosis report to the user, the user adjusts the initial diagnosis report by inputting information according to the actual diagnosis situation, and thus obtains the medical report.
[0089] The input information in this manual can be voice information or text information, and the user can select a suitable input method for input as needed.
[0090] During the user input process, the server predicts the user's subsequent input based on the user's input information and determines the prediction information.
[0091] The prediction information is a prediction of the subsequent content of the input information based on the semantics of the input information. The generation process of the prediction information can be implemented through language models in the field of natural language processing, such as the GPT series model, the BERT model, etc. The server can use a large amount of corpus of commonly used diagnostic conclusions in the medical field as sample data to train the above language model to obtain a first language model for subsequent text prediction tasks. The first language model can understand the semantics of the user's input information and predict the subsequent text of the input information.
[0092] Specifically, the server inputs the user's input information into the first language model in real time, so that the first language model predicts subsequent content of the input information, and obtains prediction information output by the first language model.
[0093] The content after the input information and the predicted information are integrated can be used to adjust the initial diagnosis report. Through this method, the user no longer needs to write a medical report from scratch, but only needs to input a small amount of previous content for adjusting the medical report to realize the generation of the medical report, which greatly reduces the workload of the user in the process of generating the medical report.
[0094] S104: Adjust the initial diagnosis report according to the input information and the prediction information to determine a medical report.
[0095] The server may integrate the user's input information and prediction information, determine adjustment information, and use the integrated adjustment information to adjust the initial diagnosis report.
[0096] Specifically, when the input information is voice information, the server converts the voice information into text information, integrates the text information and the adjustment information according to the semantics of the text information and the semantics of the prediction information, and determines the adjustment information. When the input information is text information, the server integrates the input text information and the prediction information according to the semantics of the input text information and the semantics of the prediction information, and determines the adjustment information.
[0097] The server then performs semantic analysis on the adjustment information, and based on the semantics of the adjustment information, determines a text segment in the initial diagnosis report that is consistent with the semantics of the adjustment information, adjusts the text segment based on the adjustment information, and determines a medical report based on the adjusted initial diagnosis report.
[0098] Semantic analysis is the identification and analysis of the contextual semantics of natural language text. Through the results of semantic analysis, many different types of natural language processing tasks can be performed. The medical report generation method in this specification involves multiple semantic analysis steps, and each semantic analysis step corresponds to a different natural language processing task. The semantic analysis process can be implemented through a pre-trained second language model, which can perform semantic analysis according to different task prompts and perform natural language processing tasks corresponding to the task prompts.
[0099] The semantic analysis operation is used to determine the text segment consistent with the adjustment information in the initial diagnosis report. The text segment can be one or several words, one or several sentences in the initial diagnosis report, and why the specific content of the text segment is related to the semantics of the adjustment information.
[0100] The adjustment operation here may include three types: deletion, addition, and partial modification of the original content. Among them, the addition operation and the partial modification operation of the original content are the semantic integration process of the adjustment information and the text segment corresponding to the adjustment information. The deletion operation is an operation performed on the text segment corresponding to the adjustment information, and does not require semantic integration.
[0101] Therefore, in the adjustment operation of this specification, the server performs semantic analysis on the adjustment information to determine whether the adjustment information is a deletion operation. When the adjustment information is not a deletion operation, the server performs semantic analysis on the adjustment information again, and semantically integrates the adjustment information with the text segment corresponding to the adjustment information according to the semantics of the adjustment information. The first semantic analysis operation in this step is used to identify the deletion operation, and the second semantic analysis operation is used to semantically integrate the adjustment information with the text segment corresponding to the adjustment information according to the semantics of the adjustment information.
[0102] Specifically, the server determines whether the adjustment information contains keywords representing the deletion semantics. If so, the text segment in the initial diagnosis report is deleted. If not, the adjustment information and the text segment are further semantically analyzed. When it is determined that there is a contradiction between the adjustment information and the text segment, the text segment is replaced with the adjustment information. When it is determined that there is a duplication in the semantics between the adjustment information and the text segment, the adjustment information is merged with the text segment.
[0103] In this embodiment, because the prediction information is a prediction of the subsequent content of the input information, it is difficult to guarantee absolute accuracy. In order to improve the accuracy of the adjustment of the medical report, the server can first display the prediction information to the user, respond to the user's confirmation operation on the prediction information, and then determine the adjustment information based on the prediction information and the input information for adjusting the initial diagnosis report.
[0104] The initial diagnosis report in this manual is determined based on the report template. The content of the report template is already a fixed description commonly used in the diagnosis scenario. Usually, in the actual diagnosis process, the patient has only a few parts with lesions or abnormal conditions. In this case, the user does not need to modify most of the content of the initial diagnosis report. It is only necessary to adjust the text fragments corresponding to the parts of the patient with lesions or abnormal conditions in the initial diagnosis report. Therefore, the content that the user needs to adjust in the initial diagnosis report is usually only a small amount of content in the initial diagnosis report, which improves the efficiency of generating medical reports.
