Method and system for constructing video and audio report based on diagnostic logic of structured report

Abstract

The invention provides a method for constructing a video and audio report based on diagnostic logic of a structured report, and the method comprises the steps: creating a corresponding video template based on a preset single disease examination item condition; setting video template content for each video template; configuring video and audio materials used by each scene segment; and after the image structured report is checked, selecting a video template corresponding to the image structured report, obtaining related data of the content of the video template, performing data filling and rendering on each scene segment in the video template based on the video and audio materials, and synthesizing a complete and coherent video and audio image report. The invention further discloses a system for constructing the video and audio report based on the diagnostic logic of the structured report. According to the method, the video and audio stream for interpreting the report is fully automatically constructed for the patient / clinician on the basis of the internal normativity and diagnostic logic of the iconography structured report, so that an imaging department doctor does not need to communicate with the patient / clinician face to face, and the face-to-face communication effect is achieved.

Description

Technical Field

[0001] The present invention relates to the field of medical information, and more specifically, to a method and system for constructing an audiovisual report based on the diagnostic logic of a structured report. Background Art

[0002] The diagnostic results in the radiology department are mainly delivered in the form of paper diagnostic reports and pictures carried on physical films. Currently, the delivery is gradually transitioning to the electronic release mode of cloud films, but the essential content remains unchanged. In terms of form, the report is the report and the image is the image, and the two are separated and have no connection. In some scenarios, if the radiology department uses a structured report and integrates it with image AI or post-processing, a report form with mixed text and images can be generated, and this report can also be released in electronic form. The structured report can prospectively control the quality of the report, and the report with mixed text and images is convenient to read, and the text and images are relevant, which is of great help for understanding the condition and can greatly improve the satisfaction of clinical departments and patients, thereby improving the quality of medical care.

[0003] There are also some that use voice controls to automatically read the report; at the same time, when browsing the report on the PC side, when the mouse hovers over certain keywords, you can click to view the relevant glossary. These formal improvements further enhance the satisfaction of patients and clinicians.

[0004] However, no matter which of the above forms, the deep diagnostic basis and logic of the report cannot be interpreted, and there is still much room for improvement. To conduct a high-level interpretation of the report, the materials on paper are inefficient and difficult to understand, and video communication is necessary. This is why the face-to-face communication between radiologists and clinicians, and between radiologists and patients is much more effective than reading the report itself. Summary of the Invention

[0005] In view of this, the main object of the present invention is to provide a method and system for constructing an audiovisual report based on the diagnostic logic of a structured report. Based on the internal standardization and diagnostic logic of the radiological structured report, the method automatically constructs an audiovisual stream for interpreting the report for patients / clinicians. Without the need for radiologists to communicate with patients / clinicians face-to-face, it can largely achieve the effect of face-to-face communication, and essentially improve the satisfaction of clinicians and patients with the work of the radiology department without affecting the work efficiency of the radiology department.

[0006] To achieve the above object, the technical solution of the present invention is realized as follows:

[0007] On the one hand, the present invention provides a method for constructing an audio-visual report based on a diagnostic logic of a structured report, including: creating a corresponding video template based on a preset single-disease examination item condition and defining the name of the video template; setting the video template content for each of the video templates, where the video template content includes various combinations of the following scene segments: patient information, diagnostic Guideline, main lesion, secondary lesion, diagnostic conclusion, logical reasoning diagram of the diagnostic conclusion, recommended subsequent examination and follow-up plan; configuring the audio-visual materials used for each of the scene segments, including: visual content, background music, reading content, and whether to generate subtitles; after the imaging structured report is reviewed, selecting the video template corresponding to the imaging structured report, obtaining the relevant data of the video template content from the imaging structured report, and filling and rendering each scene segment in the video template based on the audio-visual materials to synthesize a complete and coherent audio-visual imaging report.

[0008] Preferably, the imaging structured report has a built-in diagnostic logic, and the diagnostic conclusion and the logical reasoning diagram of the diagnostic conclusion are generated through the diagnostic logic and imaging manifestations.

[0009] Preferably, the video stream displayed by the scene segment includes the following content: the patient information, whose video stream displays the clinical purpose of the imaging examination, the scanning time of the imaging examination, and the scanning method; the diagnostic Guideline, the diagnostic logic involved in the imaging structured report, whose video stream displays the criteria of the diagnostic Guideline; the main lesion, whose video stream displays the picture of the main lesion, the lesion information marked in the picture, and the measurement value of the lesion; the secondary lesion, whose video stream displays the picture of the secondary lesion; the diagnostic conclusion, whose video stream displays the text of the diagnostic conclusion; the logical reasoning diagram of the diagnostic conclusion, whose video stream displays the logical reasoning diagram of the diagnostic conclusion, and the logical reasoning diagram of the diagnostic conclusion shows how to infer the current diagnostic conclusion according to the diagnostic Guideline and is automatically drawn and generated based on the built-in diagnostic logic of the imaging structured report and the imaging manifestations; the recommended subsequent examination and follow-up plan are generated based on preset rules or the built-in logic of the imaging structured report.

[0010] Preferably, based on the audiovisual material, data filling and rendering are performed on each scene segment in the video template to synthesize a complete and coherent audiovisual image report. The steps include: designing a video stream template to export a JSON format processed by a computer script; using the computer script to replace the placeholder content with the actual data required for the video template content and rendering to generate visual content; based on the text-to-speech technology, converting the recited content into speech and a matching subtitle stream with the subtitle synchronized with the speech using time codes; after synchronizing the video stream and audio stream of each scene segment, synthesizing a complete and coherent audiovisual image report.

[0011] Preferably, for the relevant information of the main lesion and the secondary lesion, the acquisition method includes: integrating the imaging structured report with the imaging-assisted diagnosis AI module or the post-processing module to obtain the key images and measurement values output by the imaging-assisted diagnosis AI module or the post-processing module, filling them into specific positions in the imaging structured report, and assigning corresponding ontology tags; the imaging structured report fills in the imaging structured report through the DICOM images transmitted by the imaging device and the attached GSPS status small file, and assigns corresponding ontology tags; based on a dedicated interface, using a batch processing tool, the patient images and measurement values are sent to the imaging structured report in real time, and at the same time, the ID of the current patient image and the GSPS status are obtained.

