PICC accurate navigation positioning method based on magnetic induction

Abstract

The invention discloses a PICC (peripherally inserted central catheter) accurate navigation and positioning method based on magnetic induction. The method comprises the following steps: acquiring spatial displacement data of a PICC through a sensor module based on magnetic induction; performing offset feature extraction on the spatial position data to obtain an offset; the offset is judged and corrected, and adjustment parameters of dynamic positioning are obtained; controlling the PICC catheter according to the adjustment parameters, obtaining motion trail data of the PICC catheter, carrying out abnormal contact state judgment on the motion trail data, obtaining a distance signal of the PICC catheter and a blood vessel wall in a blood vessel model after the abnormal contact state judgment, obtaining a contact risk level according to the distance signal, and determining the blood vessel model according to the contact risk level. The method comprises the following steps: acquiring motion change characteristics of motion trail data, performing path adjustment on a PICC catheter according to the motion change characteristics to obtain adjusted motion trail data, and performing calibration through the adjusted motion trail data to obtain an accurate positioning result.

Description

Technical Field

[0001] This invention belongs to the field of PICC catheter navigation and positioning technology, and particularly relates to a precise navigation and positioning method for PICC catheters based on magnetic induction. Background Technology

[0002] Precise positioning of PICC catheters in complex vascular environments remains a significant challenge. Current navigation methods largely rely on ultrasound or X-ray imaging, which, while assisting catheter placement to some extent, have limitations in real-time performance and accuracy. Ultrasound imaging is constrained by operator experience and equipment resolution, often failing to clearly discern the subtle positional relationship between the catheter tip and the vessel wall; X-ray imaging involves radiation risks and cannot continuously monitor catheter dynamics, making it difficult to address changes in the complex intravascular environment. These limitations frequently lead to catheter deviation or misplacement into non-target vessels, increasing the risk of complications.

[0003] In PICC catheter navigation, the core challenge lies in achieving dynamic and precise positioning of the catheter tip in three-dimensional space. The catheter's movement within the blood vessel is influenced by factors such as blood flow, vascular curvature, and tissue resistance, making it prone to deviating from the preset path. Without real-time detection and quantification of this positional shift, accurate catheter orientation adjustment is difficult. Positional shift detection and contact status monitoring are complementary; the former determines the correct catheter path, while the latter ensures operational safety. Together, they constitute the core technological bottleneck of the navigation system.

[0004] Therefore, how to monitor the positional deviation of the catheter tip in real time and accurately identify its abnormal contact with tissue has become a key issue in improving the accuracy and safety of PICC catheter navigation. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention proposes a precise navigation and positioning method for PICC catheters based on magnetic induction, thereby resolving the issues present in the prior art.

[0006] To achieve the above objectives, the present invention provides a method for precise navigation and positioning of PICC catheters based on magnetic induction, comprising:

[0007] Spatial displacement data of the PICC catheter is acquired using a magnetic induction-based sensor module. Offset features are extracted from the spatial position data to obtain the offset. The offset is then judged and corrected to obtain dynamic positioning adjustment parameters. The PICC catheter is controlled according to the adjustment parameters to acquire its motion trajectory data. Abnormal contact status is assessed based on the motion trajectory data. After the abnormal contact status is determined, the distance signal between the PICC catheter and the vessel wall in the vascular model is acquired. The contact risk level is obtained based on the distance signal. Based on the contact risk level, the motion change characteristics of the motion trajectory data are acquired. The path of the PICC catheter is adjusted based on the motion change characteristics to obtain the adjusted motion trajectory data. The adjusted motion trajectory data is then calibrated to obtain an accurate positioning result.

[0008] Optionally, the process of acquiring spatial displacement data includes:

[0009] By analyzing the changes in the external magnetic field and the electromagnetic induction sensor installed at the end of the PICC conduit, the position of the PICC conduit end at different times can be obtained, and spatial displacement data can be obtained.

[0010] Optionally, the offset acquisition process includes:

[0011] The position offset is calculated by calculating the corresponding difference between the spatial displacement data and the preset path. The position offset is then decomposed to obtain preliminary offset features. Abnormal fluctuations in the preliminary offset features are decomposed using feature analysis technology to obtain a set of interference signals. The set of interference signals is judged and filtered. Based on the filtered interference signals and the position offset data, the actual position offset of the catheter tip is corrected and calculated to obtain the offset.

[0012] Optionally, the process for obtaining adjustment parameters for dynamic positioning includes:

[0013] The offset is initially processed to obtain its magnitude and direction. Based on the magnitude and direction of the offset, it is determined whether path adjustment is needed. If path adjustment is needed, the adjustment parameters for dynamic positioning are obtained based on the mapping relationship between the magnitude and direction of the offset and the adjustment parameters.

[0014] Optionally, the process of determining abnormal contact status in motion trajectory data includes:

[0015] The polarity of the motion trajectory data is processed to obtain positional offset features. Based on the positional offset features and the distribution characteristics of bending resistance, the force on the catheter tip in the vascular environment is obtained. If the force distribution exceeds the preset normal range, the motion trajectory data is calibrated a second time to obtain calibrated trajectory change information. Based on the calibrated trajectory change information and the judgment criteria of critical state, the support vector machine algorithm is used to classify and predict whether the catheter tip is close to abnormal contact, and the classification result of abnormal contact is obtained.

[0016] Optionally, the process for obtaining the exposure risk level includes:

[0017] A geometric model of the blood vessel is constructed, and the geometric model and motion trajectory data are unified in a coordinate system. The distance data between the motion trajectory data and the blood vessel wall in the geometric model is obtained. Based on the distance data, a machine learning model is used to identify and obtain the contact risk level.