[0105] By applying the medical report generation method in this specification, the user only needs to enter part of the previous content in the adjustment information used to adjust the initial diagnosis report, and the server can automatically predict the user's subsequent input. The user does not need to enter all the adjustment information. During the user input process, the efficiency of medical report generation is further improved by generating predictive information.
[0106] In the above step S102, in order to improve the accuracy of the predicted information, in one or more embodiments of this specification, the server may predict the user's subsequent input based on the input information, determine multiple candidate information, and display each candidate information to the user. This specification does not limit the number of candidate information, which can be set according to needs.
[0107] Then, in response to the user's selection operation on each candidate information, the candidate information selected by the user is used as prediction information to adjust the initial diagnosis report.
[0108] Furthermore, in order to improve the user's decision efficiency in determining the prediction information from each candidate information, in another embodiment, the server may respectively determine the matching probability between each candidate information and the input information.
[0109] Each candidate information in this embodiment is also output by the first language model in the above step S102. When the server obtains the first language model to predict each candidate information, it obtains the confidence of each candidate information as the matching probability of each candidate information and the input information. The confidence of a candidate information is the probability that the first language model considers that the candidate information is consistent with the subsequent content of the actual input information during the model inference process.
[0110] Then, according to the size of each matching probability, the candidate information corresponding to each matching probability is sorted, the sorted candidate information is determined, and the sorted candidate information is displayed to the user. Among them, the sorting method can be selected from high to low or from low to high according to the matching probability as needed.
[0111] Because the candidate information with the highest matching probability has the highest corresponding confidence, it means that when the first language model is reasoning, the greater the possibility that the candidate information is consistent with the content of the actual input information, the more likely the candidate information is to be selected by the user. Through this embodiment, the user can notice the candidate information with the highest matching probability more quickly and make a selection, thereby further improving the efficiency of generating medical reports.
[0112] Figure 2 This is a schematic diagram of a prediction information generation process provided in an embodiment of this specification. Figure 2 As shown in the figure, when the input information is "left lower lobe posterior basal segment", the server generates 4 candidate information, namely "solid nodule shadow", "ground glass nodule shadow", "linear shadow", and "patchy high-density shadow". Figure 2 As shown on the right, the server displays the four generated candidate information to the user. If the user selects the candidate information "solid nodule shadow", the server determines that "solid nodule shadow" is the prediction information for adjusting the initial diagnosis report.
[0113] In one or more embodiments of the present specification, the prediction information may also include a missing value, which is a numerical value related to the actual lesion or abnormal area morphology.
[0114] The server responds to the user's selection operation on each candidate information, and after determining the prediction information, the user enters the missing values in the prediction information according to the actual diagnosis situation to obtain complete prediction information, and integrates the complete prediction information and the input information for adjusting the initial diagnosis report.
[0115] Figure 3 This is a schematic diagram of a prediction information generation process provided in an embodiment of this specification. Figure 3 As shown in the figure, when the input information is "a solid nodule is seen in the posterior basal segment of the left lower lobe of the lung", the server generates two candidate information, namely "size is about cm×cm" and "longest diameter is about cm". Figure 3 As shown on the right, the server displays the two generated candidate information to the user. The user selects the candidate information "size is approximately cm×cm", and the server determines that "size is approximately cm×cm" is the predicted information.
[0116] In one or more embodiments of the present specification, because each initial diagnosis text in the report template is divided according to the body part, in the above step S104, before adjusting the initial diagnosis report, the server can first perform part identification to determine the body part corresponding to the input information and the predicted information, and adjust the initial diagnosis text for the corresponding body part to improve the accuracy of the medical report.
[0117] The server determines the adjustment information based on the input information and the predicted information. And according to the semantics of the adjustment information, the human body part corresponding to the adjustment information is determined as the target part among the human body parts included in the initial diagnosis report. According to the adjustment information, the initial diagnosis text of the target part in the initial diagnosis report is adjusted to determine the medical report.
[0118] This specification does not limit the number of target parts, that is, the input information input by the user at one time may only include the adjustment content of the initial diagnosis text of one target part, or may include the adjustment content of the initial diagnosis text of multiple target parts.
[0119] When there are multiple target parts, the server performs semantic analysis on the adjustment information, and determines the adjustment content of each target part according to the semantics of the adjustment information. Then, for each target part, the initial diagnosis text of the target part is adjusted according to the adjustment content of the target part.
[0120] The operation of adjusting the initial diagnosis text of the target part is similar to the adjustment operation in the above step S104, and may also include three types: deletion, addition and partial modification of the original content.
[0121] Specifically, the server determines whether the adjustment content of the target part contains keywords representing the deletion semantics.
[0122] When the adjustment content of the target part contains keywords representing the deletion semantics, the initial diagnosis text of the target part is deleted.
[0123] When the adjustment content of the target part does not contain the keyword representing the deletion semantics, the adjustment content of the target part is integrated with the initial diagnosis text of the target part.
[0124] In order to perform the integration operation more accurately, the server may continue to perform semantic analysis on the adjusted content of the target part to determine whether the adjusted content of the target part is semantically contradictory or semantically repeated with the initial diagnosis text of the target part.