[0012] On the other hand, the present invention also provides a system for constructing an audio-visual report based on a diagnostic logic of a structured report. The system includes: a video template creation module, a template content setting module, an audio-visual configuration module, and an audio-visual synthesis module. Among them, the video template creation module is connected to the template content setting module, and is used to create a corresponding video template based on preset single-disease examination item conditions, and define the name of the video template; the template content setting module is respectively connected to the video template creation module, the audio-visual configuration module, and the audio-visual synthesis module, and is used to set video template content for each video template. The video template content includes various combinations in the following scene segments: patient information, diagnostic Guideline, main lesions, secondary lesions, diagnostic conclusion, logical reasoning diagram of the diagnostic conclusion, recommended follow-up examinations and follow-up plans; the audio-visual configuration module is respectively connected to the template content setting module and the audio-visual synthesis module, and is used to configure the audio-visual materials used for each scene segment, including: visual content, background music, reading content, and whether to generate subtitles; the audio-visual synthesis module is respectively connected to the template content setting module and the audio-visual configuration module, and is used to, after the imaging structured report is reviewed, select the video template corresponding to the imaging structured report, obtain the relevant data of the video template content from the imaging structured report, and based on the audio-visual materials, perform data filling and rendering on each scene segment in the video template, and synthesize a complete and coherent audio-visual imaging report.

[0013] Preferably, the imaging structured report has a built-in diagnostic logic, and the diagnostic conclusion and the logical reasoning diagram of the diagnostic conclusion are generated through the diagnostic logic and imaging manifestations.

[0014] Preferably, the video stream displayed by the scene segment includes the following content: the patient information, and its video stream displays the clinical purpose of the imaging examination, the scanning time of the imaging examination, and the scanning method; the diagnostic Guideline, the diagnostic logic involved in the imaging structured report, and its video stream displays the criteria of the diagnostic Guideline; the main lesions, and its video stream displays pictures of the main lesions, the lesion information marked in the pictures, and the measurement values of the lesions; the secondary lesions, and its video stream displays pictures of the secondary lesions; the diagnostic conclusion, and its video stream displays the text of the diagnostic conclusion; the logical reasoning diagram of the diagnostic conclusion, and its video stream displays the logical reasoning diagram of the diagnostic conclusion. The logical reasoning diagram of the diagnostic conclusion shows how to infer the current diagnostic conclusion according to the diagnostic Guideline, and is automatically drawn and generated based on the built-in diagnostic logic and imaging manifestations of the imaging structured report; the recommended follow-up examinations and follow-up plans are generated based on preset rules or the built-in logic of the imaging structured report.

[0015] Preferably, the video-audio synthesis module includes: a format conversion unit, a data processing unit, a synchronization unit, and an output unit. Among them, the format conversion unit is used to export a JSON format processed by a computer script through video stream template design; the data processing unit is used to replace the placeholder content with the actual data required by the video template content by a computer script and render visual content; the synchronization unit is used to convert the recited content into speech and a matching subtitle stream based on the text-to-speech technology, and synchronize the subtitle with the speech using a time code; the output unit is used to synchronize the video stream and the audio stream of each scene segment and then synthesize a complete and coherent video-audio image report.

[0016] Preferably, the method for obtaining the relevant information of the main lesion and the secondary lesion includes: by integrating the imaging structured report with the imaging-assisted diagnosis AI module or the post-processing module, obtaining the key images and measurement values output by the imaging-assisted diagnosis AI module or the post-processing module, filling them into specific positions of the imaging structured report, and assigning corresponding ontology labels; the imaging structured report fills in the imaging structured report through the DICOM images transmitted by the imaging device and the attached GSPS status small file, and assigns corresponding ontology labels; based on a dedicated interface, using a batch processing tool, the patient images and measurement values are sent to the imaging structured report in real time, and at the same time, the ID of the current patient image and the GSPS status are obtained.

[0017] The technical effects of the present invention:

[0018] 1. The method of the present invention is based on the inherent standardization and diagnostic logic of the imaging structured report, and automatically constructs a video-audio stream for interpreting the report for patients / clinicians. Without the need for radiologists to communicate face-to-face with patients / clinicians, it can largely achieve the effect of face-to-face communication. Without affecting the work efficiency of the radiology department, it essentially improves the satisfaction of clinicians and patients with the work of the radiology department;

[0019] 2. The most core part of the video stream demonstrated by the method of the present invention is the diagnostic Guideline and the diagnostic logic reasoning diagram, which can display the medical practice guidelines and evaluation criteria corresponding to the imaging manifestations, and the diagnostic logic reasoning diagram on how to draw a diagnostic conclusion based on the guidelines and imaging manifestations. In the scenario where radiologists do not need to explain face-to-face, clinicians / patients can deeply understand the condition through the video stream;

[0020] 3. The method of the present invention promotes the popularization and deepening of high-quality structured reports by improving the output quality of radiology department reports, and is expected to become a charging item in some medical institutions, improving the performance of department doctors and the doctor's initiative work ability. Description of the Drawings

[0021] The accompanying drawings described herein are used to provide a further understanding of the present invention and form a part of this application. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0022] Figure 1 A flowchart of a method for constructing an audiovisual report based on a structured report's diagnostic logic according to Embodiment 1 of the present invention is shown;

[0023] Figure 2 A logical reasoning diagram of a diagnostic conclusion in the method for constructing an audiovisual report based on a structured report's diagnostic logic according to Embodiment 1 of the present invention is shown;

[0024] Figure 3 A schematic structural diagram of a system for constructing an audiovisual report based on a structured report's diagnostic logic according to Embodiment 2 of the present invention is shown;

[0025] Figure 4 A schematic structural diagram of a system for constructing an audiovisual report based on a structured report's diagnostic logic according to Embodiment 3 of the present invention is shown;

[0026] Figure 5 A logical reasoning diagram of a diagnostic conclusion in the system for constructing an audiovisual report based on a structured report's diagnostic logic according to Embodiment 3 of the present invention is shown. Detailed Embodiments

[0027] The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0028] Embodiment 1

[0029] Figure 1 A flowchart of a method for constructing an audiovisual report based on a structured report's diagnostic logic according to Embodiment 1 of the present invention is shown; As Figure 1 shown, the method includes the following steps:

[0030] Overall description:

[0031] This method relies on structured imaging performance data. It is necessary to use an imaging structured report to be able to use this method to construct audiovisuals. Because using a local large model to extract imaging performance and diagnostic conclusions is still very inaccurate, the hallucinations generated by the large model can lead to serious medical accidents and medical disputes, and its quality is completely incomparable with that of a structured report based on strict logic. This is also the reason why the present invention is based on the application of imaging structured reports.

[0032] This method relies on the built-in diagnostic logic of the imaging structured report. Therefore, the structured report must also have internal diagnostic logic, such as ACR / AJCC or other clear specialty diagnostic logics. If it is only structured in terms of imaging findings and the diagnostic conclusion does not depend on the built-in logic but a simple structured template handwritten by the doctor, this method cannot be used to generate audio-visuals. The characteristic of the imaging structured report is that through the imaging findings filled in by the reporting doctor based on the patient's images, combined with built-in guidelines such as ACR / AJCC, the diagnostic conclusion is automatically output, and each chapter of the imaging structured report has its relevance.

