High-pressure injector stroke monitoring system and method

Abstract

The invention relates to a high-pressure injector stroke monitoring system and method, and belongs to the technical field of medical electronic instrument safety control, and the system comprises a sensing fusion module and a control module; the sensing fusion module comprises a first potentiometer, a second potentiometer, an optical encoder and a magnetostrictive sensor, and is used for respectively acquiring initial strokes of the injector according to the first potentiometer, the second potentiometer, the optical encoder and the magnetostrictive sensor and sending data of each initial stroke to the control module; the control module is used for carrying out weighted calculation on the initial strokes collected by the first potentiometer, the second potentiometer, the optical encoder and the magnetostrictive sensor respectively, determining the actual stroke of the injector, determining the actual deviation between the actual stroke and the theoretical stroke of the injector, and calculating the actual stroke of the injector according to the size relation between the actual deviation and a deviation threshold value. And judging whether the injector is abnormal or not. The technical problem that in the prior art, the stroke monitoring precision of the high-pressure injector is low is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of medical electronic equipment safety control, and in particular to a high-pressure injector stroke monitoring system and method. Background Art

[0002] During medical imaging procedures, high-pressure injectors are used to quickly and accurately inject contrast agents and other liquids. Currently, stroke monitoring for high-pressure injectors on the market primarily relies on single potentiometers. However, single potentiometers have limited accuracy and are susceptible to wear and aging over time, resulting in inaccurate test data, difficulty detecting minor anomalies, and inability to effectively address interference under complex operating conditions. Furthermore, existing abnormal dose warning methods lack unified standards, suffer from poor real-time performance, and are difficult to assess under specific injection modes. These issues severely impact injection effectiveness and patient safety.

[0003] Specifically, high-pressure injectors typically rely on sensors to monitor stroke, such as potentiometers, encoders, and limit switches. However, these sensors have inherent accuracy limitations and can suffer from wear and aging over time, leading to inaccurate detection data. For example, during use, the contacts of a potentiometer may become loose, causing deviations in the reported stroke data and compromising the accurate assessment of stroke anomalies. In actual clinical use, high-pressure injectors are subject to interference from various factors, such as electromagnetic interference, mechanical vibration, and temperature fluctuations. These factors can cause fluctuations in stroke detection data, making anomaly detection difficult. Current anomaly detection methods often struggle to effectively eliminate the influence of these interfering factors, thus compromising accuracy. High-pressure injectors generate a large amount of stroke data during operation, requiring timely processing and analysis to determine whether an anomaly exists. However, the limited data processing capabilities of some devices can lead to data processing lags and inability to provide real-time feedback on stroke anomalies. By the time an anomaly is detected, it may have already adversely affected the injection process, potentially causing the test to fail or even harming the patient. Even when a stroke anomaly is detected, the alarm systems of some high-pressure injectors may not respond promptly. For example, if the alarm delay is too long, medical staff may not be able to take timely measures, thus affecting the patient's examination and treatment effects.

[0004] The core operating principle of a single potentiometer sensor is that mechanical displacement causes the wiper in the installed potentiometer to slide, changing the resistance value. This changes the voltage across the potentiometer, which is then converted into data via an ADC to calculate position information. Single potentiometers are susceptible to changes in contact resistance, leading to false positives and immediate system shutdowns. Pressure feedback serves only as auxiliary verification and cannot independently distinguish between mechanical jamming and sensor failure. The Bayer Medrad Centargo uses a single potentiometer with pressure feedback. Wear of the potentiometer's carbon film resulted in a 20ml under-injection of CT contrast agent. The pressure sensor was also affected by blood clots, falsely triggering an emergency shutdown. The core operating principle of a limit switch sensor is that mechanical displacement causes contact, causing a spring-compressed contact to switch the output signal, halting the entire injection system. However, the front and rear in-position switches can only be used to indicate the zero and maximum injectable doses of the entire range. In practice, the actual dose variation error is unknown, resulting in excessive deviations during real-time injections. Summary of the Invention

[0005] In view of this, it is necessary to provide a high-pressure injector stroke monitoring system and method to solve the technical problem of low accuracy of high-pressure injector stroke monitoring in the prior art.

