Intraoperative nerve localization system for thoracic sympathectomy

Abstract

This invention relates to an intraoperative nerve localization system for thoracic sympathectomy, belonging to the field of minimally invasive surgical assistance and medical information processing technology. During thoracoscopic surgery, this system identifies and analyzes candidate regions of the thoracic sympathetic nerve by acquiring intraoperative thoracoscopic images and patient physiological signals. It also verifies the sympathetic nerve properties of the target structure by combining the physiological functional responses induced by intraoperative nerve stimulation. Based on this, the system fuses and calculates the image structural information and functional response information to generate the localization reliability of the thoracic sympathetic nerve, providing the surgeon with localization references and safety prompts for thoracic sympathectomy. This system enables real-time, objective localization of the thoracic sympathetic nerve during surgery, reducing reliance on surgeon experience, lowering the risk of mis- or missed sympathectomy, and improving the safety and consistency of thoracic sympathectomy, thus possessing high clinical application value.

Description

Technical Field

[0001] This invention relates to the field of minimally invasive surgical assistance and medical information processing technology, and more specifically, to an intraoperative nerve localization system for thoracic sympathectomy. Background Technology

[0002] Thoracic sympathectomy is a minimally invasive surgical procedure performed under thoracoscopic guidance, primarily used to treat primary palmar hyperhidrosis, axillary hyperhidrosis, facial flushing, and some sympathetic nerve dysfunction-related diseases. This surgery typically involves identifying and severing or clamping specific segments of the thoracic sympathetic nerve chain under thoracoscopic visualization to inhibit abnormal sympathetic nerve activity. Due to its minimal invasiveness and rapid recovery, thoracic sympathectomy has been widely adopted in clinical practice.

[0003] However, in actual surgery, the precise localization of the thoracic sympathetic nerves remains a critical issue affecting surgical outcomes and safety. The thoracic sympathetic chain is located paravertebrally, medial to the rib heads, and its anatomical location is significantly influenced by individual differences. There are substantial variations in the course of the nerve chain, segmental distribution, and branching structure among different patients. Some patients may also have anatomical variations such as bypass nerves, accessory branches, or segmental fusion. Under the limited field of vision and complex anatomical structures of thoracoscopic surgery, relying solely on anatomical landmarks for localization is prone to errors.

[0004] Current thoracic sympathectomy primarily relies on the surgeon visually identifying nerve structures under thoracoscopy and determining nerve segments based on anatomical experience such as rib numbers and paravertebral locations. This method is highly dependent on the surgeon's experience and skill level. Inexperienced surgeons or those with unclear anatomical structures are prone to mis-cutting, missed cuts, or incorrect segment identification. Inappropriate segment selection can lead to unsatisfactory surgical outcomes and even postoperative complications such as compensatory hyperhidrosis, abnormal heart rate, and Horner's syndrome.

[0005] Furthermore, current neural recognition technologies are mostly structural in nature, relying primarily on differences in nerve appearance, color, or tissue location for differentiation. They lack real-time verification methods to determine whether a target structure truly possesses sympathetic nerve function. In the intraoperative environment, factors such as fat coverage, local bleeding, tissue traction, or changes in lighting conditions can make it difficult to clearly distinguish neural structures from surrounding tissues, further increasing the risk of misjudgment.

[0006] While some studies have attempted to introduce intraoperative electrical stimulation or monitoring to assist in nerve identification, these methods are mostly used for postoperative functional assessment or observation of single physiological indicators, and a systematic localization scheme combining thoracoscopic image information has not yet been formed. These methods typically fail to address the temporal correspondence between image information and physiological responses, and lack a technical approach for comprehensive analysis and unified judgment of multi-source information, making it difficult to provide stable, intuitive, and repeatable reference data for intraoperative nerve resection decisions.