[0033] For the reviewer doctor to modify the report conclusion, it must be through the special technology of the company that corresponds the text to the structure, locate the structured report part of the previous imaging findings description for modification, and regenerate the diagnostic conclusion automatically. If the diagnostic conclusion uses the manual modification mode, the complete corresponding relationship between the imaging findings and the diagnostic conclusion cannot be guaranteed. In many cases, the reviewer doctor does not want to carefully modify the imaging findings and only modifies the diagnostic conclusion. This scenario is very common. This non-correspondence before and after, in essence, is a medical error. This non-correspondence problem is not so prominent in the text report because the text report is difficult to understand; but in the audio-visual of the interpretation report, because clear explanations need to be made, this kind of error is more likely to be exposed, leading to medical disputes.

[0034] Step S101, based on the preset single-disease examination item conditions, create a corresponding video template and define the name of the video template;

[0035] The following table is given as an example:

[0036] Match the video content template used according to the rule table:

[0037]

[0038] Find the video template with ID 1002 to be used according to the rule conditions:

[0039] Number Name Template Content 1002 Lung-RADS Explanation Template for Pulmonary Nodules … 1003 Explanation Template for Breast T-Stage …

[0040] Step S102, set the video template content for each video template. The video template content includes various combinations of the following scenario segments: patient information, diagnostic guidelines, main lesions, secondary lesions, diagnostic conclusion, the logical reasoning diagram of the diagnostic conclusion, recommended follow-up examinations and follow-up plans;

[0041] Among them, the imaging structured report has built-in diagnostic logic, and the diagnostic conclusion and the logical reasoning diagram of the diagnostic conclusion are generated through the diagnostic logic and imaging findings;

[0042] Among them, the video stream shown in the scene segment includes the following content: the patient information, whose video stream shows the clinical purpose of the imaging examination, the scanning time of the imaging examination, and the scanning method; the diagnostic Guideline, the diagnostic logic involved in the image structured report, whose video stream shows the criteria of the diagnostic Guideline; the main lesion, whose video stream shows the picture of the main lesion, the lesion information marked in the picture, and the measurement value of the lesion; the secondary lesion, whose video stream shows the picture of the secondary lesion; the diagnostic conclusion, whose video stream shows the text of the diagnostic conclusion; the logical reasoning diagram of the diagnostic conclusion, whose video stream shows the logical reasoning diagram of the diagnostic conclusion. This logical reasoning diagram of the diagnostic conclusion shows how to infer the current diagnostic conclusion according to the diagnostic Guideline and is automatically drawn and generated based on the built-in diagnostic logic of the image structured report and the image manifestation; the recommended subsequent examination and follow-up plan are generated based on preset rules or the built-in logic of the image structured report.

[0043] Among them, the methods for obtaining the relevant information of the main lesion and the secondary lesion include: through the integration of the image structured report with the image-assisted diagnosis AI module or the post-processing module, obtaining the key images and measurement values output by the image-assisted diagnosis AI module or the post-processing module, filling them into specific positions of the image structured report, and assigning corresponding ontology tags; the image structured report fills in the image structured report through the DICOM images transmitted by the imaging device and the attached GSPS status small file, and assigns corresponding ontology tags; based on a dedicated interface, using a batch processing tool, the patient images and measurement values are sent to the image structured report in real time, and at the same time, the ID of the current patient image and the GSPS status are obtained.

[0044] The video is seamlessly connected by several scene segments, and each scene segment is automatically rendered by a computer or uses existing video materials. Which scene segments a video interpretation consists of is defined by the content of the video template, and which content template a structured report uses to generate the video is based on predefined rules. The scene segments may use data such as patient clinical information, examination information, lesion pictures, measurement values, imaging manifestations, diagnostic conclusions, diagnostic Guidelines, diagnostic logic, recommended subsequent examination and follow-up information, etc. These data are provided by the structured report system and related peripheral modules such as the AI module, the post-processing module, and the pacs system interface, and some information depends on the built-in logic of the structured report.

[0045] A typical content template for generating a video interpretation of a structured report and the data it depends on are as follows:

[0046] The first scene segment is the introduction of the patient's examination information. The video stream shows the clinical purpose of the imaging examination, the scanning time of the imaging examination, and the scanning method (the type of scanning protocol such as equipment type, location, and whether to enhance, the number of sequences, and their names), while the audio stream reads out these contents.

[0047] Among them, the clinical application purpose is extracted from the RIS system; the scanning time is extracted from the department management platform that records the scanning time of the computer room; the scanning method (the type of scanning protocol such as equipment type, location, and whether to enhance, the number of sequences, and their names) is extracted from the DICOM file header of the PACS.

[0048] The second scene segment is the introduction of the Guideline involved in the report. This paragraph uses fixed and shared video materials. As long as this report involves this diagnostic Guideline, this video will be included to introduce the diagnostic criteria.

[0049] The third scene segment is the introduction of the main lesions. The video stream shows the pictures of the main lesions, highlights the lesion information marked in the pictures, and shows the lesion measurement values. The audio stream introduces the imaging manifestations, measurement values, and other information of the lesions.

[0050] Among them, there are three sources for obtaining the lesion pictures, marked information, and measurement values:

[0051] Method 1: Through the deep integration of the structured report with the AI module or the post - processing module, obtain the key images and measurement values output by them. These key images and measurement values are filled in specific positions of the report and corresponding ontology labels are obtained.

[0052] Method 2: The structured report fills in the relevant content of the report through the DICOM images transmitted by the imaging equipment and the attached GSPS status small file (calibrating the magnification, window width, annotation, and measurement values of the image), and assigns relevant ontology labels to these data objects. When using the GSPS status information to reference the image, the image browser needs to be called using the image ID to display the image.

[0053] Method 3: Use the company's PACS system. When using the dedicated interface developed by it and the batch processing tool for measurement, the measurement values will be sent to the structured report system in real - time and automatically, and at the same time, the ID number of the current image and the GSPS status will be obtained. Among them, the text of the imaging manifestation is generated by the structured report system and corresponds to a set of standardized structured data, which can form a correspondence with the information of the lesion pictures.

[0054] The fourth scene segment is the introduction of the display of minor lesions, generally not exceeding 4.

[0055] The fifth scene segment is to display the text of the diagnostic conclusion and read it aloud. This part of the content is generated by the structured reporting system through built-in logic and structured imaging performance data. The built-in logic of the structured report is set according to the Guideline. Therefore, the logic of generating the diagnostic conclusion from the imaging performance data is very clear.

[0056] The sixth scene segment is to display the logic reasoning diagram for generating this diagnosis. This diagram shows how to infer the current diagnostic conclusion according to the Guideline and is automatically drawn using the built-in logic of the structured report and the imaging performance data.

[0057] The seventh scene segment is to display the content of the recommended follow-up examinations and follow-up plans. This part is generated according to the rules and may also be taken from the built-in logic of the structured report.

[0058] Step S103, configure the audio-visual materials used for each of the said scene segments, including: visual content, background music, reading content, and whether to generate subtitles;

[0059] Step S104, after the imaging structured report is reviewed, select the video template corresponding to the imaging structured report, obtain the relevant data of the content of the video template from the imaging structured report, and based on the audio-visual materials, fill in the data and render each scene segment in the video template to synthesize a complete and coherent audio-visual imaging report.