[0018] Optionally, the process of acquiring the adjusted motion trajectory data includes:

[0019] When the risk level is high, corresponding motion rate change features are extracted from the motion trajectory data. A support vector machine (SVM) algorithm is used to classify and analyze these features, determining the degree of matching between the features and the risk level. Based on this matching, it is determined whether the catheter tip has deviated from the expected path. Reference data for path adjustment is extracted based on this deviation. The degree of path deviation is then quantified using this reference data to obtain a preliminary adjustment plan. This preliminary adjustment plan, combined with real-time monitored motion rate changes and blood flow dynamics data, is optimized using a logistic regression algorithm to determine the final path trajectory data.

[0020] On the other hand, the present invention provides a PICC catheter precision navigation and positioning system based on magnetic induction for performing the above-described method.

[0021] Compared with the prior art, the present invention has the following advantages and technical effects:

[0022] This invention discloses a method for precise positioning of catheter tips in a vascular environment. Through multi-dimensional motion signal acquisition and processing, it achieves real-time monitoring and adjustment of the catheter tip position. First, the invention acquires spatial displacement data within the blood vessel, extracts positional offset features, separates abnormal fluctuation signals, and determines the real-time offset. When the offset exceeds a threshold, a correction command is generated and adjustment parameters are calculated. Combined with resistance characteristics, it determines whether the catheter is approaching an abnormal contact state and performs a contact risk assessment. Based on the risk level and motion change characteristics, it determines whether further adjustment is needed. Finally, a precise navigation algorithm is used to continuously calibrate the catheter tip position, obtaining a stable positioning result and continuously monitoring potential risks. This invention can effectively improve the positioning accuracy and safety of catheters in complex vascular environments, providing reliable navigation support for minimally invasive surgery. Attached Figure Description

[0023] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0024] Figure 1 This is a flowchart of a PICC catheter precision navigation and positioning method based on magnetic induction according to an embodiment of the present invention;

[0025] Figure 2 This is a schematic diagram of the PICC catheter precision navigation and positioning system based on magnetic induction, according to an embodiment of the present invention. Detailed Implementation

[0026] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0027] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0028] like Figure 1 As shown, this embodiment of a PICC catheter precision navigation and positioning method and system based on magnetic induction may specifically include:

[0029] S101. The sensor data acquisition module acquires multidimensional motion signals within the vascular environment from the catheter tip to obtain spatial displacement data.

[0030] The magnetic induction sensor data acquisition module acquires multidimensional motion signals within the vascular environment from the catheter tip, obtaining raw spatial displacement information. Based on the raw spatial displacement information, filtering technology is used to denoise the acquired data, thus obtaining the final spatial displacement information.

[0031] For example, by embedding a miniature electromagnetic induction sensor at the tip of the PICC catheter, and using an electromagnetic generator placed under or near the patient to generate a low-intensity, low-frequency electromagnetic field, a signal from the sensor catheter is received by an electromagnetic receiver / tracker integrated into the main unit. By analyzing the signal from the sensor, the aforementioned raw spatial displacement information is generated. In this process, the catheter tip transmits these weak induced voltage signals to the electromagnetic receiver / tracker in the main unit via wires or wirelessly. The receiver amplifies the weak raw signal and uses a filter to remove electromagnetic noise interference from the environment, converting the amplified and filtered analog voltage signal into a digital signal. Initially, the distribution characteristics of the electromagnetic field generated by the electromagnetic generator in three-dimensional space are precisely calibrated, including: the magnetic field strength at each point in space, the magnetic field gradient at each point in space, and the attenuation of the magnetic field with distance. These are the parameters of the initial electromagnetic field without the influence of the aforementioned sensor. For the sensor placed at the catheter tip, the geometry, orientation, and sensitivity of its sensor coil, and the sensor's response function to magnetic fields of different directions and intensities (i.e., how much induced voltage it generates under a specific magnetic field) are recorded.

[0032] The induced voltage value is acquired and digitized from the sensor in real time. This real-time measured sensor induced voltage value is then substituted into a complex mathematical model. The core of this model is the electromagnetic field theory equations (such as the mathematical forms of the Biot-Savart law and Faraday's law), which describe the theoretically induced voltage that the sensor should produce at a specific location and orientation in a known magnetic field distribution.

[0033] The goal is to find a unique position (X, Y, Z) that minimizes the error between the "theoretical induced voltage" calculated based on the known emission field model and sensor response model and the actual measured induced voltage value. Alternatively, given a fixed electromagnetic field generated by the electromagnetic generator and fixed parameters of the electromagnetic sensor, the position of the electromagnetic induction sensor will produce a corresponding induced voltage. This principle is used to obtain the position at different times, yielding the original spatial displacement information. This information may manifest as changes in three-dimensional coordinates, such as the displacement of the catheter tip at a certain moment being 0.5 mm along the x-axis, 0.2 mm along the y-axis, and 0.1 mm along the z-axis, but this is mixed with noise from blood flow interference.

[0034] For example, filtering techniques can be used to improve data quality when denoising raw spatial displacement information.

[0035] One possible implementation involves using a low-pass filter to remove high-frequency noise, as blood flow interference often manifests as high-frequency vibrations, while the actual displacement changes of the catheter are typically low-frequency signals. This method removes irrelevant interference signals from the acquired raw data, yielding information closer to the true spatial displacement. For example, after filtering, the displacement data of the catheter tip might be adjusted to 0.48 mm on the x-axis, 0.19 mm on the y-axis, and 0.09 mm on the z-axis, significantly reducing data fluctuations and resulting in smoother data. This denoising operation improves the accuracy of subsequent analysis and avoids misjudgments caused by noise.

