A Doppler-guided interpharyngeal injection system for effectively treating the sensation of a foreign body in the throat.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-05-27
- Publication Date
- 2026-08-14
Smart Images
Figure CN122557166A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of medical ultrasound image guidance and interpharyngeal injection technology, and particularly relates to an interpharyngeal injection system for effectively treating foreign body sensation in the throat under Doppler precise guidance. Background Technology
[0002] The sensation of a foreign body in the throat often manifests as discomfort, tightness, difficulty swallowing, recurrent throat clearing, or throat pain. Its occurrence may be related to the local tissue condition of the pharyngeal wall, tension in the parapharyngeal soft tissues, stimulation of the retropharyngeal or parapharyngeal space, local blood flow, and response to previous treatments. When performing interpharyngeal injection, it is necessary to accurately identify the retropharyngeal space, parapharyngeal space, carotid sheath area, internal jugular vein course, glandular adjacent areas, and areas subject to acoustic interference. Based on the symptom phenotype, a suitable injection target area and safe needle insertion path must be determined. Therefore, a Doppler-guided interpharyngeal injection system is needed to effectively treat the sensation of a foreign body in the throat, requiring joint analysis of pharyngeal Doppler images, anatomical layers of the interpharyngeal space, the phenotype of the sensation of a foreign body in the throat, and the risks associated with the injection path.
[0003] The existing technology has at least the following problems that need to be improved: (1) Existing injection treatments for pharyngeal foreign body sensation mostly rely on the operator's experience to determine the injection area. There is a lack of a mechanism for identifying the anatomical layers of the pharyngeal space based on basic two-dimensional ultrasound images, color Doppler images, energy Doppler images, and spectral Doppler data. It is difficult to make a joint judgment on the boundaries of the pharyngeal wall, the pharyngeal cavity, the retropharyngeal space, the parapharyngeal space, the carotid sheath, the internal jugular vein, the glandular adjacent area, and the acoustic shadowing area, resulting in insufficient stability of the pharyngeal space injection target area.
[0004] (2) Existing injection guidance methods focus more on the needle insertion location itself and lack a phenotypic assessment mechanism that combines data on the duration of pharyngeal foreign body sensation, pharyngeal tightness score, swallowing discomfort score, frequency of repeated throat clearing, previous treatment records, and contraindication risk data. They also lack a mechanism for scoring the safety of the injection path based on the distance of the blood vessel avoidance path, the risk of pharyngeal cavity crossing, the risk of gland proximity, and the area of acoustic interference. This results in insufficient matching between the injection target area and the symptom phenotype, and insufficient safety and process verifiability of the injection path. Summary of the Invention
[0005] To address the above issues and overcome the shortcomings of existing technologies, this invention provides a Doppler-guided interpharyngeal injection system for effectively treating pharyngeal foreign body sensation. Utilizing Doppler image multi-channel fusion, convolutional neural network hierarchical recognition, pharyngeal foreign body sensation phenotype assessment, injection target area adaptation decision-making, injection path safety scoring, and process feedback updates, the system generates original Doppler image data of the pharynx, an anatomical hierarchy map of the interpharyngeal space, pharyngeal foreign body sensation phenotype assessment results, the target injection area of the interpharyngeal space, the interpharyngeal injection guidance plan, and interpharyngeal injection samples. This achieves anatomical hierarchy recognition of the interpharyngeal space, symptom phenotype adaptation, precise target area selection, path risk avoidance, and closed-loop updates during the injection process. It solves the problems of existing pharyngeal foreign body sensation injection treatments, such as reliance on experience for target area localization, insufficient symptom adaptation, inadequate path risk assessment, and insufficient process feedback updates.
[0006] This invention provides a Doppler-guided system for effective treatment of pharyngeal foreign body sensation through interpharyngeal injection, comprising a pharyngeal Doppler image acquisition module, an interpharyngeal anatomical layer recognition module, a pharyngeal foreign body sensation phenotype assessment module, an interpharyngeal injection target area decision module, an injection path safety guidance module, and an injection process feedback update module.