[0060] Among them, based on the audio-visual materials, filling in the data and rendering each scene segment in the video template to synthesize a complete and coherent audio-visual imaging report, the steps include: through video stream template design, export in JSON format for processing by computer scripts; the computer script replaces the placeholder content with the actual data required for the content of the video template and renders to generate visual content; based on text-to-speech technology, use the reading content as the sound, convert it to speech and a matching subtitle stream, and synchronize the subtitles and speech using time codes; after synchronizing the video stream and audio stream of each scene segment, synthesize a complete and coherent audio-visual imaging report.

[0061] The video template content is a collection of scene segments on the timeline. Some of these scene segments use existing video materials, while others require computer rendering. The scene segments that need to be rendered usually need to contain at least one video stream template. This template can be designed by video editing software such as Adobe After Effects and exported as a JSON format that can be processed by scripts. The computer script replaces the placeholder content with actual data and renders to generate visual content. For the beginning, end, and transition scenes of the guideline video, existing video materials can be used. The personalized content that needs to incorporate clinical information, report content, and diagnostic logic requires computer rendering. The guideline interpretation video can be pre-recorded according to the content of the guideline file. Since the types of guidelines are limited, for the guideline used, the corresponding scene segment is added.

[0062] The order of the scene segments is predefined. Switching different templates is equivalent to switching the scene order.

[0063] The scene segments can include background music or use the commentary content as the sound. In this case, the Text-to-Speech (TTS) technology is used to convert it into speech and a matching subtitle stream. The synchronization of the subtitles and the speech is performed using timecodes. Then, tools such as FFmpeg are used to synchronize the various video streams and audio streams along the timeline and then perform rendering and integration to form a complete and coherent video.

[0064] Methods for the release, storage, and regeneration of video-audio reports:

[0065] After the reviewing doctor has reviewed the report, a video stream file for the report interpretation is produced.

[0066] When the patient uses cloud films or the clinician uses the Web browsing method, the original separated graphic and text browsing mode and this video stream mode are provided side by side for selection. That is, it does not replace the original release mode but serves as a supplement.

[0067] The video stream file can be deleted according to about one year, similar to the short-term storage of images, but no copies need to be saved. Because the video copy can be regenerated at any time. If the video has expired, at the explicit request of the patient or clinician, the video can be regenerated non-real-time and republished.

[0068] The following uses a specific example to illustrate the method of this embodiment:

[0069] After writing a report using the report template 'Lung Nodule CT', a commentary video is generated:

[0070] First, match the used video content template according to the rule table:

[0071]

[0072] Find the video template with ID 1002 according to the rule conditions:

[0073] Number Name Template Content 1002 Lung-RADS Explanation Template for Pulmonary Nodules … 1003 Explanation Template for Breast T-Stage …

[0074] Video template content:

[0075]

[0076] Imaging findings:

[0077] Lung nodule assessment:

[0078] (Main) A solid nodule in the anterior segment of the upper lobe of the right lung, with an average diameter of about 7.00 mm, conforming to Lung-RADS: category 3.

[0079] (Minor) A nodule with fat components in the medial basal segment of the lower lobe of the right lung, conforming to Lung-RADS: category 1.

[0080] Other findings on LDCT:

[0081] The chest is symmetric, and the trachea is in the middle. The size of the heart is normal, and there is no pericardial effusion. There is no pleural effusion in both sides.

[0082] Diagnostic conclusion:

[0083] LDCT lung nodule screening, conforming to Lung-RADS category 3 (v1.1 2019)

[0084] Patient information: Wang XX, 29 years old. The first LDCT examination for lung nodules was performed in our hospital on March 5, 2019. Low-dose chest CT scan was performed. No smoking history, no surgical history.

[0085] Prepare the video stream and audio stream for each chapter according to the video template content:

[0086] Chapter 1: Patient information

[0087] Video stream: Scroll and play the following text information

[0088] Patient name: Wang XX

[0089] Age: 29 years old

[0090] Examination date: March 5, 2019

[0091] Examination purpose: LDCT screening for lung nodules

[0092] Scanning method: Low-dose chest CT scan

[0093] Scanning protocol type: Non-enhanced

[0094] Number and name of sequences: 1 sequence, low-dose chest CT scan

[0095] Audio stream: Read patient information aloud and play BGM

[0096] "Patient name: Wang XX, 29 years old. The first LDCT examination for pulmonary nodules was conducted in our hospital on March 5, 2019. The examination used low-dose chest CT scan, non-enhanced, with a total of 1 sequence."

[0097] Chapter 2: Guideline Introduction

[0098] Video stream: Introduction video and assessment criteria of Lung-RADS (public materials)

[0099] Audio stream: Included with the materials

[0100] Chapter 3: Main Lesion

[0101] Video stream: Show pictures of the main lesion, marked information, and measurement values

[0102] Marked on the picture: Solid nodule in the anterior segment of the upper lobe of the right lung, with an average diameter of about 7.00 mm

[0103] Audio stream: Read the imaging findings aloud and play BGM

[0104] Solid nodule in the anterior segment of the upper lobe of the right lung, with an average diameter of about 7.00 mm, conforming to Lung-RADS: Category 3."

[0105] Chapter 4: Secondary Lesions

[0106] Video stream: Show pictures of the remaining secondary lesions

[0107] Audio stream: BGM

[0108] Chapter 5: Diagnostic Conclusion

[0109] Video stream: Scroll and play the diagnostic conclusion text

[0110] Text content: LDCT pulmonary nodule screening, conforming to Lung-RADS Category 3 (v1.1 2019)

[0111] Audio stream: Read the diagnostic conclusion aloud

[0112] "Conclusion: LDCT pulmonary nodule screening, conforming to Lung-RADS Category 3 (v1.1 2019)."

[0113] Chapter 6: Logic Reasoning Diagram of Diagnostic Conclusion

[0114] Video stream: Inference process picture (generated based on report data and Lung-RADS guidelines)

[0115] Figure 2 Shows the logical reasoning diagram of the diagnostic conclusion in the method for constructing an audiovisual report based on the structured report in the first embodiment of the present invention; as Figure 2 shown.

[0116] Chapter 7: The recommended follow-up examinations and follow-up plans are scrolled and read aloud:

[0117] It is recommended to perform the next LDCT scan after 3-6 months to observe the changes in the nodules. If the nodules increase or new suspicious features appear during the follow-up, further diagnostic examinations such as PET-CT scan or biopsy may be required.

[0118] If the nodules remain stable during short-term follow-up, it may be recommended to perform an LDCT scan once a year to monitor the long-term changes in the nodules.