[0036] For example, in practical applications, the acquisition and processing of multidimensional motion signals can be optimized from both hardware and algorithm perspectives. On the hardware side, higher-sensitivity miniature sensors can be integrated into the catheter tip to capture finer signals unaffected by changes in blood flow and resistance. On the algorithm side, in addition to low-pass filtering, adaptive filtering technology can be introduced to dynamically adjust filtering parameters based on real-time changes in the vascular environment. This multi-dimensional optimization approach supports each other, ensuring comprehensive data acquisition and reliable noise reduction. Ultimately, these technologies bring significant benefits: on the one hand, doctors can determine the catheter's position and status within the blood vessel based on more accurate spatial displacement information; on the other hand, they reduce the risk of misjudgment, improving the safety and success rate of surgical procedures, especially in precise navigation in complex vascular environments.

[0037] S102. Based on the spatial displacement data, a signal processing algorithm is used to extract features of the positional offset, separate the abnormal fluctuation signal caused by the influence of blood flow, and determine the real-time offset of the catheter tip.

[0038] Based on spatial displacement data, signal processing methods are used to initially decompose the positional offset, separating signal components that may contain abnormal fluctuations and determining preliminary offset characteristics. For these preliminary offset characteristics, feature analysis techniques are used to further identify abnormal fluctuations in the signal, extracting specific patterns related to blood flow interference, resulting in a set of interference signals. If significant fluctuation patterns exist in the interference signal set, a preset threshold is used for comparison to determine which signal components are abnormalities caused by blood flow interference, resulting in filtered interference signals. Based on the filtered interference signals and real-time offset data, the actual positional offset of the catheter tip is corrected and the corrected offset is determined.

[0039] For example, when processing spatial displacement data, signal processing methods can be used to initially decompose the positional offset of the catheter tip. The core of signal processing lies in breaking down complex displacement data into components of multiple frequencies or amplitudes to identify potentially hidden abnormal fluctuations. Assuming that displacement data acquired in a vascular environment contains both periodic and aperiodic fluctuations, signal decomposition can classify periodic fluctuations as signals caused by normal blood flow, while aperiodic fluctuations may be related to abnormal interference. This decomposition method helps to initially identify the signal components of interest, laying the foundation for subsequent analysis.

[0040] For example, feature analysis techniques for initial displacement characteristics can employ a combination of time and frequency domain methods for in-depth identification. Time domain analysis focuses on the characteristics of signal changes over time; for instance, a peak displacement data point of 0.5 mm within a certain time period may indicate the instantaneous impact of blood flow interference. Frequency domain analysis, on the other hand, focuses on the frequency distribution of the signal; if an abnormal concentration of energy is found in a certain frequency band, it may be related to a specific blood flow pattern. Through this dual analysis, specific patterns related to blood flow interference can be extracted, forming a set of interference signals. The advantage of this approach is that it can characterize signal features from multiple dimensions, improving the comprehensiveness of identification.

[0041] For example, when performing threshold comparison on a set of interfering signals, a threshold for fluctuation amplitude can be preset, such as 0.3 mm. Signals exceeding this threshold are considered abnormalities caused by blood flow interference. Suppose that in a sampling process, a set of signals with a fluctuation amplitude of 0.4 mm is detected, significantly higher than the threshold, and is then classified as an abnormal signal. This threshold comparison method is simple and intuitive, and can quickly filter out signal components requiring special attention, thus providing a reliable basis for subsequent correction. Such a filtering mechanism helps reduce the workload of data processing and improves efficiency.

[0042] For example, when performing correction calculations based on filtered interference signals, real-time offset data can be combined to adjust the actual positional offset of the catheter tip using a weighted average. Assuming the original displacement data is 1.2 mm and the influence of the interference signal is 0.3 mm, the corrected offset might be 0.9 mm after correction calculation. The advantage of this correction method is its ability to dynamically adapt to changes in the vascular environment, ensuring the accuracy of catheter position data and providing a more reliable reference for subsequent operations. Through the above analysis and examples, it can be seen that each step from signal decomposition to correction calculation is closely related to the processing of displacement data in the vascular environment, with each link supporting the others to form a complete processing chain. This method not only improves the precision of data processing but also provides crucial support for the stable operation of catheters in complex environments.

[0043] S103. If the real-time offset exceeds the preset threshold range, a correction instruction is generated through the path adjustment module, and the adjustment parameters for dynamic positioning are calculated based on the magnitude and direction of the offset.

[0044] The offset data undergoes preliminary processing to determine its magnitude and direction. Based on the real-time offset magnitude and direction, a comparison is made with preset thresholds. If the offset exceeds the preset threshold range, the path adjustment module is triggered to determine the type of instruction requiring correction. The path adjustment module analyzes the offset direction and magnitude and, using a pre-established mapping relationship, calculates the adjustment parameters required for dynamic positioning.

[0045] For example, when processing offset data, a preliminary analysis can be performed on the collected spatial displacement information to extract the magnitude and direction of the offset. This process typically involves reading and parsing sensor data to determine the actual deviation of the catheter tip in space. Suppose that in a certain operation, the system detects an offset of 2.5 mm at the catheter tip, deviating to the right. In this case, the preliminary processing will transform this data into structured information suitable for subsequent analysis, providing a basis for further judgment.

[0046] For example, a reasonable range can be set for comparing the magnitude and direction of the offset with a preset threshold, such as an offset threshold of 1.5 mm. If the detected offset of 2.5 mm exceeds this range, the system will automatically determine it as abnormal and trigger the path adjustment module. The core of this comparison mechanism is to ensure the stability of the catheter position and avoid affecting the operational accuracy due to excessive offset.