[0007] Therefore, current thoracic sympathectomy procedures still lack a technique that can simultaneously combine thoracoscopic images and neurological response information to accurately locate the thoracic sympathetic nerve in real time, providing the surgeon with reliable cutting references and safety tips. This deficiency limits the standardization and overall safety of the procedure to some extent and requires further improvement.

[0008] Therefore, there is an urgent need for an intraoperative nerve localization system for thoracic sympathectomy to solve these problems. Summary of the Invention

[0009] The purpose of this invention is to solve the technical problems mentioned in the background section and to provide an intraoperative nerve localization system for thoracic sympathectomy, comprising: The system includes a data acquisition module, a neural candidate region identification module, a neural function response detection module, a multimodal fusion localization module, a severance strategy and safety prompt module, and a human-computer interaction and display module. The data acquisition module is used to acquire intraoperative thoracoscopic image data and physiological signal data related to sympathetic nerve function during thoracic sympathectomy. The nerve candidate region identification module identifies and generates nerve candidate regions based on the intraoperative thoracoscopic image data, which are suspected to be the location of the thoracic sympathetic nerve chain. The neural function response detection module is used to stimulate the candidate neural region during surgery and collect the corresponding physiological signal changes before and after stimulation in order to obtain neural function response information. The multimodal fusion localization module is used to fuse the candidate nerve regions, nerve function response information, and intraoperative image information to output the localization results of the thoracic sympathetic nerve and the confidence information of the corresponding segment. The severance strategy and safety warning module generates thoracic sympathetic nerve severance suggestions and risk warning information based on the positioning results; The human-computer interaction and display module is used to display the positioning results, cutting suggestions and risk warning information to the surgeon in real time.

[0010] As a preferred technical solution of the present invention: the data acquisition module includes a thoracoscopic video acquisition unit, a physiological signal acquisition unit, and a time synchronization unit, wherein the physiological signal acquisition unit is used to acquire heart rate signals, skin conductance signals, and / or local skin temperature signals.

[0011] As a preferred technical solution of the present invention: the neural candidate region identification module determines the possible segment range of the thoracic sympathetic nerve based on the rib sequence identification results, and performs candidate labeling of the neural structures within the segment range in combination with the morphological characteristics of the neural tissue.

[0012] As a preferred technical solution of the present invention: the neural function response detection module includes a stimulation control unit and a response analysis unit, wherein the stimulation control unit is used to apply low-intensity electrical stimulation to the neural candidate region, and the response analysis unit is used to analyze the amplitude, trend and duration of changes in physiological signals collected before and after stimulation.

[0013] As a preferred technical solution of the present invention: the multimodal fusion localization module comprehensively analyzes the image feature information and neural function response information of the neural candidate region to determine whether the target structure is the thoracic sympathetic nerve with sympathetic nerve function.

[0014] As a preferred technical solution of the present invention: the multimodal fusion localization module is further used to identify anatomical variations of the thoracic sympathetic nerve, including accessory branches, bypass nerves or segmental fusion structures, and the identification results are used as a reference for calculating localization reliability.

[0015] As a preferred technical solution of the present invention: the cutting strategy and safety prompt module is used to output a safety prompt message to the surgeon that does not recommend cutting or adjusting the cutting position when a high-risk complication is detected in the target nerve.

[0016] As a preferred technical solution of the present invention: the cutting strategy and safety prompt module can generate differentiated cutting suggestions according to the clinical indications corresponding to different thoracic sympathetic nerve segments.

[0017] As a preferred technical solution of the present invention: the human-computer interaction and display module is used to overlay and display the thoracic sympathetic nerve location markers, segment numbers and corresponding confidence information on the real-time thoracoscopy image.

[0018] As a preferred technical solution of the present invention, the system further includes a preoperative anatomical prior model import module, which is used to provide reference constraints for the identification of neural candidate regions and multimodal fusion localization during surgery.