[0119] The embodiment of the present invention is based on the inherent standardization and diagnostic logic of the imaging structured report, and automatically constructs an audiovisual stream for interpreting the report for patients / clinicians. Without the need for radiologists to communicate face-to-face with patients / clinicians, it can largely achieve the effect of face-to-face communication. Without affecting the work efficiency of the radiology department, it essentially improves the satisfaction of clinicians and patients with the work of the radiology department; the most core part of the video stream demonstrated by the method of the embodiment of the present invention is the diagnostic Guideline and the diagnostic logic reasoning diagram, which can display the medical practice guidelines and evaluation criteria corresponding to the imaging manifestations, as well as the diagnostic logic reasoning diagram on how to draw the diagnostic conclusion based on the guidelines and imaging manifestations. In the scenario where radiologists do not need to explain face-to-face, clinicians / patients can deeply understand the condition through the video stream; the method of the embodiment of the present invention promotes the popularization and deepening of high-quality structured reports by improving the output quality of radiology department reports, and is expected to become a charge item in some medical institutions, improving the performance of department doctors and the initiative of doctors to work.

[0120] Embodiment Two

[0121] Overall narrative:

[0122] This method relies on structured imaging manifestation data, and it is necessary to use an imaging structured report to be able to use this method to construct an audiovisual. Because using a local large model to extract imaging manifestations and diagnostic conclusions is still very inaccurate, the hallucinations generated by the large model will cause serious medical accidents and medical disputes, and its quality is completely incomparable with that of a structured report based on strict logic. This is also the reason why the present invention is based on the application of imaging structured reports.

[0123] This method relies on the built-in diagnostic logic of the imaging structured report. Therefore, the structured report must also have internal diagnostic logic, such as ACR / AJCC or other clear specialty diagnostic logics. If it is only structured in terms of imaging findings and the diagnostic conclusion does not depend on the built-in logic but a simple structured template written by a doctor, this method cannot be used to generate the audio-visual content. The characteristic of the imaging structured report is that through the imaging findings filled in by the reporting doctor based on the patient's images and combined with the built-in guidelines such as ACR / AJCC, the diagnostic conclusion is automatically output, and each chapter of the imaging structured report has its relevance.

[0124] When the reviewing doctor modifies the report conclusion, it must be through the special technology of the company that corresponds the text with the structure to locate the structured report part of the previous imaging findings description for modification and automatically generate the diagnostic conclusion again. If the diagnostic conclusion uses the manual modification mode, the complete corresponding relationship between the imaging findings and the diagnostic conclusion cannot be guaranteed. In many cases, the reviewing doctor does not want to carefully modify the imaging findings and only modifies the diagnostic conclusion. This scenario is very common. This non-correspondence before and after is essentially a medical error. This non-corresponding phenomenon is not so prominent in the text report because the text report is difficult to understand; but in the audio-visual content of the interpreted report, because clear explanations need to be made, this error is more likely to be exposed, leading to medical disputes.

[0125] Figure 3 The schematic structural diagram of the system for constructing an audio-visual report based on the diagnostic logic of the structured report according to the second embodiment of the present invention is shown; as Figure 3 shown, the system includes: a video template creation module 10, a template content setting module 20, an audio-visual configuration module 30, and an audio-visual synthesis module 40, where

[0126] The video template creation module 10 is connected to the template content setting module 20 and is used to create a corresponding video template based on the preset single-disease examination item conditions and define the name of the video template;

[0127] An example is as follows in the table:

[0128] Video content template used according to the rule table matching:

[0129]

[0130] Find the video template with ID 1002 to be used according to the rule conditions:

[0131] Number Name Template Content 1002 Lung-RADS Explanation Template for Pulmonary Nodules … 1003 Explanation Template for Breast T-Stage …

[0132] The template content setting module 20 is respectively connected to the video template creation module 10, the video-audio configuration module 30, and the video-audio synthesis module 40, and is used to set video template content for each of the video templates. The video template content includes various combinations in the following scene segments: patient information, diagnostic guidelines, main lesions, secondary lesions, diagnostic conclusions, logical reasoning diagrams of the diagnostic conclusions, recommended subsequent examination and follow-up plans;

[0133] Among them, the imaging structured report has built-in diagnostic logic, and the diagnostic conclusions and the logical reasoning diagrams of the diagnostic conclusions are generated through the diagnostic logic and imaging manifestations;

[0134] Among them, the video stream displayed in the scene segment includes the following content: the patient information, and its video stream displays the clinical purpose of the imaging examination, the scanning time of the imaging examination, and the scanning method; the diagnostic guidelines, the diagnostic logic involved in the imaging structured report, and its video stream displays the criteria of the diagnostic guidelines; the main lesions, and its video stream displays pictures of the main lesions, the lesion information marked in the pictures, and the measurement values of the lesions; the secondary lesions, and its video stream displays pictures of the secondary lesions; the diagnostic conclusions, and its video stream displays the text of the diagnostic conclusions; the logical reasoning diagrams of the diagnostic conclusions, and its video stream displays the logical reasoning diagrams of the diagnostic conclusions. The logical reasoning diagrams of the diagnostic conclusions show how to infer the current diagnostic conclusions according to the diagnostic guidelines, and are automatically drawn and generated based on the built-in diagnostic logic of the imaging structured report and the imaging manifestations; the recommended subsequent examination and follow-up plans are generated based on preset rules or the built-in logic of the imaging structured report.

[0135] Among them, the methods for obtaining the relevant information of the main lesions and secondary lesions include: through the integration of the imaging structured report with the imaging-assisted diagnosis AI module or the post-processing module, obtaining the key images and measurement values output by the imaging-assisted diagnosis AI module or the post-processing module, filling them into specific positions of the imaging structured report, and assigning corresponding ontology labels; the imaging structured report fills in the imaging structured report through the DICOM images transmitted by the imaging device and the attached GSPS status small files, and assigns corresponding ontology labels; based on a dedicated interface, using a batch processing tool, the patient images and measurement values are sent to the imaging structured report in real time, and at the same time, the ID of the current patient image and the GSPS status are obtained.

[0136] The video is seamlessly composed of several scene segments, each of which is automatically rendered by a computer or uses existing video materials. Which scene segments a video interpretation consists of is defined by the content of the video template, while which content template is used to generate the video in a structured report is based on predefined rules. The scene segments may use data such as patient clinical information, examination information, lesion images, measurement values, imaging findings, diagnostic conclusions, diagnostic guidelines, diagnostic logic, recommended follow-up examinations and follow-up information, etc. These data are provided by the structured reporting system and related peripheral modules such as AI modules, post-processing modules, pacs system interfaces, etc., and some information depends on the logic built into the structured report.

[0137] A typical content template for generating a structured report interpretation video and the data it depends on are as follows:

[0138] The first scene segment is the introduction of the patient's examination information. The video stream shows the clinical purpose of the imaging examination, the scanning time of the imaging examination, and the scanning method (the type of scanning protocol such as equipment type, location, whether enhanced, etc., the number and name of sequences), while the audio stream reads aloud these contents.