[0047] It should be noted that the threshold setting can be dynamically adjusted according to the actual application scenario. For example, under high precision requirements, the threshold can be lowered to 1.0 mm to improve the control sensitivity.

[0048] For example, during the path adjustment module's analysis, the system calculates the adjustment parameters required for dynamic positioning based on the offset direction and magnitude, combined with a pre-established mapping relationship. Assuming the offset direction is 2.5 mm to the right, the mapping relationship might indicate the need to apply a corresponding reverse adjustment to the left. Specific parameters might be empirical values ​​derived from training with historical data, such as an adjustment angle of 3 degrees and a moderate force. Establishing such mapping relationships typically relies on the accumulation of extensive experimental data to ensure a high degree of match between the adjustment parameters and the actual offset.

[0049] It should be noted that this method can quickly respond to offset changes in real-time operation, reducing the need for human intervention.

[0050] For example, when determining the type of instruction requiring correction, the path adjustment module might generate a specific set of control instructions, such as adjusting the guide tube's advance angle or changing its movement speed. Assuming a 2.5 mm offset to the right as described above, the system would generate an angle correction instruction, specifically a slight leftward adjustment of 3 degrees, while simultaneously reducing the advance speed to 80% of its original speed to ensure a smooth adjustment process. This type of instruction is typically chosen based on a comprehensive analysis of the magnitude and direction of the offset, aiming to achieve position restoration with minimal adjustment.

[0051] It should be noted that this refined instruction design can effectively avoid over-adjustment or under-adjustment, ensuring that the catheter tip can quickly return to the expected path.

[0052] For example, in the entire processing flow, the calculation of dynamic positioning parameters and the issuance of commands form a closed loop. The system continuously monitors the adjusted offset data and further optimizes the parameters based on real-time feedback. Suppose that the initial offset decreases to 0.8 mm, but still does not reach the ideal state, the system may fine-tune the angle or speed again until the offset fully meets the threshold requirements. The introduction of this closed-loop mechanism significantly improves the reliability of catheter positioning, especially when operating in complex environments, effectively responding to sudden changes.

[0053] S104. Based on the adjusted parameters, obtain the latest motion trajectory data of the catheter tip within the vascular environment, and combine it with the distribution characteristics of bending resistance to determine whether the catheter is approaching an abnormal contact state.

[0054] By adjusting parameters, the latest motion trajectory data of the catheter tip is extracted from the vascular environment. Preliminary processing of trajectory changes yields the positional offset characteristics of the catheter tip. Based on these offset characteristics and the distribution characteristics of bending resistance, the force distribution on the catheter tip in the vascular environment is analyzed to determine if it conforms to a preset normal range. If the force distribution exceeds the preset normal range, the motion trajectory of the catheter tip is recalibrated using environmental impact data to obtain the calibrated trajectory change information. Based on the calibrated trajectory change information and the criteria for judging critical states, a support vector machine algorithm is used to classify and predict whether the catheter tip is approaching abnormal contact, resulting in a classified abnormal contact outcome.

[0055] The vascular environment can be acquired through prior medical imaging, which characterizes the distribution and specific data of blood vessels. This is achieved by unifying the geometric model and image of the blood vessel distribution with the electromagnetically induced position in the same coordinate system, and then incorporating the aforementioned positional data into the vascular environment.

[0056] Specifically, during the extraction of catheter tip motion trajectory data, high-precision sensors can be used to collect the catheter's position coordinates within the vascular environment in real time. Assuming the catheter tip's coordinates at a certain moment are x = 2.5 mm, y = 1.8 mm, z = 0.3 mm, compared to the previous moment, the offsets are Δx = 0.2 mm, Δy = 0.1 mm, and Δz = 0.05 mm. Through preliminary processing of these offset data, the positional offset characteristics of the catheter tip are obtained, such as the offset direction deviating to the left of the vessel wall, and the offset amplitude being 0.23 mm. This feature extraction helps in subsequent analysis of whether the catheter is experiencing abnormal resistance.

[0057] In one possible implementation, the distribution characteristics of bending resistance can be analyzed by combining the geometric model of the vessel wall and the material properties of the catheter to assess the force on the catheter tip in different vessel segments. Assuming a vessel inner diameter of 3.0 mm and a catheter diameter of 1.5 mm, the resistance experienced in the bending segment is 0.15 N, significantly higher than the 0.08 N in the straight segment. By comparing these force data with the preset normal range of 0.05-0.12 N, it is determined that the current force distribution exceeds the normal range, suggesting a potential risk of abnormal contact between the catheter and the vessel wall. The resistance can be calculated by measuring the relevant frictional force using a force sensor worn by the physician or a force sensor fixed near the input position that is not affected by external forces. The frictional force is then converted into the propulsive force and the corresponding velocity of the motion trajectory.

[0058] Specifically, if the force distribution is abnormal, the motion trajectory can be recalibrated using environmental impact data.

[0059] For example, using the elastic modulus of the blood vessel wall (0.02 N / mm) and the blood flow velocity (0.5 m / s), the lateral force affecting the catheter is calculated, and the trajectory coordinates are adjusted to reduce the calibrated offset Δx to 0.15 mm. This calibration can more accurately reflect the actual movement of the catheter in complex vascular environments.

[0060] For example, based on the calibrated trajectory change information and the critical state judgment criteria, a support vector machine algorithm can be used to predict abnormal contact. Assuming the critical state is defined as the distance between the catheter tip and the vessel wall being less than 0.1 mm, by inputting the calibrated trajectory data Δx = 0.15 mm, Δy = 0.09 mm, and a force of 0.15 N, the algorithm outputs a classification result of "approaching abnormal contact".