[0125] When the server determines that the adjusted content of the target part is inconsistent with the semantics of the initial diagnosis text of the target part, the adjusted content of the target part replaces the initial diagnosis text of the target part. When the server determines that the adjusted content of the target part is repeated with the semantics of the initial diagnosis text of the target part, the adjusted content of the target part is merged with the initial diagnosis text of the target part.
[0126] In one or more embodiments of the present specification, in order to improve the generation efficiency of the medical report generation method of the present specification, in the above step S102, before the server determines the prediction information, the input information may be first identified to determine the type of the input information. Here, the type of the input information includes a report type and a non-report type.
[0127] Especially when the input information is voice information, the user's voice input is easily affected by environmental factors. For example, in a noisy environment, the server may record idle conversations, background noise and other content that is irrelevant to report generation, interfering with the generation of medical reports.
[0128] When the server determines that the input information is a report type, the subsequent content of the input information is predicted. When the server determines that the voice information is not a report type, the voice information is discarded and no further processing is performed on the voice information, thereby saving computing resources.
[0129] Through this embodiment, the server can filter out non-report input information that the user may input in advance, so as to avoid affecting the accuracy of the report after the server generates irrelevant input information based on the medical report to determine the medical report.
[0130] In one or more embodiments of the present specification, to further facilitate user operations, the input information input by the user may be instruction information. The user sends instructions to the server through the input information, and the server performs corresponding operations according to the instructions sent by the user.
[0131] The type of input information may include report type, non-report type and instruction type. When the server determines that the input information is an instruction type, it determines the preset instruction that matches the input information among the preset instructions as the target instruction and executes the operation corresponding to the target instruction.
[0132] Specifically, the server can determine the keywords of each preset instruction, and then determine the text information corresponding to the input information, and judge whether the text information matches the semantics of the keyword of any preset instruction. If so, the preset instruction corresponding to the keyword matching the text information is used as the target instruction.
[0133] For example, in one scenario, after the user completes the adjustment of the initial diagnosis text of each body part, it is usually necessary to manually operate the page to send a confirmation instruction to the server, and the server determines to generate a medical report in response to the user's confirmation instruction. If the keyword of the confirmation instruction is "generate report", the server determines that the user's input information contains the word "generate report" or a word with the same semantics as "generate report", and the server performs the operation of determining to generate a medical report.
[0134] The language in medical reports usually needs to be standardized professional terms, but the user's input information may be non-standard, such as the presence of non-professional terms, omitted descriptions, etc.
[0135] When the input information is voice information, the input information may also contain some spoken content, such as repeated words, modal particles, pause words, etc.
[0136] Therefore, in one or more embodiments of the present specification, before the server adjusts the initial diagnosis report in the above step S104, it can perform normalization processing on the input information, and the normalization processing includes removing repeated words, modal particles, pause words, converting non-medical professional terms into corresponding medical professional terms, etc., so that after adjusting the initial diagnosis report through the input information and the predicted information, a more standardized medical report can be determined.
[0137] For example, if the user's input information is "trachea is normal", the input information is an abbreviated description, and the normalized input information determined by this embodiment is "trachea and bronchus are unobstructed". If the user's input information is "thoracic vertebrae, oh no, lumbar vertebrae bone hyperplasia", the input information contains the pause word "oh no", and the normalized input information determined by this embodiment is "lumbar vertebrae bone hyperplasia".
[0138] In one or more embodiments of the present specification, the initial diagnosis text of a human body part in the initial diagnosis report includes the diagnosis results of different positions in the human body part. For example, if a human body part in the initial diagnosis report is a lung, the initial diagnosis text of the human body part may include the diagnosis result of the left lung and the diagnosis result of the right lung.
[0139] When the user adjusts the initial diagnosis report by inputting information, only the initial diagnosis text of the target area including part of the diagnosis results or individual words may be adjusted, and the entire content of the initial diagnosis text will not be adjusted.
[0140] After the server has identified the part, it can determine the text segment corresponding to the adjustment information in the initial diagnosis text of the target part according to the method in step S104, and adjust the text segment. The adjustment here is the same as the adjustment in step S104, including deletion, addition and partial modification of the original content.
[0141] For example, in the initial diagnosis report, a human body part "thoracic cavity / pleura / diaphragm" is included, and the initial diagnosis text of the human body part is: no effusion is found in the bilateral thoracic cavities; no obvious thickening is found in the bilateral pleura. It can be seen that the initial diagnosis text contains the diagnosis results of two different structural areas, the thoracic cavity and the pleura.
[0142] If the input information is "slightly thickened right pleura", then through the method in this embodiment, the server can use "thoracic cavity / pleura / diaphragm" as the target part, and in the initial diagnosis text of the human body part "no effusion in bilateral thoracic cavities; no obvious thickening in bilateral pleura", determine "no obvious thickening in bilateral pleura" as the text segment corresponding to the input information. Because the semantics of the input information and the semantics of the text segment are inconsistent, the server replaces the text segment with the input information, and the adjusted initial diagnosis text of the target part is "no effusion in bilateral thoracic cavities; slightly thickened right pleura".