[0139] Among them, the clinical application purpose is extracted from the RIS system; the scanning time is extracted from the department management platform that records the scanning time of the computer room; the scanning method (the type of scanning protocol such as equipment type, location, whether enhanced, etc., the number and name of sequences) is extracted from the DICOM file header of the PACS.

[0140] The second scene segment is the introduction of the guideline involved in the report. This section uses fixed and shared video materials, and as long as this report involves this diagnostic guideline, this section of the video is included to introduce the diagnostic criteria.

[0141] The third scene segment is the introduction of the main lesion. The video stream shows the picture of the main lesion, focuses on showing the lesion information marked in the picture, and shows the lesion measurement value. The audio stream introduces information such as the imaging findings and measurement value of the lesion.

[0142] Among them, the acquisition of lesion pictures, annotation information, and measurement values can have three sources:

[0143] Method 1: Through the deep integration of the structured report with the AI module or the post-processing module, obtain the key images and measurement values output by them. These key images and measurement values are filled in specific positions of the report and have obtained corresponding ontology labels.

[0144] Method 2: The structured report fills in the relevant content of the report through the DICOM images transmitted by the imaging device and the attached GSPS status small file (calibrating the magnification, window width, annotations, and measurement values of the images), and assigns relevant ontology tags to these data objects. When referring to an image using the GSPS status information, the image browser needs to be accessed using the image ID to display the image.

[0145] Method 3: Using the company's PACS system and its developed dedicated interface, when using the batch measurement tool, the measurement values will be sent to the structured report system in real-time and automatically, and at the same time, the ID number of the current image and the GSPS status will be obtained. The radiological manifestation text is generated by the structured report system and corresponds to a set of standardized structured data, which can correspond to the information of the lesion pictures.

[0146] The fourth scene segment is to display the introduction of secondary lesions, generally not exceeding 4 items.

[0147] The fifth scene segment is to display the text of the diagnosis conclusion and read it aloud. This part of the content is generated by the structured report system through the built-in logic and structured radiological manifestation data, and the built-in logic of the structured report is set according to the Guideline. Therefore, the logic of generating the diagnosis conclusion from the radiological manifestation data is very clear.

[0148] The sixth scene segment is to display the logical reasoning diagram for generating this diagnosis. This diagram shows how to infer the current diagnosis conclusion according to the Guideline, and it is automatically drawn using the built-in logic of the structured report and the radiological manifestation data.

[0149] The seventh scene segment is to display the content of the recommended follow-up examinations and follow-up plans. This part is generated according to the rules and may also be taken from the built-in logic of the structured report.

[0150] The audio-visual configuration module 30 is respectively connected to the template content setting module 20 and the audio-visual synthesis module 40, and is used to configure the audio-visual materials used for each scene segment, including: visual content, background music, reading content, and whether to generate subtitles;

[0151] The audio-visual synthesis module 40 is respectively connected to the template content setting module 20 and the audio-visual configuration module 30, and is used to, after the imaging structured report is reviewed, select the video template corresponding to the imaging structured report, obtain the relevant data of the video template content from the imaging structured report, and based on the audio-visual materials, fill in and render the data for each scene segment in the video template to synthesize a complete and coherent audio-visual imaging report.

[0152] Methods for publishing, storing, and regenerating audio-visual reports:

[0153] After the reviewing doctor finishes reviewing the report, a video stream file for report interpretation is produced.

[0154] When the patient uses the cloud film or the clinician uses the Web browsing method, the original separated graphic and text browsing mode and this video stream mode are provided side by side for selection. That is, it does not replace the original release mode, but makes a supplement.

[0155] The video stream file can be deleted according to about one year, similar to the short-term storage of images, but there is no need to save a copy. Because the video copy can be regenerated at any time. If the video expires, at the explicit request of the patient or the clinician, the video can be regenerated non-real-time and republished.

[0156] The embodiment of the present invention is based on the inherent standardization and diagnostic logic of the imaging structured report, and automatically constructs an audio-visual stream for interpreting the report for patients / clinicians. Without the need for the radiologist to communicate face-to-face with the patient / clinician, it can largely achieve the effect of face-to-face communication. Without affecting the work efficiency of the radiology department, it essentially improves the satisfaction of clinicians and patients with the work of the radiology department; the most core part of the video stream demonstrated by the method of the embodiment of the present invention is the diagnostic Guideline and the diagnostic logic reasoning diagram, which can display the medical practice guidelines and evaluation criteria corresponding to the imaging manifestations, and the diagnostic logic reasoning diagram on how to draw a diagnostic conclusion according to the guidelines and imaging manifestations. In the scenario without the radiologist explaining face-to-face, it can enable clinicians / patients to deeply understand the condition through the video stream.

[0157] Embodiment Three

[0158] Figure 4 shows a schematic structural diagram of a system for constructing an audio-visual report based on the diagnostic logic of a structured report according to Embodiment Three of the present invention; as Figure 4 shown, the audio-visual synthesis module 40 includes: a format conversion unit 402, a data processing unit 404, a synchronization unit 406, and an output unit 408, wherein,

[0159] The format conversion unit 402 is used to export a JSON format processed by a computer script through video stream template design;

[0160] The data processing unit 404 is used to replace the placeholder content with the actual data required by the video template content by a computer script and render to generate visual content;

[0161] The synchronization unit 406 is used to convert the reading content into speech and a matching subtitle stream based on text-to-speech technology, and synchronize the subtitle with the speech using a time code;

[0162] The output unit 408 is configured to synchronize the video stream and the audio stream of each scene segment and then synthesize a complete and coherent audio-visual image report.

[0163] Among them, based on the audio-visual materials, data filling and rendering are performed on each scene segment in the video template to synthesize a complete and coherent audio-visual image report. The steps include: designing a video stream template to export a JSON format processed by a computer script; using the computer script to replace the placeholder content with the actual data required for the video template content and rendering to generate visual content; based on the text-to-speech technology, converting the recited content into speech and a matching subtitle stream, and synchronizing the subtitle with the speech using time codes; synchronizing the video stream and the audio stream of each scene segment and then synthesizing a complete and coherent audio-visual image report.

[0164] The video template content is a collection of scene segments on the timeline. Some of the scene segments use existing video materials, and some require computer rendering. The scene segments that need to be rendered usually need to include at least one video stream template, which can be designed by video editing software such as Adobe After Effects and exported as a JSON format that can be processed by a script. The computer script replaces the placeholder content with actual data and renders to generate visual content. The beginning, end, and transition scenes of the guideline video can all use existing video materials. The personalized content that needs to bring in clinical information, report content, and diagnostic logic needs to be rendered by a computer. The guideline interpretation video can be pre-recorded according to the content of the guideline file. Since the types of guidelines are limited, the corresponding scene segment is added when using a certain guideline.

[0165] The order of the scene segments is predefined. Switching different templates is equivalent to switching the scene order.