[0007] The pharyngeal Doppler image acquisition module acquires basic two-dimensional ultrasound images, color Doppler images, energy Doppler images, spectral Doppler data, probe posture data, probe compression data, and image frame timestamp data of the pharyngeal region of the treated subject, and generates raw pharyngeal Doppler image data; The pharyngeal space anatomical layer recognition module receives raw Doppler image data of the pharynx and identifies the pharyngeal wall boundary, parapharyngeal soft tissue layers, retropharyngeal space region, parapharyngeal space region, carotid sheath region, internal jugular vein course region, glandular adjacent region and acoustic shadowing interference region based on convolutional neural network, and generates a pharyngeal space anatomical layer map; The pharyngeal foreign body sensation phenotype assessment module acquires data on the duration of pharyngeal foreign body sensation, pharyngeal tightness score, swallowing discomfort score, frequency of repeated throat clearing, pharyngeal pain score, previous treatment records, and contraindication risk data of the treated subject, and generates pharyngeal foreign body sensation phenotype assessment results. The pharyngeal space injection target area decision module receives the anatomical layer map of the pharyngeal space and the pharyngeal foreign body sensation phenotype assessment results. Based on the assessment results of the retropharyngeal space region, parapharyngeal space region, pharyngeal wall boundary, carotid sheath region, internal jugular vein course region, acoustic shadowing region, and pharyngeal foreign body sensation phenotype, it calculates the pharyngeal space injection target area score and selects the pharyngeal space target injection area based on the pharyngeal space injection target area score. The injection path safety guidance module receives the target injection area and anatomical layer diagram of the pharyngeal space. Starting from the needle entry point on the body surface and ending at the target injection area in the pharyngeal space, it generates multiple candidate injection paths. Based on the spatial relationship between the candidate injection paths and the carotid sheath region, the internal jugular vein region, the pharyngeal cavity boundary, the glandular adjacent region, and the acoustic shadowing interference region, it calculates the injection path safety score and generates a pharyngeal space injection guidance plan based on the injection path safety score. The injection process feedback update module receives the pharyngeal interstitial injection guidance plan, and collects real-time Doppler image data, needle tip position data, local blood flow change data, local tissue echo change data, injection diffusion range data, and the patient's reaction data during the pharyngeal interstitial injection process. It generates injection process monitoring results, and generates pharyngeal interstitial injection samples based on the injection process monitoring results, postoperative follow-up results, and manual verification results. Based on the pharyngeal interstitial injection samples, it updates the parameters of the pharyngeal interstitial anatomical layer identification module, the pharyngeal foreign body sensation phenotype assessment module, the pharyngeal interstitial injection target area decision module, and the injection path safety guidance module.
[0008] Furthermore, the pharyngeal Doppler image acquisition module performs sharpness detection, motion artifact detection, acoustic shadowing detection, and blood flow display integrity detection on the basic two-dimensional ultrasound image, color Doppler image, and energy Doppler image, and removes image frames with image quality below the quality threshold to obtain effective pharyngeal Doppler image frames; speckle noise suppression, grayscale normalization, edge enhancement, and local contrast enhancement are performed on the effective pharyngeal Doppler image frames to obtain enhanced pharyngeal Doppler image frames.
[0009] Furthermore, the pharyngeal space anatomical layer recognition module performs channel stitching on the enhanced pharyngeal Doppler image frame, color Doppler image, energy Doppler image, and pharyngeal blood flow display data to obtain a multi-channel input image of the pharynx. The multi-channel input image of the pharynx is then input into a convolutional neural network to extract pharyngeal wall boundary features, pharyngeal cavity boundary features, parapharyngeal soft tissue layer features, glandular adjacent area features, retropharyngeal space area features, parapharyngeal space area features, carotid sheath area features, internal jugular vein course area features, and acoustic shadowing interference area features.
[0010] Furthermore, the pharyngeal interstitial injection target area decision module extracts multiple injection candidate areas based on the probability maps of the retropharyngeal space, parapharyngeal space, pharyngeal wall boundary, and pharyngeal cavity boundary; based on the probability maps of the carotid sheath region and the internal jugular vein course region, it eliminates the risk of injection candidate areas near the carotid sheath region or the internal jugular vein course region to obtain safe injection candidate areas; based on the interstitial space integrity, vascular avoidance distance, pharyngeal foreign body sensation phenotype assessment results, acoustic interference degree, and contraindication risk data of the safe injection candidate areas, the pharyngeal interstitial injection target area score is calculated.
[0011] Furthermore, the injection path safety guidance module calculates the minimum distance between each candidate injection path line and the carotid sheath region and the internal jugular vein course region to obtain the path vessel avoidance distance; based on the overlap relationship between each candidate injection path line and the pharyngeal cavity boundary, it determines whether there is a risk of pharyngeal cavity boundary crossing; based on the spatial relationship between each candidate injection path line and the glandular adjacent region, it determines whether there is a risk of glandular proximity; based on the path vessel avoidance distance, pharyngeal cavity boundary crossing risk, glandular proximity risk, acoustic shadowing interference region, and pharyngeal interstitial injection target area score, the injection path safety score is calculated.
[0012] Furthermore, the injection process feedback update module generates injection process monitoring results based on needle tip position deviation, injection diffusion range judgment results, blood flow abnormality judgment results, and the treatment subject's reaction data; it generates pharyngeal interstitial injection samples based on the injection process monitoring results, postoperative follow-up results, and manual verification results, and updates the parameters of the pharyngeal interstitial anatomical layer identification module, pharyngeal foreign body sensation phenotype assessment module, pharyngeal interstitial injection target area decision module, and injection path safety guidance module based on the pharyngeal interstitial injection samples.
[0013] The beneficial effects of this invention are as follows: (1) Using multi-channel fusion of Doppler images and hierarchical recognition of convolutional neural networks, the pharyngeal Doppler image acquisition module acquires basic two-dimensional ultrasound images, color Doppler images, energy Doppler images, spectral Doppler data, probe posture data, probe compression data and image frame timestamp data to generate original pharyngeal Doppler image data; the enhanced pharyngeal Doppler image frame, color Doppler image, energy Doppler image and pharyngeal blood flow display data are spliced by the pharyngeal space anatomical layer recognition module, and the pharyngeal wall boundary, parapharyngeal soft tissue layer, retropharyngeal space region, parapharyngeal space region, carotid sheath region, internal jugular vein course region, gland adjacent region and acoustic shadowing interference region are identified based on convolutional neural network. The technical effect is to improve the integrity of pharyngeal space anatomical layer recognition and target area localization basis, and solve the problem that the existing injection treatment for pharyngeal foreign body sensation mainly relies on experience localization and is difficult to accurately distinguish the pharyngeal space from surrounding risk structures.