[0143] If the input information is "delete the description about the right pleura", then through the method in this embodiment, the server can use "thoracic cavity / pleura / diaphragm" as the target part, and in the initial diagnosis text of the human body part "no effusion in bilateral thoracic cavities; no obvious thickening in bilateral pleura", determine "no obvious thickening in bilateral pleura" as the text segment corresponding to the input information. Because the input information contains the semantics of deletion, the server deletes the text segment according to the input information, and the adjusted initial diagnosis text of the target part is "no effusion in bilateral thoracic cavities".
[0144] In one or more embodiments of the present specification, the medical report generation method provided in the present specification can also be used to generate medical reports of medical images.
[0145] Then in the above step S100, the server may also determine the initial diagnosis report according to the following method.
[0146] First, the server acquires the medical image and determines the region of interest in the medical image.
[0147] The specific type of medical images is not limited in this specification, and any type of medical images can be used. In this embodiment, the specific method of determining the region of interest is not limited, and it can be obtained by any machine learning model used for image processing tasks, or by manual annotation, or by both machine learning models and manual annotation.
[0148] Secondly, the server determines the human body part corresponding to the region of interest from among the human body parts as the focus part. The human body part corresponding to the region of interest is the region of interest marked in the image area of the human body part.
[0149] Due to the characteristics of medical images themselves, the number of regions of interest determined by the server may be one or more, and the number of corresponding determined sites of interest may also be one or more.
[0150] Then, the server determines the initial diagnosis text of the concerned part according to the region of interest, and determines the initial diagnosis report according to the initial diagnosis text of the concerned part. The initial diagnosis report is composed of the initial diagnosis text of each concerned part.
[0151] Because, in this embodiment, not all human body parts are necessarily marked with regions of interest, the determined regions of interest are only some of all human body parts, and the initial diagnosis report only includes the initial diagnosis text of these parts.
[0152] In another embodiment, in order to make the initial diagnosis report more comprehensive, the server treats the human body parts that do not correspond to the regions of interest as non-interested parts. The human body parts that do not correspond to the regions of interest are those that are not marked with regions of interest in the image area of the human body parts.
[0153] The server uses the preset diagnosis text of each non-concern part in the report template as the initial diagnosis text of each non-concern part. Then the initial diagnosis report includes the initial diagnosis text of each concern part determined according to the region of interest and the preset diagnosis text of each non-concern part.
[0154] Compared to the method in step S100 of determining the initial diagnostic text entirely based on a preset template, the initial diagnostic text of the area of concern determined according to the region of interest in this embodiment can make the content of the initial diagnostic report more accurate, thereby reducing the content that the user needs to adjust through voice information input by voice or text information input by keyboard, thereby further improving the efficiency of generating reports.
[0155] When the server determines the region of interest based on the machine learning model, in order to facilitate user viewing, the server may display the initial diagnosis text of the human body part corresponding to the region of interest determined by the machine learning model in the initial diagnosis report with a highlighted display effect to remind the user to check. This specification does not limit the specific form of the highlighted display effect here, and any display effect such as highlighting, bolding, flashing, etc. can be selected according to actual needs.
[0156] Combine the following Figure 4 , taking a CT chest plain scan scene as an example, the method of this embodiment is described. Figure 4 Schematic diagram of an initial diagnostic report provided in the embodiments of this specification. Figure 4 As shown, each human body part in the initial template includes 7 parts, namely, lungs, trachea, mediastinum, heart and great vessels, chest cavity / pleura / diaphragm, bones and others. The initial diagnosis text of each human body part is listed in the initial diagnosis report. Figure 4 The bold part, i.e., the initial diagnosis text corresponding to the lungs and bones, indicates the initial diagnosis text of the human body parts corresponding to the regions of interest determined by the machine learning model.
[0157] In the scenario of generating medical reports of medical images, it is usually necessary to first determine the image findings. Image findings are used to describe the directly observed image information. Compared with medical reports, the content of image findings is usually more professional. For example, the content of image findings can include professional image data such as density, signal strength, morphology, size, and position. After determining the image findings, the medical report is determined based on the image information corresponding to the image findings.
[0158] In one or more embodiments of the present specification, the preset diagnostic text for each body part in the report template may be image information describing the medical image. The initial diagnostic report determined in the above step S100 is used to describe the image information of the medical image, indicating the initial image content.
[0159] The user views the medical image, and based on the information observed in the medical image, verifies the content of the initial diagnosis report displayed by the server. When it is determined that there is content in the initial diagnosis report that needs to be adjusted, the user inputs voice information or text information. The server determines the prediction information based on the voice information or text information input by the user, and adjusts the initial diagnosis text of each part of the human body based on the voice information or text information input by the user and the generated prediction information.
[0160] In the scenario of this embodiment, the adjusted initial diagnosis text of each human body part determined by the server is used to describe the adjusted image information. Then, the server inputs the adjusted initial diagnosis text of each human body part into the third language model to generate the medical image examination conclusion. The server determines the medical report based on the adjusted initial diagnosis report and the generated medical image examination conclusion. The third language model in this embodiment can be any machine learning model trained for natural language processing tasks.