[0019] Compared with the prior art, the intraoperative nerve localization system for thoracic sympathectomy provided by the present invention has at least the following beneficial effects: First, this invention introduces a nerve candidate region identification mechanism based on thoracoscopic images during surgery, combined with image time alignment processing, to systematically analyze the possible locations of the thoracic sympathetic nerves, avoiding reliance solely on surgeon experience and single anatomical landmarks. By comprehensively evaluating the neural structural features in consecutive image frames, it effectively reduces the risk of misjudgment caused by individual anatomical differences, limited field of vision, or tissue occlusion, thus improving the stability and consistency of intraoperative localization of the thoracic sympathetic nerves.

[0020] Secondly, based on structural recognition, this invention further introduces a neural stimulation and physiological function response detection mechanism. By quantitatively analyzing the changes in physiological signals before and after stimulation, it achieves real-time verification of the sympathetic nerve functional attributes of the target structure. This approach overcomes the shortcomings of existing technologies that rely solely on appearance to determine nerve properties, enabling intraoperative localization not only based on "visible locations" but also on "verifiable functional responses." This effectively reduces the possibility of mistakenly cutting non-target tissues or missing real sympathetic nerves, thus improving the reliability of surgical localization results.

[0021] Third, this invention fuses thoracoscopic image information with physiological functional response information to form a unified localization reliability judgment result, thereby providing surgeons with clear localization references and cutting prompts. This fusion judgment method can transform multi-source information into intuitive and interpretable intraoperative auxiliary decision-making basis, helping to reduce operational differences between different surgeons, improve the standardization and repeatability of thoracic sympathectomy, and at the same time, reduce the probability of postoperative complications to a certain extent, thereby improving the overall surgical safety and clinical application value. Attached Figure Description

[0022] Figure 1 This is a system diagram of the present invention; Figure 2 This is a diagram showing the composition of the data acquisition module of the present invention; Figure 3 This is a diagram illustrating the composition of the neural function response module of the present invention. Figure 4 This is a schematic diagram of the intraoperative field of view of the candidate region of the thoracic sympathetic nerve under thoracoscopic guidance according to the present invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this invention clearer, the following description is provided in conjunction with embodiments and appendices. Figures 1-4 The present invention will be further described in detail below. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0024] Example 1: This invention provides an intraoperative nerve localization system for thoracic sympathectomy. Applied during thoracoscopic thoracic sympathectomy, it is used to locate and confirm the thoracic sympathetic nerve chain and its corresponding segments intraoperatively, and to provide the surgeon with suggestions on the cutting location and safety prompts. The system includes: a thoracoscopy image acquisition module, a physiological signal acquisition module, a nerve candidate region identification module, a nerve stimulation and functional response detection module, a multimodal fusion localization module, a cutting strategy and safety prompt module, and a display and interaction module. These modules are connected via a data bus or control interface to form a real-time intraoperative localization closed loop.

[0025] The thoracoscopy image acquisition module is an imaging component of an existing thoracoscopy system. It includes a thoracoscope body inserted into the thoracic cavity through a puncture port in the patient's chest wall, an image sensor located at the distal end of the thoracoscope body, and an image acquisition interface module electrically connected to the image sensor. The image sensor is a CMOS or CCD image sensor used to convert optical images of intrathoracic tissues into electrical signals; the image acquisition interface module is used to perform analog-to-digital conversion on the electrical signals and output continuous digital thoracoscopy video data.

[0026] The system uses frame rate (Unit: frames / second) Acquire thoracoscopic video images, record the first... Frame image is in, Indicates the first The pixel coordinates in the frame of the thoracoscopic image are Image pixel values; Two-dimensional pixel coordinates (unit: pixels); The image frame number (dimensionless).

[0027] The pixel value is a digital signal output by the image sensor, used to characterize the optical brightness or color information of the tissue at the corresponding location. The acquisition time corresponding to a frame image is defined as in, For the first Timestamp of the frame image (in seconds). Based on the image pixel values. The neural candidate region recognition module calculates a pixel-level neural probability map for each frame of the image. ,in Represents pixels The probability value (dimensionless) of belonging to neural tissue. The neural probability is calculated by a recognition model based on image features such as pixel intensity, local texture, and morphological features.