[0014] (2) Using the pharyngeal foreign body sensation phenotype assessment and target area adaptation decision, the pharyngeal foreign body sensation phenotype assessment module normalizes and splices the data on the duration of pharyngeal foreign body sensation, pharyngeal tightness score, swallowing discomfort score, frequency of repeated throat clearing, pharyngeal pain score, previous treatment record data and contraindication risk data to generate a pharyngeal foreign body sensation phenotype feature vector, and identifies the persistent foreign body sensation type, swallowing-induced type, tightness and compression type, repeated throat clearing type and recurrence risk type; the pharyngeal interspace injection target area decision module calculates the pharyngeal interspace injection target area score by combining the pharyngeal interspace anatomical layer map and the pharyngeal foreign body sensation phenotype assessment results, and selects the pharyngeal interspace target injection area. The technical effect is to improve the matching between the injection target area and the pharyngeal foreign body sensation symptom phenotype, and solve the problem that the existing injection treatment does not fully combine the symptom persistence characteristics, inducing factors and previous treatment response.
[0015] (3) By using injection path safety scoring and process feedback updates, multiple injection path candidate lines are generated by the injection path safety guidance module, starting from the needle entry point on the body surface and ending at the target injection area in the pharyngeal space. The injection path safety score is calculated based on the path vascular avoidance distance, pharyngeal cavity boundary crossing risk, glandular proximity risk, acoustic shadow interference area, and pharyngeal space injection target area score, and a pharyngeal space injection guidance plan is generated. The injection process feedback update module generates injection process monitoring results based on needle tip position deviation, injection diffusion range judgment results, blood flow abnormality judgment results, and the treatment subject's reaction data. The parameters are updated based on the pharyngeal space injection sample. The technical effect is to improve the safety of injection path selection and the verifiability of the injection process, and solve the problem of the lack of comprehensive judgment on vascular avoidance, pharyngeal cavity boundary crossing, and glandular proximity risk in existing injection paths. Attached Figure Description
[0016] The accompanying drawings are provided to further understand the present invention and form part of the specification. They are used together with the embodiments of the present invention to explain the invention and do not constitute a limitation thereof.
[0017] Figure 1 This is a flowchart of the overall process of the interpharyngeal injection system proposed in this invention; Figure 2 This is a flowchart of the pharyngeal space anatomical layer recognition module proposed in this invention; Figure 3 This is a flowchart of the pharyngeal interstitial injection target area decision-making and path safety guidance proposed in this invention. Detailed Implementation
[0018] Example 1, see Figures 1-3The present invention provides a Doppler-guided system for effective treatment of pharyngeal foreign body sensation through interpharyngeal injection, comprising a pharyngeal Doppler image acquisition module, an interpharyngeal anatomical layer recognition module, a pharyngeal foreign body sensation phenotype assessment module, an interpharyngeal injection target area decision module, an injection path safety guidance module, and an injection process feedback update module. The pharyngeal Doppler image acquisition module acquires basic two-dimensional ultrasound images, color Doppler images, energy Doppler images, spectral Doppler data, probe posture data, probe compression data, and image frame timestamp data of the pharyngeal region of the treated object, generates raw pharyngeal Doppler image data, and sends the raw pharyngeal Doppler image data to the pharyngeal space anatomical layer recognition module. The pharyngeal space anatomical layer recognition module receives raw Doppler image data of the pharynx, and identifies the pharyngeal wall boundary, parapharyngeal soft tissue layers, retropharyngeal space region, parapharyngeal space region, carotid sheath region, internal jugular vein course region, glandular adjacent region and acoustic shadowing interference region based on convolutional neural network, generates a pharyngeal space anatomical layer map, and sends the pharyngeal space anatomical layer map to the pharyngeal space injection target area decision module and injection path safety guidance module; The pharyngeal foreign body sensation phenotype assessment module acquires data on the duration of pharyngeal foreign body sensation, pharyngeal tightness score, swallowing discomfort score, frequency of repeated throat clearing, pharyngeal pain score, previous treatment records, and contraindication risk data of the treated subject, generates pharyngeal foreign body sensation phenotype assessment results, and sends the pharyngeal foreign body sensation phenotype assessment results to the pharyngeal interstitial injection target area decision module. The pharyngeal space injection target area decision module receives the anatomical layer map of the pharyngeal space and the pharyngeal foreign body sensation phenotype assessment results. Based on the assessment results of the retropharyngeal space region, parapharyngeal space region, pharyngeal wall boundary, carotid sheath region, internal jugular vein course region, acoustic shadowing region, and pharyngeal foreign body sensation phenotype, it calculates the pharyngeal space injection target area score and selects the pharyngeal space target injection area based on the pharyngeal space injection target area score. The injection path safety guidance module receives the target injection area and anatomical layer diagram of the pharyngeal space. Starting from the needle entry point on the body surface and ending at the target injection area in the pharyngeal space, it generates multiple candidate injection path lines. Based on the spatial relationship between the candidate injection path lines and the carotid sheath region, the internal jugular vein region, the pharyngeal cavity boundary, the glandular adjacent region, and the acoustic shadowing interference region, it calculates the injection path safety score and generates a pharyngeal space injection guidance plan based on the injection path safety score.