[0161] In this embodiment, because the content of the initial diagnosis report directly adjusted by the user represents the content of the initial image, after the user completes the adjustment of the initial diagnosis report, the server generates a medical report based on the adjusted initial diagnosis report through the third language model. In order to achieve a better user experience, after the user completes the adjustment of the initial diagnosis report, the server can generate a medical report in response to the user's confirmation instruction and display the generated medical report to the user.
[0162] For the convenience of user viewing, in one or more embodiments of the present specification, the medical report generated by the server may only include the diagnosis results indicating the presence of lesions or abnormalities in the patient.
[0163] According to the above content, the type of input information in this specification may include an instruction class. In this embodiment, the user's instruction to generate a medical report may be a manual operation of the user specifying an identifier on the graphical user page, or may be triggered by input information determined by the user's voice input or text input.
[0164] The following will be combined Figure 5 The method of this embodiment is described. Figure 5 A schematic diagram of a medical report generation process provided in an embodiment of this specification.
[0165] like Figure 5As shown, the upper figure is the adjusted initial diagnosis report, and the lower figure is the medical report, wherein the medical report includes the adjusted initial diagnosis report and the generated medical image examination conclusion. If the keyword of the determination instruction is "generate report", after the server determines the adjusted initial diagnosis report, in response to the input information input by the user containing the "generate report" semantics, it determines that the input information is a determination instruction. Then, the server executes the determination instruction, generates the medical image examination conclusion, and displays the adjusted initial diagnosis report and the generated medical image examination conclusion to the user as a medical report.
[0166] Figure 5 The diagnostic results of each part of the human body in the medical report are listed in a structured manner. It can be seen that the adjusted initial diagnosis report is not exactly the same as the examination conclusion of the medical image. The medical report only contains the diagnostic results that indicate the presence of lesions or abnormalities in the patient.
[0167] In one or more embodiments of the present specification, in order to further improve the accuracy of the medical report, after determining the medical report in the above step S104, the server may further check the determined medical report based on the patient information, and when it is found that the medical report contains content that does not match the patient information, the user is reminded so that the user can check the content of the medical report again.
[0168] The above-mentioned patient information may include any information related to diagnosis and treatment provided by the patient, such as basic information, chief complaint, relevant test results, medical history, etc.
[0169] Specifically, after determining the medical report, the server obtains the patient information and performs semantic analysis on the patient information and the medical report distribution. The semantic analysis operation is used to determine whether there is any mismatch between the medical report and the patient information. If so, the prompt information is determined based on the corresponding statement of the mismatch in the medical report. The mismatch may be content that the user has overlooked and needs to be adjusted in the initial diagnosis report.
[0170] This specification does not limit the specific form of the prompt information. It can be a highlighted statement of the non-matching content in the medical report, or a pop-up message, as long as it can prompt the user to re-check the non-matching content.
[0171] If not, the medical report is displayed to the user. It is determined that the medical report does not detect any problem, and the user is prompted that the medical report is in an optimal state.
[0172] When the corresponding sentence of the mismatched content in the medical report is highlighted, the server may use any highlighting effect to display the corresponding sentence of the mismatched content in the medical report, such as highlighting, bolding, flashing, etc.
[0173] For example, in the basic information of the patient information, the server can determine that the patient is male. However, an initial diagnosis text determined according to the report template is "ovarian cyst". During the adjustment process, the user did not adjust the initial diagnosis text by inputting information. After determining the medical report, the server performs semantic analysis on the patient information and the medical report respectively. Based on the patient information, it recognizes that the initial diagnosis text does not match the patient's gender, so the initial diagnosis text is displayed to the user in bold.
[0174] The chief complaint is the symptom that causes the patient to seek medical treatment. When the patient information includes the chief complaint, the server can determine the body part corresponding to the symptom from among the body parts included in the report template based on the symptoms corresponding to the chief complaint.
[0175] Then, the server determines in the medical report whether the diagnosis result of the human body part corresponding to the symptom indicates an abnormality. If so, it is determined that the medical report does not detect the problem. If not, the prompt information is determined based on the diagnosis result of the human body part corresponding to the symptom in the medical report.
[0176] Through this embodiment, the server further verifies the content of the determined medical report based on the patient information, so that the user can further adjust the content that may be omitted or erroneous in the medical report and determine a more accurate medical report.
[0177] The process of this embodiment is a further optimization process for the determined medical report. When this embodiment is applied, although the accuracy of the medical report can be further improved, the generation time of the medical report will be increased accordingly. Therefore, in order to improve the user experience, this embodiment can be executed by the server in response to the user's operation, so that the user can freely choose whether to adopt the method of this embodiment according to the actual medical report generation needs.
[0178] According to the above content, the type of input information in this specification may include instruction type. The execution of the instruction corresponding to the method of this embodiment may be triggered by a manual operation of a user specifying an identifier on a graphical user page, or may be triggered by voice information.
[0179] When triggered by voice information, in one or more embodiments of this specification, the server may predetermine that the keyword for executing the instruction corresponding to the method of this embodiment is "one-key quality control", then the server determines that the input information contains the word "one-key quality control" or a word with the same semantics as "one-key quality control", and the server executes the method of this embodiment to further check the medical report. Of course, "one-key quality control" is only a way to set a keyword, and any keyword can be set according to actual needs to trigger the method of this embodiment.