[0028] In the In the frame image, the system performs [action] within the paravertebral region. Threshold segmentation and connected component analysis are performed to obtain several neural candidate regions, denoted as the first... The candidate regions are ,in, This represents a set of regions (pixel set) consisting of a number of pixels. For each candidate region, the system calculates its image structure evidence, i.e., the image neural score: in, The number of pixels in the candidate region (dimensionless). Image neural scoring for candidate regions (dimensionless, range of values) ).

[0029] To ensure consistency between image analysis and subsequent stimuli and physiological responses on the timeline, when the system stimulates a candidate region, the stimulation start time is recorded as follows: And based on frame rate Determine the set of image frames corresponding to the stimulus event: in, This is a dimensionless set of image frame indices corresponding to stimulus events. The length of the time window selected before and after stimulation (in seconds).

[0030] The system performs time-averaged image neural scores in the image frame set to obtain structural evidence corresponding to the stimulus event: in, Candidate region Time-consistent image scoring (dimensionless) under stimulus events.

[0031] The neural stimulation and functional response detection module includes a stimulation probe and a stimulation control unit. The stimulation probe enters the thoracic cavity through a thoracoscopic operating channel, and a stimulation electrode is positioned at its distal end to apply low-intensity electrical stimulation to candidate areas. The stimulation control unit controls the amplitude, frequency, and pulse width of the stimulation current.

[0032] The stimulation current signal is represented as: in, Stimulation current (unit: mA). The stimulus amplitude (mA) is the value of the stimulus. The stimulation frequency (Hz) Pulse width (seconds) The number of pulses (dimensionless). It is a rectangular pulse function.

[0033] During stimulation, the physiological signal acquisition module simultaneously acquires heart rate signals. (Unit: times / minute). Within the pre-stimulation window. With the post-stimulation window Calculate the mean heart rate separately: in, All values ​​are average heart rates (unit: beats / minute). The time window length (in seconds).

[0034] The stimulus-induced change in heart rate is defined as: in, The unit is beats per minute. To eliminate the influence of individual differences, the standard deviation of the baseline heart rate before stimulation was calculated: in, The unit is times per minute.

[0035] Standardizing the aforementioned change yields the dimensionless functional response: in, Standardized heart rate response (dimensionless). To prevent the division of infinitesimal constants with a denominator of zero, their units are... same.

[0036] The multimodal fusion localization module fuses image structural evidence with functional response evidence. First, it fuses structural evidence... Mapped to logarithmic odds form: in, The structure logarithmic probability (dimensionless). This is a numerically stable term (dimensionless). The fusion discriminant function is then constructed: in, The fusion discriminant value (dimensionless). For bias terms, These are weighting coefficients (all dimensionless).

[0037] The discriminant value is mapped using a Sigmoid function to obtain the confidence level that the candidate region belongs to the thoracic sympathetic nerve: in, Location reliability of the thoracic sympathetic nerve (dimensionless, 0-1).

[0038] The system selects the candidate region with the highest confidence as the localization result for the current frame: The corresponding candidate regions are then output as the thoracic sympathetic nerve localization regions.

[0039] The system pre-sets a threshold Γ for determining the location of the thoracic sympathetic nerve, where Γ is a dimensionless parameter used to distinguish between valid and invalid nerve localization. When the localization reliability of the candidate region... When the confidence level is greater than or equal to the threshold Γ, the system determines that the candidate region is the result of thoracic sympathetic nerve localization and outputs it as the intraoperative nerve localization region, and outputs the localization identifier and segment reference information to the display interaction module; when the confidence level is lower than the threshold, the system outputs a safety prompt of review or not recommending transection.

[0040] The display and interaction module overlays the positioning area, confidence level, and prompt information onto the real-time thoracoscopic image, thereby providing the surgeon with reliable intraoperative nerve positioning assistance without interfering with the surgical procedure.