[0006] In order to solve the above problems, the present invention provides a high-pressure injector stroke monitoring system, comprising: a sensor fusion module and a control module;

[0007] The sensor fusion module includes a first potentiometer, a second potentiometer, an optical encoder and a magnetostrictive sensor, and is used to respectively collect the initial stroke of the syringe according to the first potentiometer, the second potentiometer, the optical encoder and the magnetostrictive sensor, and send the data of each initial stroke to the control module;

[0008] The control module is used to perform weighted calculation on the initial strokes respectively collected by the first potentiometer, the second potentiometer, the optical encoder, and the magnetostrictive sensor to determine the actual stroke of the syringe, and determine the actual deviation between the actual stroke of the syringe and the theoretical stroke, and determine whether the syringe has an abnormality based on the relationship between the actual deviation and the deviation threshold.

[0009] In a possible implementation, the first potentiometer and the second potentiometer are symmetrically mounted on both sides of the high-pressure injector push rod, and are used to collect analog signals of the push rod stroke in real time;

[0010] The optical encoder measures the displacement of the syringe plunger in a non-contact manner and calibrates the zero drift error of the first potentiometer and the second potentiometer through absolute position feedback.

[0011] In a possible implementation, the control module includes a confidence weight calculation unit;

[0012] The confidence weight calculation unit is used to dynamically allocate the confidence weight of each sensor in the sensor fusion module using a preset weighted adaptive fusion algorithm.

[0013] In one possible implementation, the preset weighted adaptive fusion algorithm is Bayesian filtering, and the weight update rules include:

[0014] When the difference between any one of the first potentiometer, the second potentiometer, or the magnetostrictive sensor and the optical encoder exceeds a weight deviation threshold, reducing the weight of the first potentiometer, the second potentiometer, or the magnetostrictive sensor whose difference with the optical encoder exceeds the weight deviation threshold to a first weight value;

[0015] In a nuclear magnetic resonance environment, the weight of the magnetostrictive sensor is increased to a second weight value.

[0016] In a possible implementation, the control module further includes a deviation threshold adjustment unit;

[0017] The deviation threshold adjustment unit is used to dynamically adjust the dosage deviation threshold using the injection rate, injection liquid viscosity and historical deviation data of the syringe as input data using a preset reinforcement learning model.

[0018] In a possible implementation, the control module further includes an abnormality diagnosis unit;

[0019] The abnormality diagnosis unit is used to compare whether the actual deviation is greater than the deviation threshold;

[0020] If so, compare whether the difference between the first potentiometer and the second potentiometer is greater than the first fault threshold. If so, determine that it is a sensor failure and switch the sensor fusion module to the optical encoder independent working mode;

[0021] If the difference between the first potentiometer and the second potentiometer is less than the first fault threshold, it is determined whether there is a stroke difference between the optical encoder and the magnetostrictive sensor. If so, it is an electromagnetic interference fault; if not, it is a mechanical jamming fault.

[0022] In a possible implementation, a fault warning module is also included;

[0023] The fault warning module is in communication with the control module, and when the control module determines that an abnormality occurs in the syringe, the fault alarm module is triggered to sound an alarm.

[0024] In a possible implementation, it further includes a data processing module;

[0025] The data processing module is communicatively connected to the sensor fusion module, and is used to receive the data of each initial stroke collected by the sensor fusion module, filter the data of each initial stroke, and transmit the processed data of each initial stroke to the control module.

[0026] In a possible implementation, a prediction module is further included;

[0027] The prediction module is used to predict future dose deviations using a well-trained LSTM neural network, wherein the input parameters of the neural network include historical travel data, pressure and temperature sequences.

[0028] In a second aspect, the present invention further provides a high-pressure injector stroke monitoring method, which is applied to any of the above-mentioned high-pressure injector stroke monitoring systems, comprising:

[0029] Acquire multiple syringe initial stroke data collected by multiple different types of sensors in real time;

[0030] Performing weighted calculation on the plurality of initial stroke data to determine an actual stroke of the syringe, and determining an actual deviation between the actual stroke of the syringe and a theoretical stroke;

[0031] Based on the relationship between the actual deviation and the deviation threshold, it is determined whether the syringe has an abnormality.

[0032] The present invention has the following beneficial effects: by providing a sensor fusion module comprising a first potentiometer, a second potentiometer, an optical encoder, and a magnetostrictive sensor, and collecting syringe stroke data based on multiple different sensor types, the problem of large stroke acquisition errors caused by single sensor failure can be significantly reduced. Furthermore, the magnetostrictive sensor ensures that data acquisition is not subject to magnetic field interference in a nuclear magnetic resonance environment, providing accurate stroke data. Dynamic calibration of sensor data using a weighted adaptive algorithm significantly improves the accuracy and anti-interference capability of stroke detection. A dynamic threshold adjustment mechanism is used to optimize the accuracy of anomaly detection, further enhancing the accuracy of syringe stroke detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 This is a system architecture diagram of an embodiment of a high-pressure injector stroke monitoring system provided by the present invention;

[0034] Figure 2 This is a flow chart of an embodiment of a high-pressure injector stroke monitoring method provided by the present invention. DETAILED DESCRIPTION

[0035] The preferred embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings, wherein the accompanying drawings constitute a part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, and are not used to limit the scope of the present invention.