[0019] The injection process feedback update module receives the pharyngeal interstitial injection guidance plan, and collects real-time Doppler image data, needle tip position data, local blood flow change data, local tissue echo change data, injection diffusion range data, and the patient's reaction data during the pharyngeal interstitial injection process. It generates injection process monitoring results, and generates pharyngeal interstitial injection samples based on the injection process monitoring results, postoperative follow-up results, and manual verification results. Based on the pharyngeal interstitial injection samples, it updates the parameters of the pharyngeal interstitial anatomical layer identification module, the pharyngeal foreign body sensation phenotype assessment module, the pharyngeal interstitial injection target area decision module, and the injection path safety guidance module.
[0020] With the cooperation of the above modules, this system can combine Doppler image acquisition, anatomical layer identification of the pharyngeal space, phenotypic assessment of pharyngeal foreign body sensation, injection target area decision-making, safe guidance of injection path, and feedback updates of the injection process, thereby improving the safety and stability of target area localization, path selection, and process monitoring in pharyngeal space injection.
[0021] Example 2: This example is based on all the above examples. The pharyngeal Doppler image acquisition module specifically includes: Doppler image acquisition; the pharyngeal Doppler image acquisition module acquires basic two-dimensional ultrasound images, color Doppler images, energy Doppler images, and spectral Doppler data of the pharyngeal region of the treated subject.
[0022] The acquisition status record module synchronously records probe posture data, probe compression data, and image frame timestamp data. The probe posture data includes probe tilt angle data, probe rotation angle data, and probe movement speed data.
[0023] Image quality screening: The pharyngeal Doppler image acquisition module performs sharpness detection, motion artifact detection, acoustic shadowing detection, and blood flow display integrity detection on basic two-dimensional ultrasound images, color Doppler images, and energy Doppler images, and removes image frames with image quality below the quality threshold to obtain valid pharyngeal Doppler image frames.
[0024] Image enhancement processing: The pharyngeal Doppler image acquisition module performs speckle noise suppression, grayscale normalization, edge enhancement, and local contrast enhancement on the effective pharyngeal Doppler image frames to obtain enhanced pharyngeal Doppler image frames.
[0025] Blood flow display area extraction: The pharyngeal Doppler image acquisition module extracts the carotid artery blood flow display area, internal jugular vein blood flow display area, and parapharyngeal small vessel blood flow display area from color Doppler images and energy Doppler images to obtain pharyngeal blood flow display data.
[0026] Data binding: The pharyngeal Doppler image acquisition module binds the enhanced pharyngeal Doppler image frame, pharyngeal blood flow display data, spectral Doppler data, probe posture data, probe compression data, and image frame timestamp data to generate the original pharyngeal Doppler image data.
[0027] In this embodiment, the pharyngeal Doppler image acquisition module retains the temporal relationship of continuous image frames, changes in probe posture, and changes in probe pressure, enabling subsequent modules to determine whether the position of the pharyngeal space is affected by swallowing, breathing, or probe pressure.
[0028] Example 3: This example is based on all the above examples. The pharyngeal interspace anatomical layer identification module is one of the core modules of this invention, and specifically includes: The module receives raw Doppler image data of the pharynx; the pharyngeal space anatomical layer recognition module receives raw Doppler image data of the pharynx sent by the Doppler image acquisition module.
[0029] Multi-channel input image construction: The pharyngeal space anatomical layer recognition module stitches together the enhanced pharyngeal Doppler image frames, color Doppler images, energy Doppler images, and pharyngeal blood flow display data to obtain a multi-channel input image of the pharynx.
[0030] Convolutional Neural Network Feature Extraction: The pharyngeal space anatomical layer recognition module inputs multi-channel pharyngeal images into a convolutional neural network. Through shallow convolutional units, it extracts pharyngeal wall boundary features and pharyngeal cavity boundary features. Through mid-layer convolutional units, it extracts parapharyngeal soft tissue layer features and glandular adjacent area features. Through deep semantic units, it extracts features of the retropharyngeal space region, parapharyngeal space region, carotid sheath region, internal jugular vein course region, and acoustic shadowing interference region.
[0031] Anatomical region probability map generation; The pharyngeal space anatomical layer recognition module generates a probability map of the retropharyngeal space based on the features of the retropharyngeal space region, a probability map of the parapharyngeal space based on the features of the parapharyngeal space region, a probability map of the carotid sheath region based on the features of the carotid sheath region, a probability map of the internal jugular vein region based on the features of the internal jugular vein course region, and a probability map of acoustic interference based on the features of the acoustic interference region.
[0032] The anatomical layer diagram of the pharyngeal space is generated. The anatomical layer recognition module of the pharyngeal space generates the anatomical layer diagram of the pharyngeal space based on the probability diagrams of the pharyngeal wall boundary, pharyngeal cavity boundary, retropharyngeal space, parapharyngeal space, carotid sheath region, internal jugular vein course region, glandular adjacent region, and acoustic shadowing interference.
[0033] Regarding parameter adjustments: Step 1: Adjusting the probability threshold of the retropharyngeal space; When the retropharyngeal space region has a continuous boundary and the image clarity is high, the probability threshold of the retropharyngeal space is increased to reduce false identification of non-target soft tissue. When the sound and shadow interference area covers the edge of the retropharyngeal space area, the probability threshold of the retropharyngeal space is reduced and the number of consecutive frames fused is increased.