[0061] In one possible implementation, the training data for the support vector machine algorithm can be derived from historical surgical records.

[0062] For example, by selecting 1000 sets of catheter motion data, with 300 sets labeled as abnormal contact and 700 sets as normal motion, a model can be trained using feature vectors to achieve a prediction accuracy of over 90%. This high-precision prediction can effectively assist catheter navigation and ensure surgical safety.

[0063] It should be noted that the above-mentioned steps are closely interconnected, forming a complete catheter motion monitoring system from trajectory extraction to anomaly prediction. Each step is implemented around the dynamic behavior of the catheter within the vascular environment, ensuring the reliability of the analysis results and significantly improving operational accuracy.

[0064] S105. If it is determined that the catheter is approaching an abnormal contact state, the distance signal between the catheter tip and the blood vessel wall is extracted through the contact warning module, and the risk of tissue damage is quantitatively assessed to obtain the contact risk level.

[0065] The contact early warning module collects real-time distance data between the catheter tip and the vessel wall in the geometric model to initially monitor abnormal contact states and obtain an initial distance signal. Based on this initial distance signal, a preset threshold range is used for comparison. If the distance signal is lower than a preset lower threshold, it is determined to be a high-risk abnormal contact state, and a high-risk warning sign is generated. For the high-risk warning sign, the positional change data of the catheter tip within the current time period is acquired. Combined with the vessel wall distance information, a support vector machine algorithm is used to classify and predict the probability of tissue damage and determine the potential damage level.

[0066] For example, a contact warning module can collect real-time distance data between the catheter tip and the vessel wall. This data is then used to calculate the distance based on the aforementioned geometric model and positional data within the same coordinate system. In one possible implementation, when comparing the initial distance signal with a preset threshold range, a lower limit for the safe distance can be set at 0.5 mm, and a warning distance at 0.3 mm. If the signal indicates a distance of 0.4 mm, the system determines this as an abnormal contact and triggers a warning. The comparison process must consider the elasticity of the vessel wall and the rigidity of the catheter material.

[0067] For example, polyurethane catheters may undergo minute deformation at a distance of 0.3 mm, increasing the risk of contact. Threshold settings are based on clinical data, such as the statistical distribution of vessel wall thickness from 0.2 to 0.8 mm, to ensure accurate assessment.

[0068] Specifically, after a high-risk warning sign is generated, data on the positional changes of the catheter tip needs to be extracted. Positional changes can be detected using an electromagnetic positioning system with an accuracy of 0.05 mm, recording the catheter's three-dimensional coordinate changes over 5 seconds.

[0069] For example, a displacement of 0.2 mm on the X-axis and 0.1 mm on the Y-axis at the catheter tip indicates possible obstruction due to vascular curvature. Combining this with vessel wall distance information, such as a local vessel diameter of 2 mm, it can be inferred whether the catheter is close to the vessel wall. Electromagnetic positioning data and ultrasonic ranging data complement each other, forming complete spatial information.

[0070] For example, when using the support vector machine algorithm to predict the probability of tissue damage, the input features include a distance of 0.3 mm, a displacement velocity of 0.1 mm / s, and a vessel wall stiffness index of 0.6. The algorithm is trained based on historical cases and classifies damage levels as low, medium, and high.

[0071] For example, at a distance of 0.2 mm and a speed of 0.15 mm / s, a moderate risk of injury is predicted, prompting the operator to slow down the advance speed. Support vector machines use nonlinear kernel functions to process complex feature relationships, ensuring reliable prediction results. The prediction process considers the biomechanical properties of the blood vessel wall, such as differences in elastic modulus, to avoid misjudgments.

[0072] In one embodiment, after determining the potential damage level, the system can automatically generate operational suggestions.

[0073] For example, in cases of moderate risk of injury, it is recommended to reduce the catheter advancement force to 0.5 Newtons or adjust the catheter angle by 10 degrees to avoid high-risk areas. It is advisable to combine real-time data with historical experience to reduce human error. The entire process optimizes operational safety through closed-loop feedback.

[0074] Understandably, the above scheme forms a complete chain from data collection to prediction, with each link verifying the others.

[0075] For example, the model constructed from medical images is cross-calibrated with electromagnetic positioning data, and threshold comparison and algorithm prediction complement each other. This multi-faceted support approach ensures consistent and reliable monitoring results, making it suitable for catheter operations in complex vascular environments.

[0076] S106. Based on the exposure risk level, obtain the motion change characteristics in the real-time monitoring data, and combine them with the dynamic fluctuations of blood flow to determine whether further path adjustment is needed at the catheter tip.

[0077] When the risk level is high, corresponding monitoring feature data is extracted based on the movement trajectory. A support vector machine (SVM) algorithm is used to classify and analyze these features, determining the degree of matching between the features and the risk level. By combining the matching degree with blood flow dynamics and correlation data, the impact of these fluctuations on movement changes is analyzed to determine if the fluctuation correlation exceeds the safe range. If the fluctuation correlation exceeds the safe range, real-time position data of the catheter tip is acquired. Combined with the analysis results of dynamic influence, it is determined whether the catheter tip has deviated from the expected path. Based on the results of whether the catheter tip has deviated from the expected path, reference data for path adjustment is extracted. The degree of path deviation is quantified to obtain a preliminary adjustment direction plan. Using this preliminary adjustment direction plan, combined with real-time monitored changes in movement rate and blood flow dynamics data, a logistic regression algorithm is used to optimize the adjustment plan, determining the final path trajectory data.