[0036] A specific embodiment of the present invention discloses a high-pressure injector stroke monitoring system, see Figure 1 , including: sensor fusion module 1 and control module 2;

[0037] The sensor fusion module includes a first potentiometer 11, a second potentiometer 12, an optical encoder 13 and a magnetostrictive sensor 14, which is used to respectively collect the initial stroke of the syringe according to the first potentiometer, the second potentiometer, the optical encoder and the magnetostrictive sensor, and send the data of each initial stroke to the control module;

[0038] It should be noted that the first and second potentiometers are symmetrically mounted on either side of the high-pressure syringe plunger to collect real-time analog signals of the plunger's travel. The optical encoder provides non-contact displacement measurement, preventing wear of the potentiometer's carbon film. Zero drift errors in the first and second potentiometers are calibrated using absolute position feedback. The magnetostrictive sensor is resistant to electromagnetic interference and compatible with nuclear magnetic resonance (NMR) environments.

[0039] By integrating an optical encoder and a magnetostrictive sensor on the basis of a dual potentiometer, a triple-redundant monitoring system is formed, and the accuracy is improved through multi-source data fusion. An automatic calibration algorithm is added based on the optical encoder, and the redundant data is used to correct the potentiometer drift error in real time, avoiding the problem of inaccurate syringe stroke data collection caused by carbon film wear due to long-term use, thereby further improving the accuracy of stroke collection in this application.

[0040] After installation, the first and second potentiometers, optical encoder, and magnetostrictive sensor are calibrated to ensure that their output signals are consistent when the syringe is in its initial position. When the syringe is operating normally, the stroke data detected by the first and second potentiometers, optical encoder, and magnetostrictive sensor should be essentially consistent. However, if a stroke anomaly occurs, such as a stuck or misaligned pushrod, the actual data detected by the first and second potentiometers, optical encoder, and magnetostrictive sensor may differ. The control module uses a pre-set algorithm to determine whether a stroke anomaly exists.

[0041] The control module is used to perform weighted calculation on the initial strokes respectively collected by the first potentiometer, the second potentiometer, the optical encoder, and the magnetostrictive sensor to determine the actual stroke of the syringe, and determine the actual deviation between the actual stroke of the syringe and the theoretical stroke, and determine whether the syringe has an abnormality based on the relationship between the actual deviation and the deviation threshold.

[0042] It should be noted that the inherent defects of a single sensor can be offset by performing weighted calculation on the initial strokes respectively collected by the first potentiometer, the second potentiometer, the optical encoder and the magnetostrictive sensor through the control module.

[0043] In a specific embodiment, see Figure 1 , the control module includes a confidence weight calculation unit 21;

[0044] The confidence weight calculation unit is used to dynamically allocate the confidence weight of each sensor in the sensor fusion module using a preset weighted adaptive fusion algorithm.

[0045] In this embodiment, the traditional fixed weight scheme cannot adapt to complex scenarios such as high-speed injection (large flow fluctuations) or high-viscosity liquids (sensitive to small deviations), which can easily lead to false alarms or missed detections. Weighted calculation can offset the inherent defects of a single sensor and dynamically adjust the weight based on the real-time data quality of the sensor: during high-speed injection, the optical encoder weight is relaxed to 0.5 (anti-jitter advantage) and the potentiometer weight is reduced to 0.2; in high-viscosity liquids (such as iodized oil): the magnetostrictive sensor weight is tightened to 0.6 (high precision advantage) to ensure ±0.5ml deviation detection sensitivity; in a nuclear magnetic resonance environment, the magnetostrictive sensor weight is increased to 0.7, the optical encoder weight is 0.25, and the potentiometer weight is forced to be reduced to 0.05 (to avoid magnetic field interference);

[0046] In a specific embodiment, the preset weighted adaptive fusion algorithm is Bayesian filtering, and the weight update rules include:

[0047] When the difference between any one of the first potentiometer, the second potentiometer, or the magnetostrictive sensor and the optical encoder exceeds a weight deviation threshold, reducing the weight of the first potentiometer, the second potentiometer, or the magnetostrictive sensor whose difference with the optical encoder exceeds the weight deviation threshold to a first weight value;

[0048] In a nuclear magnetic resonance environment, the weight of the magnetostrictive sensor is increased to a second weight value.