[0034] Step 2: Adjustment of parapharyngeal space probability threshold; When the layers of parapharyngeal soft tissue are clear and the boundaries of the carotid sheath region are well defined, increase the probability threshold of the parapharyngeal space. When the parapharyngeal soft tissue layers are displaced due to probe compression, the parapharyngeal space probability threshold is reduced, and the weight of probe compression data in regional correction is increased.
[0035] Step 3: Adjusting the weights for blood vessel region recognition; When the carotid artery blood flow display area or the internal jugular vein blood flow display area is close to the candidate area of the pharyngeal space, increase the weight of the probability map of the carotid sheath area and the probability map of the internal jugular vein course area. Increase the weight of spectral Doppler data and energy Doppler images in vascular region identification when the blood flow display area is unstable in consecutive image frames.
[0036] Step 4: Adjusting the sound and shadow interference suppression parameters; When the acoustic shadowing interference probability map covers the center of the candidate region of the pharyngeal space, the penalty weight corresponding to the acoustic shadowing interference probability map is increased. When the acoustic shadowing probability map is only located at the edge of the candidate region of the pharyngeal space, the intensity of acoustic shadowing suppression is reduced, and the corresponding region is retained as a candidate target area.
[0037] By performing the above operations, the pharyngeal space anatomical layer recognition module can stably identify the spatial hierarchical relationship between the pharyngeal wall, pharyngeal cavity, retropharyngeal space, parapharyngeal space, and adjacent vascular structures.
[0038] Example 4: This example is based on all the above examples. The pharyngeal foreign body sensation phenotype assessment module specifically includes: Symptom data acquisition: The pharyngeal foreign body sensation phenotype assessment module acquires data on the duration of pharyngeal foreign body sensation, pharyngeal tightness score, swallowing discomfort score, frequency of repeated throat clearing, and pharyngeal pain score of the treated subjects.
[0039] Previous treatment data acquisition; The pharyngeal foreign body sensation phenotype assessment module acquires the previous treatment record data of the treated subject, including previous drug treatment records, previous nebulization treatment records, previous injection treatment records, and previous recurrence records.
[0040] Contraindication risk data acquisition: The pharyngeal foreign body sensation phenotype assessment module acquires contraindication risk data of the treated subjects, including coagulation abnormality risk data, local infection risk data, drug allergy risk data, and risk data of proximity to important blood vessels.
[0041] Phenotypic feature vector generation for pharyngeal foreign body sensation: The pharyngeal foreign body sensation phenotypic assessment module normalizes and concatenates data on the duration of pharyngeal foreign body sensation, pharyngeal tightness score, swallowing discomfort score, frequency of repeated throat clearing, pharyngeal pain score, previous treatment records, and contraindication risk data to generate a pharyngeal foreign body sensation phenotypic feature vector.
[0042] The pharyngeal foreign body sensation phenotype assessment results are generated. The pharyngeal foreign body sensation phenotype assessment module identifies persistent foreign body sensation, swallowing-induced type, constriction and compression type, repetitive throat clearing type, and recurrence risk type based on the pharyngeal foreign body sensation phenotype feature vector, and generates pharyngeal foreign body sensation phenotype assessment results.
[0043] In this embodiment, the pharyngeal foreign body sensation phenotype assessment module does not only record symptoms, but also incorporates the duration of symptoms, swallowing triggers, previous treatments, and contraindications into the target area decision-making process, making the pharyngeal interstitial injection guidance plan more suitable for the patient's condition.
[0044] Example 5: This example is based on all the above examples. The pharyngeal interstitial injection target area decision module is another core module of the present invention, specifically including: The pharyngeal space anatomical layer diagram is received; the pharyngeal space injection target area decision module receives the pharyngeal space anatomical layer diagram sent by the pharyngeal space anatomical layer recognition module.
[0045] The pharyngeal foreign body sensation phenotype assessment results are received; the pharyngeal interstitial injection target area decision module receives the pharyngeal foreign body sensation phenotype assessment results sent by the pharyngeal foreign body sensation phenotype assessment module.
[0046] Injection candidate region generation; the pharyngeal interstitial injection target region decision module extracts multiple injection candidate regions based on the probability map of the retropharyngeal space, the probability map of the parapharyngeal space, the boundary of the pharyngeal wall, and the boundary of the pharyngeal cavity.
[0047] Vascular risk exclusion: The pharyngeal interstitial injection target area decision module excludes potential injection areas near the carotid sheath region or the internal jugular vein region based on the probability map of the carotid sheath region and the probability map of the internal jugular vein course region, thus obtaining safe injection candidate areas.
[0048] Target adaptation score calculation: The pharyngeal interstitial injection target zone decision module calculates the pharyngeal interstitial injection target zone score based on the interstitial integrity, vascular avoidance distance, pharyngeal foreign body sensation phenotype assessment results, acoustic interference degree, and contraindication risk data of the safe injection candidate area.
[0049] The pharyngeal interstitial injection target area is generated. The pharyngeal interstitial injection target area decision module sorts the safe injection candidate areas according to the pharyngeal interstitial injection target area score, selects the safe injection candidate area with the highest pharyngeal interstitial injection target area score and meets the safety threshold, and generates the pharyngeal interstitial injection target area.