[0078] When acquiring the displacement change rate in a real-time monitoring system, sensors continuously record the fluctuations in the catheter's displacement rate within the blood vessel. If, within a certain time period, the monitored displacement change rate suddenly increases or changes from the normal range, it indicates that the change may be due to a change in the blood environment, causing a change in the displacement rate under a given thrust. In this case, the system will initially classify the displacement change rate change as abnormal. For this anomaly, the system will extract relevant monitoring feature data, such as the frequency and amplitude of the displacement change rate, thus providing a basis for subsequent analysis.

[0079] For example, for classification analysis of feature data, the support vector machine algorithm can be used to process the feature of displacement change rate.

[0080] In one possible implementation, the system compares the feature data with a historical risk level database. If a combination of features matches 80% of the data, it indicates a high correlation with a high-risk level. This matching result provides crucial information for subsequent risk assessment, helping to more accurately identify potential problems.

[0081] For example, when the risk level is high, analyzing the impact of dynamic fluctuations in blood flow on the rate of displacement change can be done by combining real-time monitoring data to observe whether changes in blood flow velocity are synchronized with the rate of displacement change. For instance, if the blood flow velocity decreases from 5 units / second to 3 units / second within a certain period, while the rate of displacement change increases simultaneously, the system will determine that the correlation between the fluctuations may exceed safe limits. This analytical approach can help identify potential causes of abnormal states in a timely manner, improving the comprehensiveness of monitoring.

[0082] For example, if the fluctuation correlation exceeds the safe range, the system will further acquire real-time position data of the catheter tip. Assuming the catheter tip deviates from the expected path by 2 mm within the blood vessel, the system will confirm the degree of deviation by combining blood flow dynamics analysis. This method helps to quickly pinpoint the root cause of the problem and provides a reliable reference for subsequent adjustments.

[0083] For example, when extracting reference data for path adjustment, the deviation can be quantified. Assuming a deviation angle of 5 degrees, the system will initially generate a plan to adjust 3 degrees to the left. This initial plan can provide direction for subsequent optimization, ensuring a more efficient adjustment process.

[0084] For example, based on the initial plan for adjusting the direction, the system will optimize it using a logistic regression algorithm, taking into account real-time displacement change rate and blood flow dynamic data. Assuming the optimized path trajectory is adjusted to 2.5 degrees to the left, the system will generate the final path trajectory data. This optimization method can improve the accuracy of path adjustment and reduce uncertainties during the operation.

[0085] For example, the close coordination of each stage and the comprehensive application of data analysis throughout the process can effectively improve the reliability of monitoring and adjustment. By analyzing changes in motion rate and path optimization step by step, the system can minimize the impact of deviations while ensuring safety. This approach not only improves operational stability but also provides valuable practical experience for related fields.

[0086] S107. Using the motion trajectory data after path adjustment, a precise navigation algorithm is used to continuously calibrate the three-dimensional spatial position of the catheter tip to obtain the final stable positioning result.

[0087] For the path trajectory data, a data smoothing algorithm is used to process noise and obtain smooth trajectory data. A Kalman filter algorithm is then used to fuse the smoothed trajectory data with the navigation model to calibrate the 3D position of the catheter tip, obtaining a preliminary positioning result. If the deviation between the preliminary positioning result and a preset spatial coordinate threshold exceeds a set range, an iterative optimization algorithm is used to adjust the path parameters to obtain an optimized positioning result. Based on the optimized positioning result, the real-time error between the catheter tip and the target position is calculated, and the navigation model parameters are updated to obtain an updated navigation model. Using the updated navigation model, the 3D position of the catheter tip is continuously predicted, generating real-time trajectory data to obtain a dynamic positioning result. If the stability of the dynamic positioning result is lower than a preset threshold, the data smoothing algorithm parameters are adjusted, and the motion trajectory data is reprocessed to obtain a stable positioning result.

[0088] For example, when dealing with noise in path trajectory data, data smoothing algorithms often employ moving averages or Gaussian smoothing. Moving averages reduce the impact of random noise by calculating the average value within a fixed window.

[0089] For example, in catheter navigation scenarios, the trajectory data of the catheter tip within the blood vessel may be noisy due to blood flow interference or sensor jitter. If the coordinates of the acquired trajectory points fluctuate significantly within a short period, such as a point's coordinates jumping from (2.0, 3.0, 1.0) to (2.5, 3.2, 1.3) within 0.1 seconds, a moving average method with a window size of 5 can smooth the point to a more stable value. This method is simple and efficient, preserves the main trend of the trajectory, and is suitable for scenarios with high real-time requirements.

[0090] In one possible implementation, the Kalman filter algorithm is used to fuse smooth trajectory data with the navigation model. The Kalman filter, through prediction and update steps, integrates historical trajectory data with sensor data to calibrate the three-dimensional position of the catheter tip.

[0091] For example, the navigation model predicts the catheter tip should be located at (5.0, 4.0, 2.0) at a certain moment, while smoothed sensor data indicates a location of (5.2, 4.1, 2.1). Kalman filtering combines the uncertainties of both to output a preliminary positioning result closer to the true value. This method can effectively reduce positioning errors, especially at complex vascular branches, where positioning accuracy is crucial for subsequent navigation.

[0092] Specifically, if the initial positioning result deviates too much from the preset spatial coordinate threshold, such as exceeding the threshold of 0.5 mm, the iterative optimization algorithm can adjust the path parameters using the gradient descent method.

[0093] For example, if the actual position of the catheter tip deviates from the target point (10.0, 8.0, 3.0) by 0.7 mm, the optimization algorithm, through multiple iterations, adjusts the catheter advancement angle and speed to generate an optimized positioning result that is closer to the target. This method can dynamically adapt to changes in the geometry of the blood vessel, ensuring that the catheter advances along the expected path.

