High-frequency instrument power adjusting system based on image recognition and impedance recognition

Abstract

The invention provides a high-frequency instrument power adjusting system based on image recognition and impedance recognition, and the system comprises an image recognition module which is used for responding to a collection instruction, collecting an image of a tissue, and carrying out the feature recognition of the image, and obtaining an image feature; the impedance identification module is used for responding to the acquisition instruction, acquiring an electric signal of the tissue and processing the electric signal to obtain impedance information; the main control module is used for responding to the control signal, generating an acquisition instruction, sending the acquisition instruction to the image identification module and the impedance identification module, and carrying out weighted fusion calculation on image features and impedance information to obtain a power issuing parameter; and the power control module is used for adjusting the power of the high-frequency instrument based on the power issuing parameters.

Description

Technical Field

[0001] This disclosure relates to the field of biomedical engineering technology, specifically to the field of medical device technology, and in particular to a high-frequency instrument power regulation system based on image recognition and impedance recognition. Background Technology

[0002] In recent years, with the development of minimally invasive surgery, high-frequency electrosurgical units, as the most widely used electrosurgical instruments, have faced increasingly higher requirements for performance and safety. High-frequency electrosurgical units use high-frequency, high-voltage current generated at the tip of the surgical electrode to create a concentrated thermal effect on the local tissue, causing rapid vaporization of tissue cells, thereby achieving different effects such as cutting, coagulation, or hemostasis.

[0003] There are generally two types of power adjustment for existing high-frequency electrosurgical units. One is a fixed working power, which relies on the doctor's experience for manual adjustment. This is inefficient and cannot adapt to real-time impedance changes caused by changes in tissue type or state, and there is a risk of over- or under-power. The other is automatic adjustment based on impedance parameters. This involves collecting impedance information through the electrosurgical electrodes during the operation and adjusting the output power in real time based on the impedance information. However, relying on a single impedance information means that the tissue identification ability is insufficient, and factors such as tissue contact area and electrode aging can easily lead to inaccurate power control, increasing surgical risks. Summary of the Invention

[0004] In view of the above problems, this disclosure provides a high-frequency instrument power regulation system based on image recognition and impedance recognition, which at least partially solves the above technical problems.

[0005] This disclosure provides a high-frequency medical device power adjustment system based on image recognition and impedance identification, comprising: an image recognition module for acquiring an image of tissue in response to an acquisition command, and performing feature recognition on the image to obtain image features; an impedance identification module for acquiring electrical signals of the tissue in response to an acquisition command, and processing the electrical signals to obtain impedance information; a main control module for generating an acquisition command and sending it to the image recognition module and the impedance identification module in response to a control signal, and performing weighted fusion calculation on the image features and impedance information to obtain power transmission parameters; and a power control module for adjusting the power of the high-frequency medical device based on the power transmission parameters.

[0006] According to embodiments of this disclosure, the image recognition module performs feature recognition on an image to obtain image features, including: preprocessing the image to remove noise and obtain a preprocessed image; adjusting the contrast and brightness of the preprocessed image and performing image segmentation to obtain a region of interest, the region of interest being able to highlight the features of the image and the tissue features of the high-frequency instrument contact area; inputting the region of interest into a lightweight pre-trained model for feature recognition and outputting image features, the image features including image features, tissue type, and tissue morphology.

[0007] According to embodiments of this disclosure, the impedance identification module processes electrical signals to obtain impedance information, including: filtering and amplifying the acquired voltage and current signals to obtain amplified voltage and current signals; converting the amplified voltage and current signals into analog-to-digital signals to obtain a first digital signal and a second digital signal; and calculating the first impedance information and the second impedance information based on a biological ternary impedance model using the first and second digital signals. The first impedance information is used to characterize the conductivity of extracellular fluid, and the second impedance information is used to characterize the conductivity of intracellular fluid.