[0050] Regarding parameter adjustments: Step 1: Adjusting the blood vessel avoidance distance threshold; When the carotid sheath region or the course of the internal jugular vein is close to the safe injection candidate area, increase the vascular avoidance distance threshold. When the confidence level for vascular region identification is low, the risk boundary of the carotid sheath region and the course of the internal jugular vein is expanded, and the interpharyngeal injection target area score of adjacent safe injection candidate regions is reduced.
[0051] Step 2: Adjusting the weighting of the pharyngeal foreign body sensation phenotype; When the phenotype assessment result of pharyngeal foreign body sensation is swallowing-induced, increase the corresponding weights of the probability maps of pharyngeal wall boundary and retropharyngeal space. When the phenotype assessment result of pharyngeal foreign body sensation is tightness and compression type, increase the corresponding weights of the parapharyngeal space probability map and the parapharyngeal soft tissue layer features. When the pharyngeal foreign body sensation phenotype assessment result is a recurrence risk type, the corresponding weights of previous treatment record data and contraindication risk data are increased.
[0052] Step 3: Adjusting the weight of sound and shadow interference levels; When the level of acoustic interference covers the center of the safe injection candidate area, the penalty weight corresponding to the level of acoustic interference is increased. When the acoustic interference level is only located at the edge of the safe injection candidate area, the penalty weight corresponding to the acoustic interference level is reduced, and the corresponding area is retained as a candidate target area.
[0053] Step 4: Adjust the safety threshold; When the risk of contraindication increases, the safety threshold is raised and a manual review prompt is triggered. When the anatomical layer diagram of the pharyngeal space is continuous and stable and the vascular avoidance distance meets the requirements, the safety threshold can be appropriately reduced to improve the generation efficiency of the pharyngeal space target injection area.
[0054] By performing the above operations, the pharyngeal interstitial injection target area decision module can comprehensively consider the pharyngeal interstitial structure, vascular risk, symptom phenotype, and contraindication risk to screen out suitable pharyngeal interstitial target injection areas for the treatment of pharyngeal foreign body sensation.
[0055] Example 6: This example is based on all the above examples. The injection path safety guidance module specifically includes: The pharyngeal interstitial injection target area is received; the injection path safety guidance module receives the pharyngeal interstitial injection target area sent by the pharyngeal interstitial injection target area decision module.
[0056] Injection path candidate line generation: The injection path safety guidance module generates multiple injection path candidate lines, starting from the needle entry point on the body surface and ending at the target injection area in the pharyngeal space.
[0057] Blood vessel avoidance judgment; The injection path safety guidance module calculates the minimum distance between each candidate injection path line and the carotid sheath region and the internal jugular vein course region to obtain the blood vessel avoidance distance of the path.
[0058] Pharyngeal cavity boundary crossing judgment: The injection path safety guidance module determines whether there is a risk of pharyngeal cavity boundary crossing based on the overlap relationship between each candidate injection path line and the pharyngeal cavity boundary.
[0059] Glandular proximity assessment: The injection path safety guidance module determines whether there is a risk of glandular proximity based on the spatial relationship between each candidate injection path line and the adjacent area of the gland.
[0060] Injection path safety score calculation: The injection path safety guidance module calculates the injection path safety score based on the distance of blood vessel avoidance, risk of pharyngeal cavity crossing, risk of gland proximity, acoustic interference area, and injection target area score in the interpharyngeal space.
[0061] The pharyngeal interstitial injection guidance plan is generated. The injection path safety guidance module sorts multiple injection path candidate lines according to the injection path safety score, selects the injection path candidate line with the highest injection path safety score and meets the path safety threshold, and binds the selected injection path candidate line, pharyngeal interstitial target injection area, needle insertion angle prompt, needle insertion depth prompt and risk avoidance prompt to generate the pharyngeal interstitial injection guidance plan.
[0062] In this embodiment, the injection path safety guidance module comprehensively screens the injection path by considering the distance to avoid blood vessels, the risk of pharyngeal boundary crossing, and the risk of gland proxies, thus avoiding the risk area being crossed by simply inserting the needle along the shortest path.
[0063] Example 7: This example is based on all the above examples. The injection process feedback update module specifically includes: The pharyngeal interstitial injection guidance plan is received; the injection process feedback update module receives the pharyngeal interstitial injection guidance plan sent by the injection path safety guidance module.
[0064] Real-time monitoring and data acquisition; the injection process feedback and update module collects real-time Doppler image data, needle tip position data, local blood flow change data, local tissue echo change data, injection diffusion range data, and the treated subject's response data during the interpharyngeal injection process.
[0065] Needle tip position deviation calculation; The injection process feedback update module compares the needle tip position data with the pharyngeal interstitial injection target area in the pharyngeal interstitial injection guidance plan to calculate the needle tip position deviation.
[0066] Injection diffusion range determination: The injection process feedback update module determines whether the injection diffusion is located within the target injection area in the pharyngeal space based on local tissue echo change data and injection diffusion range data, and generates the injection diffusion range determination result.
[0067] Blood flow abnormality judgment; The injection process feedback update module judges whether there are abnormal blood flow changes around the injection path based on local blood flow change data, and generates blood flow abnormality judgment results.
[0068] The injection process monitoring results are generated. The injection process feedback update module generates the injection process monitoring results based on the needle tip position deviation, the injection diffusion range judgment results, the blood flow abnormality judgment results, and the treatment subject's response data.