[0166] The scene segment can include background music or use the commentary content as the sound. In this case, the text-to-speech (TTS) technology is used to convert it into speech and a matching subtitle stream. The synchronization of the subtitle and the speech is performed using time codes. Then, tools such as FFmpeg are used to synchronize the video streams and the audio streams through the timeline, and then render and integrate them to form a complete and coherent video.

[0167] The following uses a specific example to illustrate the method of this embodiment:

[0168] After writing a report using the report template 'Lung Nodule CT', an explanatory video is generated:

[0169] First, match the video content template used according to the rule table:

[0170]

[0171] Find the video template with ID 1002 to be used according to the rule conditions:

[0172] Number Name Template Content 1002 Lung-RADS Explanation Template for Pulmonary Nodules … 1003 Explanation Template for Breast T-Stage …

[0173] Video template content:

[0174]

[0175]

[0176] Imaging findings:

[0177] Lung nodule assessment:

[0178] (Main) A solid nodule in the anterior segment of the upper lobe of the right lung, with an average diameter of about 7.00 mm, conforming to Lung-RADS: category 3.

[0179] (Minor) A nodule with fat component in the medial basal segment of the lower lobe of the right lung, conforming to Lung-RADS: category 1.

[0180] Other findings on LDCT:

[0181] The chest is symmetric, and the trachea is in the middle. The size of the heart is normal, and there is no pericardial effusion. There is no pleural effusion on both sides.

[0182] Diagnostic conclusion:

[0183] LDCT lung nodule screening, conforming to Lung-RADS category 3 (v1.1 2019)

[0184] Patient information: Wang XX, 29 years old. The first LDCT examination for lung nodules was performed in our hospital on March 5, 2019, and a low-dose chest CT scan was conducted. There is no smoking history and no surgical history.

[0185] Prepare the video stream and audio stream for each chapter according to the video template content:

[0186] Chapter 1: Patient information

[0187] Video stream: Scroll and play the following text information

[0188] Patient name: Wang XX

[0189] Age: 29 years old

[0190] Examination date: March 5, 2019

[0191] Examination purpose: LDCT screening for lung nodules

[0192] Scanning method: Low-dose chest CT scan

[0193] Scanning protocol type: Non-contrast

[0194] Number and name of sequences: 1 sequence, low-dose chest CT scan

[0195] Audio stream: Read patient information aloud and play BGM

[0196] "Patient name: Wang XX, 29 years old. The first LDCT examination of pulmonary nodules was performed in our hospital on March 5, 2019. The examination used low-dose chest CT scan, non-contrast, with a total of 1 sequence."

[0197] Chapter 2: Guideline Introduction

[0198] Video stream: Introduction video and assessment criteria of Lung-RADS (public materials)

[0199] Audio stream: Included with the materials

[0200] Chapter 3: Main lesion

[0201] Video stream: Show pictures of the main lesion, marked information, and measurement values

[0202] Marked on the picture: Solid nodule in the anterior segment of the upper lobe of the right lung, with an average diameter of about 7.00 mm

[0203] Audio stream: Read the imaging findings aloud and play BGM

[0204] Solid nodule in the anterior segment of the upper lobe of the right lung, with an average diameter of about 7.00 mm, conforming to Lung-RADS: category 3."

[0205] Chapter 4: Minor lesions

[0206] Video stream: Show pictures of the remaining minor lesions

[0207] Audio stream: BGM

[0208] Chapter 5: Diagnostic conclusion

[0209] Video stream: Scroll and play the diagnostic conclusion text

[0210] Text content: LDCT pulmonary nodule screening, conforming to Lung-RADS category 3 (v1.1 2019)

[0211] Audio stream: Read the diagnostic conclusion aloud

[0212] "Conclusion: LDCT pulmonary nodule screening, conforming to Lung-RADS category 3 (v1.1 2019)."

[0213] Chapter 6: Logical Reasoning Diagram of Diagnostic Conclusions

[0214] Video Stream: Inference Process Picture (Generated Based on Report Data and Lung-RADS Guidelines)

[0215] Figure 5 Shows the logical reasoning diagram of diagnostic conclusions in the system for constructing video-audio reports based on structured reports according to Embodiment 3 of the present invention; as Figure 5 shown.

[0216] Chapter 7: Rolling Display and Reading of Suggested Follow-up Examinations and Follow-up Plans

[0217] It is recommended to perform the next LDCT scan after 3 - 6 months to observe the changes in the nodules. If the nodules increase in size or new suspicious features appear during the follow-up period, further diagnostic examinations such as PET-CT scans or biopsies may be required.

[0218] If the nodules remain stable during short-term follow-up, it may be recommended to perform an LDCT scan once a year to monitor the long-term changes in the nodules.

[0219] From the above description, it can be seen that the above embodiments of the present invention achieve the following technical effects: The embodiments of the present invention are based on the inherent standardization and diagnostic logic of imaging structured reports, and automatically construct a video-audio stream for interpreting reports for patients / clinicians. Without the need for radiologists to communicate face-to-face with patients / clinicians, it can largely achieve the effect of face-to-face communication. Without affecting the work efficiency of the radiology department, it essentially improves the satisfaction of clinicians and patients with the work of the radiology department; The most core part of the video stream demonstrated by the method of the embodiments of the present invention is the diagnostic Guideline and the diagnostic logical reasoning diagram, which can display the medical practice guidelines and evaluation criteria corresponding to the imaging manifestations, as well as the diagnostic logical reasoning diagram of how to draw diagnostic conclusions based on the guidelines and imaging manifestations. In the scenario where radiologists do not need to explain face-to-face, clinicians / patients can deeply understand the condition through the video stream.

[0220] Obviously, those skilled in the art should understand that the above-mentioned modules or steps of the present invention can be implemented by a general-purpose computing device. They can be concentrated on a single computing device or distributed on a network composed of multiple computing devices. Optionally, they can be implemented by program codes executable by the computing device, so that they can be stored in a storage device and executed by the computing device, or they can be separately made into individual integrated circuit modules, or multiple modules or steps among them can be made into a single integrated circuit module to implement. In this way, the present invention is not limited to any specific combination of hardware and software.

[0221] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for constructing an audio-visual report based on a diagnostic logic of a structured report, characterized in that, Including: Based on the preset single-disease examination item conditions, create corresponding video templates and define the names of the video templates; Set the video template content for each video template. The video template content includes various combinations of the following scene segments: patient information, diagnostic guidelines, main lesions, secondary lesions, diagnostic conclusions, logical reasoning diagrams of the diagnostic conclusions, recommended subsequent examination and follow-up plans; Configure the audio-visual materials used for each scene segment, including: visual content, background music, reading content, and whether to generate subtitles; After the image structured report is reviewed, select the video template corresponding to the image structured report, obtain the relevant data of the video template content from the image structured report, and based on the audio-visual materials, fill in the data and render each scene segment in the video template to synthesize a complete and coherent audio-visual image report.