[0180] Figure 6This is a voice-triggered optimization instruction process diagram provided in the embodiments of this specification. Figure 6 As shown in the figure, the medical report contains the diagnosis result "No obvious abnormality was found in the heart and large blood vessels". However, after the server performs semantic analysis on the patient's complaint, it determines that the body parts corresponding to the symptoms in the complaint are the heart and large blood vessels, but the diagnosis result of the body parts corresponding to the symptoms in the medical report does not indicate abnormality. Therefore, the server displays the diagnosis result "No obvious abnormality was found in the heart and large blood vessels" in the medical report to highlight the display effect.
[0181] According to the above-mentioned embodiments, Figure 7 A method for generating a medical report is shown. Figure 7 This is a flow chart of generating a medical report provided in the embodiments of this specification. Figure 7 As shown, for the parts of concern corresponding to the regions of interest, the server determines the initial diagnosis text based on the regions of interest. For the non-parts of concern that do not have corresponding regions of interest, the server determines the initial diagnosis text based on the preset diagnosis text of the report template. The initial diagnosis report is determined based on the initial diagnosis texts of each part of the human body.
[0182] Then, the server obtains the user's input information, which can be voice information or text information, and identifies the type of the input information. If the input information is determined to be a command type, the corresponding operation is performed; if the voice information is determined to be a non-report type, the input information is discarded and not processed.
[0183] If the input information is determined to be a report type, the input information is used to adjust the initial diagnosis report. The user's subsequent input is predicted based on the semantics of the input information to determine the prediction information. The initial diagnosis report is adjusted based on the input information and the prediction information to determine the medical report.
[0184] After determining the medical report, the server may further optimize the determined medical report according to the patient information to obtain an optimized medical report.
[0185] like Figure 7 As shown, the user inputs input information containing the semantics of "generate report" so that the server performs an operation of determining to generate a medical report. The user may also input input information containing the semantics of "one-key quality control" so that the server performs an operation of optimizing the medical report.
[0186] The above is the medical report generation method provided in this specification. Based on the same idea, this specification also provides a corresponding medical report generation device, such as Figure 8 shown.
[0187] Figure 8 A schematic diagram of a medical report generating device provided in this specification specifically includes:
[0188] An initial diagnosis report determination module 200 is used to determine and display an initial diagnosis report;
[0189] An input module 202, configured to obtain input information input by a user based on the initial diagnosis report, predict a subsequent input of the user based on the input information, and determine prediction information;
[0190] The medical report determination module 204 is used to adjust the initial diagnosis report according to the input information and the prediction information to determine a medical report.
[0191] Optionally, the initial diagnosis report determination module 200 is specifically used to determine a report template containing preset diagnosis texts for various human body parts, determine an initial diagnosis report based on the report template, and display the initial diagnosis report.
[0192] Optionally, the initial diagnosis report determination module 200 is specifically used to acquire medical images, determine the region of interest in the medical images, determine the human body part corresponding to the region of interest among the human body parts, as the part of interest, determine the initial diagnosis text of the part of interest based on the region of interest, determine the initial diagnosis report based on the initial diagnosis text of the part of interest, and display the initial diagnosis report.
[0193] Optionally, the input module 202 is specifically used to predict the user's subsequent input based on the input information, determine each predicted candidate information, display the each candidate information to the user, and in response to the user's selection operation on the each candidate information, use the candidate information selected by the user as the predicted information.
[0194] Optionally, the input module 202 is specifically used to determine the matching probability between each candidate information and the input information, sort the candidate information corresponding to each matching probability according to the size of each matching probability, determine the sorted candidate information, and display the sorted candidate information to the user.
[0195] Optionally, the medical report determination module 204 is specifically used to determine adjustment information based on the input information and the prediction information, determine a text segment corresponding to the adjustment information in the initial diagnosis report, and adjust the text segment based on the adjustment information.
[0196] Optionally, the input information is voice information or text information, and the medical report determination module 204 is specifically used to determine the text information corresponding to the voice information when the input information is voice information, integrate the text information with the prediction information, and determine the adjustment information; when the input information is text information, integrate the text information with the prediction information and determine the adjustment information.
[0197] Optionally, the medical report determination module 204 is specifically used to determine adjustment information based on the input information and the predicted information, determine the human body part corresponding to the adjustment information among the various human body parts as the target part, and adjust the initial diagnosis text of the target part in the initial diagnosis report based on the adjustment information to determine the medical report.
[0198] Optionally, there are multiple target parts, and the medical report determination module 204 is specifically used to determine the adjustment content of each target part in the adjustment information, and for each target part, adjust the initial diagnosis text of the target part according to the adjustment content of the target part.
[0199] Optionally, the medical report determination module 204 is specifically used to delete the initial diagnosis text of the target part when the adjustment content of the target part contains keywords representing the semantics of deletion, and to integrate the adjustment content of the target part with the initial diagnosis text of the target part when the adjustment content of the target part does not contain keywords representing the semantics of deletion.