[0094] For example, real-time error calculation and navigation model parameter updates can be achieved through an error feedback mechanism. Assuming the real-time error between the catheter tip and the target position is 0.3 mm, the navigation model adjusts its prediction parameters based on this error, such as reducing the weight of predicted velocity. This update mechanism allows the model to more accurately reflect the actual movement of the catheter within the blood vessel, thereby improving the reliability of navigation.

[0095] In one possible implementation, the stability of the dynamic positioning results can be assessed by calculating the variance of the trajectory points. If the variance exceeds a preset threshold of 0.2, it indicates significant trajectory fluctuations, requiring adjustment of the data smoothing algorithm parameters, such as increasing the moving average window to 7. This adjustment further smooths the trajectory, reducing positioning instability caused by blood flow or device vibration, ultimately generating stable positioning results. This method is particularly effective in complex vascular environments, ensuring the continuity and accuracy of the catheter tip trajectory.

[0096] It should be noted that the implementation of each step is closely aligned with the core requirements of catheter navigation, with each step optimized in turn, from noise reduction to final stable positioning, ensuring precise movement of the catheter tip within the blood vessel. This step-by-step approach provides reliable navigation support in dynamic environments.

[0097] S108. Based on the final stable positioning result, obtain long-term motion trend data of the catheter in the vascular environment, continuously monitor the potential risks of positional deviation and abnormal contact, and judge the operational stability of the navigation system.

[0098] By acquiring stable positioning data of the catheter within the vascular environment from the navigation system, a dynamic record of its long-term movement trend is constructed, yielding preliminary movement trajectory information. Based on this trajectory information, the catheter's positional deviation within the vascular environment is analyzed, and time-series analysis is used to determine the frequency and amplitude characteristics of the deviation. Considering the frequency and amplitude characteristics of the deviation, and in conjunction with the dynamic changes in the vascular environment, the presence of abnormal contact phenomena is detected, and the location and intensity distribution of contact points are determined. If abnormal contact phenomena are detected, the contact intensity is compared against a preset threshold to obtain the potential risk level of the abnormal contact and determine the risk warning level. Based on the risk warning level, the operational stability of the navigation system is evaluated, and a support vector machine model is used to classify stability indicators, obtaining a stability assessment result. Based on the stability assessment result, the monitoring parameters for the catheter's movement trend are adjusted, and long-term data records are continuously updated to determine if parameter optimization is necessary. If parameter optimization is confirmed, the positional deviation and abnormal contact data are re-analyzed based on the optimized parameters to obtain updated risk assessment results.

[0099] For example, when acquiring stable positioning data of the catheter within the vascular environment, the navigation system can utilize a high-precision, magnetically-based sensor to collect real-time positional information of the catheter in three-dimensional space. Assuming the system collects data once per second during a 30-minute procedure, a dataset of thousands of location points is generated. These data points reflect the catheter's minute movements and posture changes within the blood vessel, providing a foundation for subsequently constructing long-term motion trends.

[0100] It should be noted that the stability of the data directly affects the accuracy of subsequent analysis, so it is necessary to ensure that the sensor signals are not affected by external interference during the data acquisition process.

[0101] For example, when constructing a dynamic record of long-term movement trends, the collected location data can be arranged chronologically to form a continuous movement trajectory. Assuming the average speed of the catheter within the blood vessel is 2 mm / s, but decreases to 0.5 mm / s in some narrow areas, this trend can be visualized using visualization tools, intuitively reflecting the catheter's movement characteristics in different vascular segments. This recording method helps identify potential positional deviations.

[0102] For example, time-series analysis can be used to analyze catheter displacement and extract periodic characteristics of catheter movement. Suppose the analysis reveals a significant catheter displacement of approximately 3 millimeters every 5 minutes within a certain blood vessel segment. This frequency and amplitude characteristic may be related to the pulsation of the blood vessel or the elasticity of the catheter material. Such analysis can provide data support for subsequent anomaly detection.

[0103] For example, when detecting abnormal contact phenomena, the dynamic changes in the vascular environment can be considered, focusing on the contact points between the catheter and the vessel wall. Suppose that in a certain test, the catheter is found to frequently contact the vessel wall at a bend, with the contact intensity reaching 1.5 times a preset threshold; the system will then mark this area as a high-risk point. This detection method can promptly identify potential problem areas.

[0104] For example, when assessing the potential risk level of abnormal exposure, the exposure intensity can be compared using a preset threshold. Assuming the exposure intensity threshold is 10 units and the actual detected value is 15 units, the system will classify the risk level as a medium alert. This grading method helps healthcare professionals quickly determine whether intervention is necessary.

[0105] For example, to assess the stability of a navigation system, a support vector machine (SVM) model can be used to classify stability indicators. Assuming that during system operation, stability indicators remain within the normal range for more than 80% of the time, but occasionally fluctuate below 60%, the model will classify this as a state requiring attention. This classification provides a basis for subsequent parameter adjustments.

[0106] For example, when adjusting catheter motion trend monitoring parameters, the data acquisition frequency can be increased or the sensitivity of offset detection can be adjusted based on stability assessment results. Assuming the original acquisition frequency is 1 time / second, increasing it to 2 times / second after adjustment allows the system to capture subtle changes in the catheter more promptly. This adjustment method helps improve the accuracy of monitoring.

[0107] For example, in a reanalysis after parameter optimization, a new risk assessment report can be generated based on the updated location offset and abnormal contact data. Assuming that after optimization, the offset frequency decreases from once every 5 minutes to once every 10 minutes, and the risk level also decreases accordingly, this feedback mechanism can continuously improve the system's performance.