2. The method for constructing an audiovisual report based on a diagnostic logic of a structured report according to claim 1, wherein The image structured report has built-in diagnostic logic, and the diagnostic conclusions and the logical reasoning diagrams of the diagnostic conclusions are generated through the diagnostic logic and image manifestations.

3. The method for constructing an audiovisual report based on a diagnostic logic of a structured report according to claim 1, wherein The video stream shown in the scene segment includes the following content: The patient information, and its video stream shows the clinical purpose of the imaging examination, the scanning time of the imaging examination, and the scanning method; The diagnostic guidelines, which are the diagnostic logics involved in the image structured report, and its video stream shows the criteria of the diagnostic guidelines; The main lesions, and its video stream shows the pictures of the main lesions, the lesion information marked on the pictures, and the measurement values of the lesions; The secondary lesions, and its video stream shows the pictures of the secondary lesions; The diagnostic conclusions, and its video stream shows the text of the diagnostic conclusions; The logical reasoning diagram of the diagnostic conclusions, and its video stream shows the logical reasoning diagram of the diagnostic conclusions. The logical reasoning diagram of the diagnostic conclusions shows how to infer the current diagnostic conclusions according to the diagnostic guidelines and is automatically drawn and generated according to the built-in diagnostic logic and image manifestations of the image structured report; The recommended subsequent examination and follow-up plans are generated based on preset rules or the built-in logic of the image structured report.

4. The method for constructing an audiovisual report based on a diagnostic logic of a structured report according to claim 3, characterized in that, Based on the audio-visual materials, filling in the data and rendering each scene segment in the video template to synthesize a complete and coherent audio-visual image report, the steps include: Through video stream template design, export a JSON format to be processed by a computer script; The computer script replaces the placeholder content with the actual data required for the video template content and renders to generate visual content; Based on text-to-speech technology, convert the reading content into speech and a matching subtitle stream as sound, and synchronize the subtitles and the speech using time codes; After synchronizing the video stream and the audio stream of each scene segment, synthesize a complete and coherent audio-visual image report.

5. The method for constructing an audiovisual report based on a diagnostic logic of a structured report according to claim 3, wherein The relevant information of the main lesions and secondary lesions, and the acquisition method includes: Through the integration of the image structured report with an image-assisted diagnosis AI module or a post-processing module, obtain the key images and measurement values output by the image-assisted diagnosis AI module or the post-processing module, fill them in specific positions of the image structured report, and assign corresponding ontology labels; The imaging structured report fills in the imaging structured report by means of DICOM images transmitted by imaging equipment and attached GSPS status small files, and assigns corresponding ontology tags; Based on a dedicated interface and using batch processing tools, patient images and measurement values are sent to the imaging structured report in real time, and at the same time, the ID of the current patient image and the GSPS status are obtained.

6. A system for constructing an audiovisual report based on a diagnostic logic of a structured report, characterized in that, The system includes: a video template creation module, a template content setting module, an audio-visual configuration module, and an audio-visual synthesis module, where, The video template creation module is connected to the template content setting module and is used to create a corresponding video template based on preset single-disease examination item conditions and define the name of the video template; The template content setting module is respectively connected to the video template creation module, the audio-visual configuration module, and the audio-visual synthesis module, and is used to set the video template content for each video template. The video template content includes various combinations in the following scenario segments: patient information, diagnostic Guideline, main lesions, secondary lesions, diagnostic conclusions, the logical reasoning diagram of the diagnostic conclusions, recommended follow-up examinations and follow-up plans; The audio-visual configuration module is respectively connected to the template content setting module and the audio-visual synthesis module, and is used to configure the audio-visual materials used for each scenario segment, including: visual content, background music, reading content, and whether to generate subtitles; The audio-visual synthesis module is respectively connected to the template content setting module and the audio-visual configuration module, and is used to, after the imaging structured report is reviewed, select the video template corresponding to the imaging structured report, obtain the relevant data of the video template content from the imaging structured report, and based on the audio-visual materials, perform data filling and rendering on each scenario segment in the video template to synthesize a complete and coherent audio-visual imaging report.

7. The system for constructing an audiovisual report based on the diagnostic logic of structured reports according to claim 6, characterized in that, The imaging structured report has built-in diagnostic logic, and the diagnostic conclusions and the logical reasoning diagram of the diagnostic conclusions are generated through the diagnostic logic and imaging manifestations.

8. The system for constructing an audiovisual report based on the diagnostic logic of structured reports according to claim 6, characterized in that, The video stream displayed by the scenario segment includes the following content: The patient information, and its video stream displays the clinical purpose of the imaging examination, the scanning time of the imaging examination, and the scanning method; The diagnostic Guideline, which is the diagnostic logic involved in the imaging structured report, and its video stream displays the criteria of the diagnostic Guideline; The main lesions, and its video stream displays pictures of the main lesions, the lesion information marked in the pictures, and the measurement values of the lesions; The secondary lesions, and its video stream displays pictures of the secondary lesions; The diagnostic conclusions, and its video stream displays the text of the diagnostic conclusions; The logical reasoning diagram of the diagnostic conclusions, and its video stream displays the logical reasoning diagram of the diagnostic conclusions. The logical reasoning diagram of the diagnostic conclusions shows how to infer the current diagnostic conclusions according to the diagnostic Guideline and is automatically drawn and generated based on the built-in diagnostic logic and imaging manifestations of the imaging structured report; The recommended follow-up examinations and follow-up plans are generated based on preset rules or the built-in logic of the imaging structured report.

9. The system for constructing an audiovisual report based on a diagnostic logic of a structured report according to claim 8, wherein The video and audio synthesis module includes: a format conversion unit, a data processing unit, a synchronization unit, and an output unit. Among them, The format conversion unit is used to export a JSON format processed by a computer script through video stream template design; The data processing unit is used to replace the placeholder content with the actual data required by the video template content by a computer script and render visual content; The synchronization unit is used to convert the recited content into speech and a matching subtitle stream based on text-to-speech technology, and synchronize the subtitle with the speech using time codes; The output unit is used to synthesize a complete and coherent video and audio image report after synchronizing the video stream and audio stream of each scene segment.

10. The system for constructing an audiovisual report based on the diagnostic logic of structured reports according to claim 8, wherein, The relevant information of the main lesion and the secondary lesion, and the acquisition method includes: By integrating the image structured report with the image-assisted diagnosis AI module or the post-processing module, obtaining the key images and measurement values output by the image-assisted diagnosis AI module or the post-processing module, filling them into specific positions of the image structured report, and assigning corresponding ontology labels; The image structured report fills in the image structured report through DICOM images transmitted by imaging devices and attached GSPS status small files, and assigns corresponding ontology labels; Based on a dedicated interface, using a batch processing tool, patient images and measurement values are sent to the image structured report in real time, and at the same time, the ID of the current patient image and the GSPS status are obtained.

Images