[0069] Pharyngeal interstitial injection sample generation; The injection process feedback update module generates pharyngeal interstitial injection samples based on the injection process monitoring results, postoperative follow-up results, and manual review results.
[0070] The parameter update module updates the parameters of the pharyngeal space anatomical layer identification module, the pharyngeal foreign body sensation phenotype assessment module, the pharyngeal space injection target area decision module, and the injection path safety guidance module based on the pharyngeal space injection sample.
[0071] By performing the above operations, this system can continuously correct the target area identification, target area decision and path guidance parameters based on the needle tip position, injection diffusion range, blood flow changes and follow-up results during the injection process, thereby improving the stability and safety of subsequent interpharyngeal injection.
Claims
1. A Doppler-guided interpharyngeal injection system for effectively treating pharyngeal foreign body sensation, characterized in that: It includes a pharyngeal Doppler image acquisition module, a pharyngeal interspace anatomical layer recognition module, a pharyngeal foreign body sensation phenotype assessment module, a pharyngeal interspace injection target area decision module, an injection path safety guidance module, and an injection process feedback update module; The pharyngeal Doppler image acquisition module acquires basic two-dimensional ultrasound images, color Doppler images, energy Doppler images, spectral Doppler data, probe posture data, probe compression data, and image frame timestamp data of the pharyngeal region of the treated subject, and generates raw pharyngeal Doppler image data; The pharyngeal space anatomical layer recognition module receives raw Doppler image data of the pharynx and identifies the pharyngeal wall boundary, parapharyngeal soft tissue layers, retropharyngeal space region, parapharyngeal space region, carotid sheath region, internal jugular vein course region, glandular adjacent region and acoustic shadowing interference region based on convolutional neural network, and generates a pharyngeal space anatomical layer map; The pharyngeal foreign body sensation phenotype assessment module acquires data on the duration of pharyngeal foreign body sensation, pharyngeal tightness score, swallowing discomfort score, frequency of repeated throat clearing, pharyngeal pain score, previous treatment records, and contraindication risk data of the treated subject, and generates pharyngeal foreign body sensation phenotype assessment results. The pharyngeal space injection target area decision module receives the anatomical layer map of the pharyngeal space and the pharyngeal foreign body sensation phenotype assessment results. Based on the assessment results of the retropharyngeal space region, parapharyngeal space region, pharyngeal wall boundary, carotid sheath region, internal jugular vein course region, acoustic shadowing region, and pharyngeal foreign body sensation phenotype, it calculates the pharyngeal space injection target area score and selects the pharyngeal space target injection area based on the pharyngeal space injection target area score. The injection path safety guidance module receives the target injection area and anatomical layer diagram of the pharyngeal space. Starting from the needle entry point on the body surface and ending at the target injection area in the pharyngeal space, it generates multiple candidate injection paths. Based on the spatial relationship between the candidate injection paths and the carotid sheath region, the internal jugular vein region, the pharyngeal cavity boundary, the glandular adjacent region, and the acoustic shadowing interference region, it calculates the injection path safety score and generates a pharyngeal space injection guidance plan based on the injection path safety score. The injection process feedback update module receives the pharyngeal interstitial injection guidance plan, and collects real-time Doppler image data, needle tip position data, local blood flow change data, local tissue echo change data, injection diffusion range data, and the patient's reaction data during the pharyngeal interstitial injection process. It generates injection process monitoring results, and generates pharyngeal interstitial injection samples based on the injection process monitoring results, postoperative follow-up results, and manual verification results. Based on the pharyngeal interstitial injection samples, it updates the parameters of the pharyngeal interstitial anatomical layer identification module, the pharyngeal foreign body sensation phenotype assessment module, the pharyngeal interstitial injection target area decision module, and the injection path safety guidance module.
2. The pharyngeal interstitial injection system for effectively treating pharyngeal foreign body sensation under Doppler precision guidance according to claim 1, characterized in that: The pharyngeal Doppler image acquisition module performs sharpness detection, motion artifact detection, sound shadow occlusion detection, and blood flow display integrity detection on basic two-dimensional ultrasound images, color Doppler images, and energy Doppler images, and removes image frames with image quality below the quality threshold to obtain effective pharyngeal Doppler image frames; The effective pharyngeal Doppler image frames were subjected to speckle noise suppression, grayscale normalization, edge enhancement, and local contrast enhancement to obtain enhanced pharyngeal Doppler image frames.
3. The pharyngeal interstitial injection system for effectively treating pharyngeal foreign body sensation under Doppler precision guidance according to claim 2, characterized in that: The pharyngeal Doppler image acquisition module extracts the carotid artery blood flow display area, internal jugular vein blood flow display area, and parapharyngeal small vessel blood flow display area from color Doppler images and energy Doppler images to obtain pharyngeal blood flow display data. The enhanced pharyngeal Doppler image frame, pharyngeal blood flow display data, spectral Doppler data, probe posture data, probe compression data, and image frame timestamp data are bound together to generate the original pharyngeal Doppler image data.