[0108] like Figure 2 As shown, this invention provides a PICC catheter precision navigation and positioning system based on magnetic induction, mainly comprising:

[0109] The data acquisition module is used to acquire multidimensional motion signals within the vascular environment from the catheter tip through the sensor data acquisition module, continuously sample the effects of blood flow and bending resistance, and obtain spatial displacement data.

[0110] The signal processing module is used to extract features of positional offset based on spatial displacement data using signal processing algorithms, separate abnormal fluctuation signals caused by blood flow, and determine the real-time offset of the catheter tip.

[0111] The path adjustment module is used to generate correction instructions if the real-time offset exceeds the preset threshold range, and to calculate the adjustment parameters for dynamic positioning based on the magnitude and direction of the offset.

[0112] The status judgment module is used to obtain the latest motion trajectory data of the catheter tip in the vascular environment based on the adjusted parameters, and to determine whether the catheter is close to an abnormal contact state by combining the distribution characteristics of bending resistance.

[0113] The contact warning module is used to extract the distance signal between the catheter tip and the blood vessel wall if it is determined that the catheter is approaching an abnormal contact state, and to quantitatively assess the risk of tissue damage to obtain the contact risk level.

[0114] The risk assessment module is used to determine whether further path adjustment is needed at the catheter tip by obtaining motion change characteristics in real-time monitoring data based on the exposure risk level and combining them with the dynamic fluctuations of blood flow influence.

[0115] The navigation calibration module is used to continuously calibrate the three-dimensional spatial position of the catheter tip using precise navigation algorithms based on the motion trajectory data after path adjustment, so as to obtain the final stable positioning result.

[0116] The monitoring module is used to acquire long-term motion trend data of the catheter in the vascular environment based on the final stable positioning result, continuously monitor the potential risks of positional deviation and abnormal contact, and judge the operational stability of the navigation system.

[0117] The above are merely preferred embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for precise navigation and positioning of PICC catheters based on magnetic induction, characterized in that, include: Spatial displacement data of the PICC conduit is acquired using a magnetic induction-based sensor module. Offset features are extracted from spatial location data to obtain the offset; The offset is judged and corrected to obtain the adjustment parameters for dynamic positioning; the PICC catheter is controlled according to the adjustment parameters to obtain the motion trajectory data of the PICC catheter; abnormal contact status is judged on the motion trajectory data; after the abnormal contact status is judged, the distance signal between the PICC catheter and the blood vessel wall in the blood vessel model is obtained; the contact risk level is obtained according to the distance signal; the motion change characteristics of the motion trajectory data are obtained according to the contact risk level; the path of the PICC catheter is adjusted according to the motion change characteristics to obtain the adjusted motion trajectory data; the adjusted motion trajectory data is calibrated to obtain the accurate positioning result.

2. The method according to claim 1, characterized in that, The process of acquiring spatial displacement data includes: By analyzing the changes in the external magnetic field and the electromagnetic induction sensor installed at the end of the PICC conduit, the position of the PICC conduit end at different times can be obtained, and spatial displacement data can be obtained.

3. The method according to claim 1, characterized in that, The process of obtaining the offset includes: The position offset is calculated by calculating the corresponding difference between the spatial displacement data and the preset path. The position offset is then decomposed to obtain preliminary offset features. Abnormal fluctuations in the preliminary offset features are decomposed using feature analysis technology to obtain a set of interference signals. The set of interference signals is judged and filtered. Based on the filtered interference signals and the position offset data, the actual position offset of the catheter tip is corrected and calculated to obtain the offset.

4. The method according to claim 1, characterized in that, The process of obtaining adjustment parameters for dynamic positioning includes: The offset is initially processed to obtain its magnitude and direction. Based on the magnitude and direction of the offset, it is determined whether path adjustment is needed. If path adjustment is needed, the adjustment parameters for dynamic positioning are obtained based on the mapping relationship between the magnitude and direction of the offset and the adjustment parameters.

5. The method according to claim 1, characterized in that, The process of determining abnormal contact status in motion trajectory data includes: The polarity of the motion trajectory data is processed to obtain positional offset features. Based on the positional offset features and the distribution characteristics of bending resistance, the force on the catheter tip in the vascular environment is obtained. If the force distribution exceeds the preset normal range, the motion trajectory data is calibrated a second time to obtain calibrated trajectory change information. Based on the calibrated trajectory change information and the judgment criteria of critical state, the support vector machine algorithm is used to classify and predict whether the catheter tip is close to abnormal contact, and the classification result of abnormal contact is obtained.

6. The method according to claim 1, characterized in that, The process of obtaining the exposure risk level includes: A geometric model of the blood vessel is constructed, and the geometric model and motion trajectory data are unified in a coordinate system. The distance data between the motion trajectory data and the blood vessel wall in the geometric model is obtained. Based on the distance data, a machine learning model is used to identify and obtain the contact risk level.

7. The method according to claim 1, characterized in that, The process of acquiring the adjusted motion trajectory data includes: When the risk level is high, corresponding motion rate change features are extracted from the motion trajectory data. A support vector machine (SVM) algorithm is used to classify and analyze these features, determining the degree of matching between the features and the risk level. Based on this matching, it is determined whether the catheter tip has deviated from the expected path. Reference data for path adjustment is extracted based on this deviation. The degree of path deviation is then quantified using this reference data to obtain a preliminary adjustment plan. This preliminary adjustment plan, combined with real-time monitored motion rate changes and blood flow dynamics data, is optimized using a logistic regression algorithm to determine the final path trajectory data.

8. A PICC catheter precision navigation and positioning system based on magnetic induction, characterized in that, Used to perform the method according to any one of claims 1-7.

Images