[0041] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments, but the present invention is not limited to these embodiments. Equivalent modifications made by those skilled in the art without departing from the principles of the present invention should fall within the protection scope of the present invention.

[0042] Example 2: This example illustrates the application of the intraoperative nerve localization system described in this invention in a real surgical scenario, using a case of a patient with primary palmar hyperhidrosis undergoing thoracoscopic thoracic sympathectomy.

[0043] The patient is placed in a supine or semi-recumbent position. An operating channel is established via a standard thoracoscopic approach, and the thoracoscope is inserted into the thoracic cavity to obtain the intraoperative field of view. The thoracoscopic image acquisition module outputs real-time image data of the paravertebral region within the thoracic cavity, as shown in the figure. The images clearly display the course of the ribs, paravertebral soft tissue structures, and the distribution of blood vessels on the surface of the visceral pleura. The system continuously acquires thoracoscopic images at a preset frame rate and processes the image pixel values ​​in real time.

[0044] Within this surgical field of view, a longitudinally running cord-like structure can be observed in the paravertebral region. This structure is located medial to the adjacent rib head and, to the naked eye, may resemble the thoracic sympathetic nerve chain. However, it is surrounded by numerous blood vessels and connective tissue, making it difficult to confirm its identity as the target sympathetic nerve solely through visual inspection. To avoid misjudgment, the surgeon activates the intraoperative nerve localization system of this invention.

[0045] The system first performs a constrained search of the paravertebral region based on thoracoscopic images, and then performs pixel-level analysis on suspected nerve cord-like structures in the images, calculating the nerve pixel probability and forming nerve candidate regions. In this embodiment, the system identifies multiple candidate regions in the image and marks one candidate region located near the level of the third rib as a key candidate. This candidate region is labeled R3 in the image (as shown by the arrow in the figure), representing the nerve candidate structure identified by the system at the level of the third rib.

[0046] Subsequently, the system records the spatial position changes of the candidate region in consecutive image frames and performs temporal consistency analysis on the corresponding image neural scores to eliminate transient artifacts caused by respiratory traction or camera shake. After confirming that the candidate region maintains stable movement across multiple consecutive frames, the system prompts the surgeon to perform functional verification of the region.

[0047] The surgeon uses the thoracoscopic operating channel to bring the stimulation probe close to the R3 candidate region and applies low-intensity electrical stimulation under system control. During stimulation, the physiological signal acquisition module simultaneously acquires the patient's heart rate signal and records changes before and after stimulation. The system compares the baseline heart rate before stimulation with the heart rate after stimulation, calculates the change in heart rate, and performs standardization to obtain the corresponding functional response index.

[0048] In this embodiment, after the system detects stimulation of the candidate R3 region, the patient's heart rate exhibits a significant and repeatable change, and the magnitude of this change exceeds the system's preset functional response judgment threshold, indicating that the candidate structure has clear sympathetic nerve functional characteristics. The system further fuses this functional response result with the aforementioned image structural evidence to obtain the thoracic sympathetic nerve localization reliability corresponding to the candidate region.

[0049] After fusion calculation, the location reliability of candidate region R3 was the highest among all candidate regions, exceeding the system's preset threshold for thoracic sympathetic nerve location determination. Based on this, the system determined that candidate region R3 corresponds to the segment of the thoracic sympathetic nerve chain, highlighted the region in the real-time thoracoscopy image, and simultaneously output a "Location confirmed" message to the surgeon.

[0050] With the assistance of system prompts, the surgeon performed a transection of the thoracic sympathetic nerve corresponding to the R3 region. Postoperative observation revealed no significant bleeding or accidental damage to adjacent tissues. The patient's sweating symptoms were significantly relieved immediately after the surgery, and no adverse reactions such as abnormal heart rate fluctuations occurred. This embodiment demonstrates that the intraoperative nerve localization system described in this invention can accurately locate the thoracic sympathetic nerve under complex intraoperative visual conditions and provide a reliable reference for the transection operation.