[0200] Optionally, the medical report determination module 204 is specifically used to perform semantic analysis on the adjustment content of the target site and the initial diagnosis text of the target site. When it is determined that there is a contradiction between the semantics of the adjustment content of the target site and the initial diagnosis text of the target site, the initial diagnosis text of the target site is replaced by the adjustment content of the target site. When it is determined that there is duplication in the semantics of the adjustment content of the target site and the initial diagnosis text of the target site, the adjustment content of the target site is merged with the initial diagnosis text of the target site.
[0201] Optionally, the initial diagnosis report is used to describe the imaging information of the medical image, and the medical report determination module 204 is specifically used to adjust the initial diagnosis report according to the input information and the prediction information, input the adjusted initial diagnosis report into a third language model, generate an examination conclusion of the medical image, and determine the medical report based on the adjusted initial diagnosis report and the examination conclusion of the medical image.
[0202] Optionally, the medical report determination module 204 is specifically used to obtain patient information, the patient information including at least one of basic information, chief complaint, relevant examination results, and medical history, and determine whether there is any mismatch between the medical report and the patient information. If so, a prompt message is determined based on the corresponding statement in the medical report of the mismatching content. The prompt message is used to prompt the user to re-check the mismatching content. If not, the medical report is displayed to the user.
[0203] This specification also provides a computer-readable storage medium, which stores a computer program, which can be used to execute the above Figure 1 A method for generating a medical report is provided.
[0204] This manual also provides Fig. 9 The schematic structure diagram of the electronic device shown in FIG. Fig. 9 As mentioned above, at the hardware level, the electronic device includes a processor, an internal bus, a network interface, a memory, and a non-volatile memory, and may also include other hardware required for the business. The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it to achieve the above Figure 1 Of course, in addition to the software implementation, this specification does not exclude other implementations, such as logic devices or a combination of software and hardware, etc., that is, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or logic devices.
[0205] For the improvement of a technology, it can be clearly distinguished whether it is a hardware improvement (for example, improvement of the circuit structure of diodes, transistors, switches, etc.) or a software improvement (improvement of the method flow). However, with the development of technology, many improvements of the method flow today can be regarded as direct improvements of the hardware circuit structure. Designers almost always obtain the corresponding hardware circuit structure by programming the improved method flow into the hardware circuit. Therefore, it cannot be said that the improvement of a method flow cannot be implemented with a hardware entity module. For example, a programmable logic device (PLD) (such as a field programmable gate array (FPGA)) is such an integrated circuit whose logical function is determined by the user's programming of the device. Designers can "integrate" a digital system on a PLD by programming themselves, without having to ask chip manufacturers to design and produce dedicated integrated circuit chips. Moreover, nowadays, instead of manually making integrated circuit chips, this kind of programming is mostly implemented by "logic compiler" software, which is similar to the software compiler used when developing and writing programs, and the original code before compilation must also be written in a specific programming language, which is called hardware description language (HDL). There is not only one HDL, but many kinds, such as ABEL (Advanced Boolean Expression Language), AHDL (Altera Hardware Description Language), Confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (Java Hardware Description Language), Lava, Lola, MyHDL, PALASM, RHDL (Ruby Hardware Description Language), etc. The most commonly used ones are VHDL (Very-High-Speed Integrated Circuit Hardware Description Language) and Verilog. Those skilled in the art should also know that it is only necessary to program the method flow slightly in the above-mentioned hardware description languages and program it into the integrated circuit, and then it is easy to obtain the hardware circuit that implements the logic method flow.
[0206] The controller can be implemented in any appropriate manner, for example, the controller can take the form of a microprocessor or processor and a computer-readable medium storing a computer-readable program code (such as software or firmware) that can be executed by the (micro)processor, a logic gate, a switch, an application-specific integrated circuit (ASIC), a programmable logic controller, and an embedded microcontroller. Examples of controllers include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicone Labs C8051F320. The memory controller can also be implemented as part of the control logic of the memory. Those skilled in the art also know that in addition to implementing the controller in a purely computer-readable program code manner, the controller can be implemented in the form of a logic gate, a switch, an application-specific integrated circuit, a programmable logic controller, and an embedded microcontroller by logically programming the method steps. Therefore, this controller can be considered as a hardware component, and the devices included therein for implementing various functions can also be regarded as structures within the hardware component. Or even, the devices for implementing various functions can be regarded as both software modules for implementing the method and structures within the hardware component.
[0207] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.
[0208] For the convenience of description, the above device is described in various units according to their functions. Of course, when implementing this specification, the functions of each unit can be implemented in the same or multiple software and / or hardware.
[0209] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.
[0210] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0211] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0212] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0213] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.
[0214] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.
[0215] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic 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 (transitory media), such as modulated data signals and carrier waves.
[0216] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.
[0217] Those skilled in the art will appreciate that the embodiments of this specification may be provided as methods, systems or computer program products. Therefore, this specification may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, this specification may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0218] This specification may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data organs, etc. that perform specific tasks or implement specific abstract data types. This specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media including storage devices.
[0219] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.
[0220] The above description is only an embodiment of this specification and is not intended to limit this specification. For those skilled in the art, this specification may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this specification should be included in the scope of the claims of this application.