[0049] Among them, the weight deviation threshold, the first weight value and the second weight value can be set according to actual needs and are not limited here.

[0050] It should be noted that the theoretical stroke is calculated by multiplying the injection time by the injection speed. The stroke calculation can be converted into the measurement of the injection dose, and the determination of abnormalities based on the deviation between the actual stroke and the theoretical stroke can be converted into the determination of abnormalities based on the deviation between the actual injection volume and the theoretical injection volume.

[0051] In some embodiments, see Figure 1 The control module further includes a deviation threshold adjustment unit 22;

[0052] The deviation threshold adjustment unit is used to dynamically adjust the dosage deviation threshold using the injection rate, injection liquid viscosity and historical deviation data of the syringe as input data using a preset reinforcement learning model.

[0053] In this embodiment, a reinforcement learning model (such as Q-Learning) uses injection rate, liquid viscosity, and historical deviation data as input to adjust the deviation threshold in real time (e.g., ±0.5ml to ±1.2ml), breaking through the limitations of the traditional fixed threshold (±1ml). In high-speed injection scenarios, the threshold is automatically relaxed to ±1.2ml to avoid premature shutdowns caused by transient flow fluctuations; in high-viscosity liquid scenarios, the threshold is tightened to ±0.5ml, improving the detection sensitivity of small dosage deviations.

[0054] Furthermore, by avoiding manual intervention and accurately adjusting deviations through intelligent algorithms, it is possible to prevent false alarms or missed alarms from occurring on site.

[0055] In some embodiments, see Figure 1 , the control module further includes an abnormality diagnosis unit 23;

[0056] The abnormality diagnosis unit is used to compare whether the actual deviation is greater than the deviation threshold;

[0057] If so, compare whether the difference between the first potentiometer and the second potentiometer is greater than the first fault threshold. If so, determine that it is a sensor failure and switch the sensor fusion module to the optical encoder independent working mode;

[0058] If the difference between the first potentiometer and the second potentiometer is less than the first fault threshold, it is determined whether there is a stroke difference between the optical encoder and the magnetostrictive sensor. If so, it is an electromagnetic interference fault; if not, it is a mechanical jamming fault.

[0059] In this embodiment, a first fault threshold is set to achieve hierarchical fault determination. If any sensor fails, the system automatically downgrades to dual-sensor redundancy mode (e.g., potentiometer + encoder), avoiding downtime and interrupting injection. Furthermore, the fault location can be quickly located and repaired in a timely manner.

[0060] In some embodiments, see Figure 1 , further comprising a fault warning module 3;

[0061] The fault warning module is in communication with the control module, and when the control module determines that an abnormality occurs in the syringe, the fault alarm module is triggered to sound an alarm.

[0062] In this embodiment, once an abnormality is detected, the control module immediately triggers the early warning device, which can sound an alarm through sound, light, or other means, prompting medical staff to take appropriate measures. The control module also records abnormal information, including the time of occurrence, travel time, and dosage deviation, for subsequent analysis and processing.

[0063] In some embodiments, a data processing module 4 is further included;

[0064] The data processing module is communicatively connected to the sensor fusion module, and is used to receive the data of each initial stroke collected by the sensor fusion module, filter the data of each initial stroke, and transmit the processed data of each initial stroke to the control module.

[0065] In this embodiment, the collected data is filtered using algorithms such as mean filtering, RC low-pass filtering, and Kalman filtering to remove noise caused by electromagnetic interference, potentiometer jitter, etc. An STM32 chip is then used as the core, and the voltage data converted by the ADC is collected via DMA, and the actual dose is calculated.

[0066] In a second aspect, the present invention further provides a high-pressure injector stroke monitoring method, which is applied to any of the high-pressure injector stroke monitoring systems described above. Figure 2 ,include:

[0067] S201, acquiring in real time a plurality of syringe initial stroke data collected by a plurality of different types of sensors;

[0068] S202, performing weighted calculation on the multiple initial stroke data to determine the actual stroke of the syringe, and determining the actual deviation between the actual stroke of the syringe and the theoretical stroke;

[0069] S203. Determine whether the syringe has an abnormality based on the relationship between the actual deviation and the deviation threshold.

[0070] In this embodiment, by acquiring travel data from multiple different types of sensors, errors introduced by a single sensor can be eliminated. Furthermore, by using a weighted calculation method to assign weights to multiple travel data points, it is possible to avoid sensing errors in different scenarios caused by a single weight, thereby improving the accuracy of travel detection.