[0051] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An intraoperative nerve location system for thoracic sympathectomy, characterized in that, include: The system includes a data acquisition module, a neural candidate region identification module, a neural function response detection module, a multimodal fusion localization module, a severance strategy and safety prompt module, and a human-computer interaction and display module. The data acquisition module is used to acquire intraoperative thoracoscopic image data and physiological signal data related to sympathetic nerve function during thoracic sympathectomy. The nerve candidate region identification module identifies and generates nerve candidate regions based on the intraoperative thoracoscopic image data, which are suspected to be the location of the thoracic sympathetic nerve chain. The neural function response detection module is used to stimulate the candidate neural region during surgery and collect the corresponding physiological signal changes before and after stimulation in order to obtain neural function response information. The multimodal fusion localization module is used to fuse the candidate nerve regions, nerve function response information, and intraoperative image information to output the localization results of the thoracic sympathetic nerve and the confidence information of the corresponding segment. The severance strategy and safety warning module generates thoracic sympathetic nerve severance suggestions and risk warning information based on the positioning results; The human-computer interaction and display module is used to display the positioning results, cutting suggestions and risk warning information to the surgeon in real time.

2. An intraoperative nerve location system for use in a thoracic sympathicolysis as claimed in claim 1, characterized in that: The data acquisition module includes a thoracoscopic video acquisition unit, a physiological signal acquisition unit, and a time synchronization unit, wherein the physiological signal acquisition unit is used to acquire heart rate signals, skin conductance signals, and / or local skin temperature signals.

3. An intraoperative nerve location system for use in a thoracic sympathicolysis according to claim 1, characterized in that: The neural candidate region identification module determines the possible segment range of the thoracic sympathetic nerve based on the rib sequence identification results, and performs candidate labeling of neural structures within the segment range in combination with the morphological characteristics of the neural tissue.

4. The intraoperative nerve localization system for thoracic sympathectomy according to claim 1, characterized in that: The neural function response detection module includes a stimulation control unit and a response analysis unit. The stimulation control unit is used to apply low-intensity electrical stimulation to the neural candidate region, and the response analysis unit is used to analyze the amplitude, trend and duration of changes in physiological signals collected before and after stimulation.

5. The intraoperative nerve localization system for thoracic sympathectomy according to claim 1, characterized in that: The multimodal fusion localization module comprehensively analyzes the image feature information and neural function response information of the neural candidate region to determine whether the target structure is the thoracic sympathetic nerve with sympathetic nerve function.

6. The intraoperative nerve localization system for thoracic sympathectomy according to claim 1, characterized in that: The multimodal fusion localization module is further used to identify anatomical variations of the thoracic sympathetic nerve, including accessory branches, bypass nerves, or segmental fusion structures, and the identification results are used as a reference for calculating localization reliability.

7. The intraoperative nerve localization system for thoracic sympathectomy according to claim 1, characterized in that: The cutting strategy and safety prompt module is used to output a safety prompt message to the surgeon that does not recommend cutting or adjusting the cutting position when a high-risk complication is detected in the target nerve.

8. The intraoperative nerve localization system for thoracic sympathectomy according to claim 1, characterized in that: The described cutting strategy and safety prompt module can generate differentiated cutting suggestions based on the clinical indications corresponding to different thoracic sympathetic nerve segments.

9. The intraoperative nerve localization system for thoracic sympathectomy according to claim 1, characterized in that: The human-computer interaction and display module is used to overlay and display the thoracic sympathetic nerve location markers, segment numbers, and corresponding confidence information on the real-time thoracoscopy image.

10. The intraoperative nerve localization system for thoracic sympathectomy according to claim 1, characterized in that: The system also includes a preoperative anatomical prior model import module, which provides reference constraints for intraoperative identification of neural candidate regions and multimodal fusion localization.

Images