Claims
1. A method for generating a medical report, characterized in that: include: Determine initial diagnostic report and present; Acquire input information input by a user based on the initial diagnosis report, predict a subsequent input of the user according to the input information, and determine prediction information; The initial diagnosis report is adjusted according to the input information and the prediction information to determine a medical report.
2. The method according to claim 1, characterized in that Determine and present the initial diagnostic report, including: Determine report templates containing pre-set diagnostic text for each body part; According to the report template, an initial diagnosis report is determined, and the initial diagnosis report is displayed.
3. The method according to claim 2, characterized in that According to the report template, determine the initial diagnostic report, including: Acquire a medical image and determine a region of interest in the medical image; Among the human body parts, determining the human body part corresponding to the region of interest as the focus part; Determining an initial diagnosis text of the part of interest according to the region of interest; An initial diagnosis report is determined based on the initial diagnosis text of the concerned part, and the initial diagnosis report is displayed.
4. The method according to claim 1, characterized in that Predicting the user's subsequent input according to the input information and determining prediction information specifically includes: Predicting the user's subsequent input according to the input information to determine each predicted candidate information; Displaying each candidate information to the user; In response to the user's selection operation on each piece of candidate information, the candidate information selected by the user is used as prediction information.
5. The method according to claim 4, characterized in that Displaying the candidate information to the user specifically includes: respectively determining a matching probability between each candidate information and the input information; According to the magnitude of each matching probability, the candidate information corresponding to each matching probability is sorted to determine the sorted candidate information; The sorted candidate information is displayed to the user.
6. The method according to claim 1, characterized in that Adjusting the initial diagnosis report according to the input information and the prediction information to determine a medical report specifically includes: Determining adjustment information according to the input information and the prediction information; In the initial diagnosis report, determining a text segment corresponding to the adjustment information; The text segment is adjusted according to the adjustment information.
7. The method according to claim 6, characterized in that The input information is voice information or text information; Determining adjustment information according to the input information and the prediction information specifically includes: When the input information is voice information, determining text information corresponding to the voice information, integrating the text information with the prediction information, and determining adjustment information; When the input information is text information, the text information is integrated with the prediction information to determine the adjustment information.
8. The method according to claim 2 or 3, characterized in that: Adjusting the initial diagnosis report according to the input information and the prediction information to determine a medical report specifically includes: Determining adjustment information according to the input information and the prediction information; Determine, among the human body parts, a human body part corresponding to the adjustment information as a target part; The initial diagnosis text of the target part in the initial diagnosis report is adjusted according to the adjustment information to determine the medical report.
9. The method according to claim 8, characterized in that The number of the target sites is multiple; According to the adjustment information, the initial diagnosis text of the target part in the initial diagnosis report is adjusted, specifically including: In the adjustment information, the adjustment content of each target part is determined respectively; For each target part, the initial diagnosis text of the target part is adjusted according to the adjustment content of the target part.
10. The method according to claim 9, characterized in that According to the adjustment content of the target part, the initial diagnosis text of the target part is adjusted, including: When the adjustment content of the target part contains keywords representing deletion semantics, the initial diagnosis text of the target part is deleted; When the adjustment content of the target part does not contain the keyword representing the deletion semantics, the adjustment content of the target part is integrated with the initial diagnosis text of the target part.
11. The method according to claim 10, characterized in that Integrate the adjustment content of the target part with the initial diagnosis text of the target part, including: Performing semantic analysis on the adjustment content of the target part and the initial diagnosis text of the target part; When it is determined that the adjusted content of the target part is inconsistent with the semantics of the initial diagnosis text of the target part, the initial diagnosis text of the target part is replaced with the adjusted content of the target part; When it is determined that the semantics of the adjustment content of the target part and the initial diagnosis text of the target part are repeated, the adjustment content of the target part and the initial diagnosis text of the target part are merged.
12. The method according to claim 1, characterized in that The initial diagnosis report is used to describe the imaging information of the medical image; Adjusting the initial diagnosis report according to the input information and the prediction information to determine a medical report specifically includes: Adjusting the initial diagnosis report according to the input information and the prediction information; Inputting the adjusted initial diagnosis report into a third language model to generate an examination conclusion of the medical image; A medical report is determined based on the adjusted initial diagnosis report and the examination conclusion of the medical image.
13. The method according to claim 1, characterized in that After determining the medical report, the method further includes: Acquiring patient information, wherein the patient information includes at least one of basic information, chief complaint, relevant examination results, and medical history; Determining whether there is any mismatch between the medical report and the patient information; If yes, determining prompt information according to the sentence corresponding to the mismatching content in the medical report, wherein the prompt information is used to prompt the user to check the mismatching content again; If not, the medical report is displayed to the user.
14. A medical report generating device, characterized in that: include: An initial diagnosis report determination module determines and displays the initial diagnosis report; An input module, which obtains input information input by a user based on the initial diagnosis report, predicts a subsequent input of the user according to the input information, and determines prediction information; The medical report determination module adjusts the initial diagnosis report according to the input information and the prediction information to determine the medical report.
15. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 13 is implemented.
16. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the method according to any one of claims 1 to 13 is implemented.
Citation Information
Cited By
Medical report generation method and apparatus, storage medium, and electronic device
WO2026046388A1