[0071] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by any technician familiar with this technical field within the technical scope disclosed by the present invention should be covered by the scope of protection of the present invention.

Claims

1. A high-pressure injector stroke monitoring system, characterized in that: include: Sensor fusion module and control module; The sensor fusion module includes a first potentiometer, a second potentiometer, an optical encoder and a magnetostrictive sensor, and is used to respectively collect the initial stroke of the syringe according to the first potentiometer, the second potentiometer, the optical encoder and the magnetostrictive sensor, and send the data of each initial stroke to the control module; The control module is used to perform weighted calculation on the initial strokes respectively collected by the first potentiometer, the second potentiometer, the optical encoder, and the magnetostrictive sensor to determine the actual stroke of the syringe, and determine the actual deviation between the actual stroke of the syringe and the theoretical stroke, and determine whether the syringe has an abnormality based on the relationship between the actual deviation and the deviation threshold.

2. The high-pressure injector stroke monitoring system according to claim 1, characterized in that: The first potentiometer and the second potentiometer are symmetrically installed on both sides of the high-pressure injector push rod, and are used to collect the analog signal of the push rod stroke in real time; The optical encoder measures the displacement of the syringe plunger in a non-contact manner and calibrates the zero drift error of the first potentiometer and the second potentiometer through absolute position feedback.

3. The high-pressure injector stroke monitoring system according to claim 1, characterized in that: The control module includes a confidence weight calculation unit; The confidence weight calculation unit is used to dynamically allocate the confidence weight of each sensor in the sensor fusion module using a preset weighted adaptive fusion algorithm.

4. The high-pressure injector stroke monitoring system according to claim 3, characterized in that: The preset weighted adaptive fusion algorithm is Bayesian filtering, and the weight update rules include: When the difference between any one of the first potentiometer, the second potentiometer, or the magnetostrictive sensor and the optical encoder exceeds a weight deviation threshold, reducing the weight of the first potentiometer, the second potentiometer, or the magnetostrictive sensor whose difference with the optical encoder exceeds the weight deviation threshold to a first weight value; In a nuclear magnetic resonance environment, the weight of the magnetostrictive sensor is increased to a second weight value.

5. The high-pressure injector stroke monitoring system according to claim 1, characterized in that: The control module further includes a deviation threshold adjustment unit; The deviation threshold adjustment unit is used to dynamically adjust the dosage deviation threshold using the injection rate, injection liquid viscosity and historical deviation data of the syringe as input data using a preset reinforcement learning model.

6. The high-pressure injector stroke monitoring system according to claim 1, characterized in that: The control module further includes an abnormality diagnosis unit; The abnormality diagnosis unit is used to compare whether the actual deviation is greater than the deviation threshold; If so, compare whether the difference between the first potentiometer and the second potentiometer is greater than the first fault threshold. If so, determine that it is a sensor failure and switch the sensor fusion module to the optical encoder independent working mode; If the difference between the first potentiometer and the second potentiometer is less than the first fault threshold, it is determined whether there is a stroke difference between the optical encoder and the magnetostrictive sensor. If so, it is an electromagnetic interference fault; if not, it is a mechanical jamming fault.

7. The high-pressure injector stroke monitoring system according to claim 1, characterized in that: It also includes a fault warning module; The fault warning module is in communication with the control module, and when the control module determines that an abnormality occurs in the syringe, the fault alarm module is triggered to sound an alarm.

8. The high-pressure injector stroke monitoring system according to claim 1, characterized in that: Also includes a data processing module; The data processing module is communicatively connected to the sensor fusion module, and is used to receive the data of each initial stroke collected by the sensor fusion module, filter the data of each initial stroke, and transmit the processed data of each initial stroke to the control module.

9. The high-pressure injector stroke monitoring system according to claim 1, characterized in that: It also includes a prediction module; The prediction module is used to predict future dose deviations using a well-trained LSTM neural network, wherein the input parameters of the neural network include historical travel data, pressure and temperature sequences.

10. A high-pressure injector stroke monitoring method, applied to the high-pressure injector stroke monitoring system according to any one of claims 1 to 9, characterized in that: include: Acquire multiple syringe initial stroke data collected by multiple different types of sensors in real time; Performing weighted calculation on the plurality of initial stroke data to determine an actual stroke of the syringe, and determining an actual deviation between the actual stroke of the syringe and a theoretical stroke; Based on the relationship between the actual deviation and the deviation threshold, it is determined whether the syringe has an abnormality.

Images