4. The pharyngeal interstitial injection system for effectively treating pharyngeal foreign body sensation under Doppler precision guidance according to claim 3, characterized in that: The pharyngeal space anatomical layer recognition module performs channel stitching on enhanced pharyngeal Doppler image frames, color Doppler images, energy Doppler images, and pharyngeal blood flow display data to obtain a multi-channel input image of the pharynx. The multi-channel input image of the pharynx is then input into a convolutional neural network. Shallow convolutional units extract pharyngeal wall boundary features and pharyngeal cavity boundary features, mid-layer convolutional units extract parapharyngeal soft tissue layer features and glandular adjacent area features, and deep semantic units extract features of the retropharyngeal space region, parapharyngeal space region, carotid sheath region, internal jugular vein course region, and acoustic shadowing interference region.
5. The pharyngeal interstitial injection system for effectively treating pharyngeal foreign body sensation under Doppler precision guidance according to claim 4, characterized in that: The pharyngeal space anatomical layer recognition module generates a probability map of the retropharyngeal space based on the features of the retropharyngeal space region, a probability map of the parapharyngeal space based on the features of the parapharyngeal space region, a probability map of the carotid sheath region based on the features of the carotid sheath region, a probability map of the internal jugular vein region based on the features of the internal jugular vein course region, and a probability map of acoustic shadowing interference based on the features of the acoustic shadowing interference region; and generates a pharyngeal space anatomical layer map based on the pharyngeal wall boundary, pharyngeal cavity boundary, probability map of the retropharyngeal space, probability map of the parapharyngeal space, probability map of the carotid sheath region, probability map of the internal jugular vein course region, glandular adjacent region, and probability map of acoustic shadowing interference.
6. The pharyngeal interstitial injection system for effectively treating pharyngeal foreign body sensation under Doppler precision guidance according to claim 5, characterized in that: The pharyngeal foreign body sensation phenotype assessment module normalizes and concatenates data on the duration of pharyngeal foreign body sensation, pharyngeal tightness score, swallowing discomfort score, frequency of repeated throat clearing, pharyngeal pain score, previous treatment records, and contraindication risk data to generate a pharyngeal foreign body sensation phenotype feature vector. Based on the pharyngeal foreign body sensation phenotype feature vector, it identifies persistent foreign body sensation, swallowing-induced type, tightness and compression type, repeated throat clearing type, and recurrence risk type, and generates pharyngeal foreign body sensation phenotype assessment results.
7. The pharyngeal interstitial injection system for effectively treating pharyngeal foreign body sensation under Doppler precision guidance according to claim 6, characterized in that: The pharyngeal interstitial injection target area decision module extracts multiple injection candidate areas based on the probability maps of the retropharyngeal space, parapharyngeal space, pharyngeal wall boundary, and pharyngeal cavity boundary. Based on the probability maps of the carotid sheath region and the internal jugular vein course region, it eliminates risk from injection candidate areas near the carotid sheath region or the internal jugular vein course region, obtaining safe injection candidate areas. Based on the interstitial space integrity, vascular avoidance distance, pharyngeal foreign body sensation phenotype assessment results, acoustic interference degree, and contraindication risk data of the safe injection candidate areas, a pharyngeal interstitial injection target area score is calculated. The safe injection candidate areas are ranked according to the pharyngeal interstitial injection target area score, and the safe injection candidate area with the highest score and meeting the safety threshold is selected to generate the pharyngeal interstitial target injection area.
8. The pharyngeal interstitial injection system for effectively treating pharyngeal foreign body sensation under Doppler precision guidance according to claim 7, characterized in that: The injection path safety guidance module calculates the minimum distance between each candidate injection path line and the carotid sheath region and the internal jugular vein course region to obtain the path vessel avoidance distance; it determines whether there is a risk of pharyngeal boundary crossing based on the overlap relationship between each candidate injection path line and the pharyngeal cavity boundary; it determines whether there is a risk of glandular proximity based on the spatial relationship between each candidate injection path line and the glandular adjacent region; and it calculates the injection path safety score based on the path vessel avoidance distance, pharyngeal boundary crossing risk, glandular proximity risk, acoustic interference region, and pharyngeal interstitial injection target area score.
9. The pharyngeal interstitial injection system for effectively treating pharyngeal foreign body sensation under Doppler precision guidance according to claim 8, characterized in that: The injection path safety guidance module sorts multiple injection path candidate lines according to the injection path safety score, selects the injection path candidate line with the highest injection path safety score and meets the path safety threshold, and binds the selected injection path candidate line, the pharyngeal interstitial injection target area, the needle insertion angle prompt, the needle insertion depth prompt and the risk avoidance prompt to generate a pharyngeal interstitial injection guidance plan.
10. The pharyngeal interstitial injection system for effectively treating pharyngeal foreign body sensation under Doppler precision guidance according to claim 9, characterized in that: The injection process feedback update module compares the needle tip position data with the target injection area in the pharyngeal space in the pharyngeal space injection guidance plan to calculate the needle tip position deviation; it determines whether the injection diffusion is within the target injection area in the pharyngeal space based on local tissue echo change data and injection diffusion range data, generating an injection diffusion range judgment result; it determines whether abnormal blood flow changes occur around the injection path based on local blood flow change data, generating a blood flow abnormality judgment result; it generates an injection process monitoring result based on the needle tip position deviation, injection diffusion range judgment result, blood flow abnormality judgment result, and the treated subject's response data; and it generates a pharyngeal space injection sample based on the injection process monitoring result, postoperative follow-up result, and manual review result.