[0008] According to embodiments of this disclosure, the main control module performs weighted fusion calculations on image features and impedance information to obtain power transmission parameters, including: calculating a first output power based on a morphology-output power model using image features; calculating a second output power based on an impedance-output power model using impedance information; performing weighted fusion calculations on the first output power and the second output power based on a first static weight and a second static weight assigned to the morphology-output power model and the impedance-output power model respectively, based on prior verified knowledge, to obtain a target power; and determining the power transmission parameters based on the target power.

[0009] According to embodiments of this disclosure, the main control module performs weighted fusion calculation on image features and impedance information to obtain power transmission parameters, and further includes: calculating a first output power based on a morphology-output power model using image features; calculating a second output power based on an impedance-output power model using impedance information; assigning a first dynamic weight and a second dynamic weight to the morphology-output power model and the impedance-output power model respectively based on a first confidence level output by the morphology-output power model and a second confidence level output by the impedance-output power model; performing weighted fusion calculation on the first output power and the second output power based on the first dynamic weight and the second dynamic weight to obtain a target power; and determining the power transmission parameters based on the target power.

[0010] According to embodiments of this disclosure, the main control module is also used to control the synchronous acquisition of images and electrical signals of tissues by the image recognition module and the impedance recognition module.

[0011] According to embodiments of this disclosure, the main control module generates a trigger pulse via hardware to simultaneously trigger the image recognition module and the impedance recognition module to synchronously acquire images and tissue electrical signals; and / or, the main control module performs time alignment via software to control the synchronous acquisition of images and tissue electrical signals by the image recognition module and the impedance recognition module.

[0012] According to embodiments of this disclosure, the power control module includes: a rectifier-filter circuit for rectifying and filtering an electrical signal input to the power supply; a frequency conversion modulation circuit for frequency conversion modulation of the rectified and filtered electrical signal; a power amplifier circuit for power amplifying the frequency-modulated electrical signal and outputting the amplified signal to a high-frequency instrument to adjust the power of the high-frequency instrument; and a feedback circuit for correcting the pulse width modulation duty cycle of the rectified and filtered electrical signal input to the frequency conversion modulation circuit based on the amplified signal and power output parameters.

[0013] According to embodiments of this disclosure, the power control module has a response speed of less than 1ms, outputs to the high-frequency device at a frequency of 300kHz to 5MHz, and a power of 10W to 300W.

[0014] According to embodiments of this disclosure, the high-frequency instrument includes a high-frequency electrosurgical unit.

[0015] The high-frequency instrument power regulation system based on image recognition and impedance recognition disclosed herein has at least the following technical advantages:

[0016] This system, by setting up independent image recognition and impedance recognition modules, simultaneously acquires the image features and impedance information of tissues, enabling real-time and accurate identification of tissue type and morphology. Furthermore, by weighting and fusing the dual-modal data of image features and impedance information, it can adapt to the surgical needs of different tissue types and surgical scenarios without requiring adjustments by the surgeon, significantly improving the intelligence and adaptability of high-frequency instruments.

[0017] This system can fuse dual-mode data of image and impedance, dynamically adjust the weights, achieve real-time and accurate tissue power matching, find the optimal power output value, and greatly improve the accuracy of power regulation.

[0018] The system employs both image and impedance-based safety measures, significantly enhancing the safety of high-frequency instruments such as electrosurgical units. Attached Figure Description

[0019] The foregoing contents, as well as other objects, features, and advantages of this disclosure, will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0020] Figure 1A schematic block diagram of a high-frequency instrument power regulation system based on image recognition and impedance recognition according to an embodiment of the present disclosure is shown.

[0021] Figure 2 A flowchart illustrating the operation of an image recognition module according to an embodiment of the present disclosure is shown.

[0022] Figure 3 A schematic diagram illustrating the operation of an impedance identification module according to an embodiment of the present disclosure is provided.

[0023] Figure 4 A schematic flowchart of the main control module according to an embodiment of the present disclosure is shown.

[0024] Figure 5 An architectural diagram of a power control module according to an embodiment of the present disclosure is shown schematically. Detailed Implementation

[0025] The embodiments of the present disclosure will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the disclosure. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of the present disclosure for ease of explanation. However, it will be apparent that one or more embodiments may be practiced without these specific details. Furthermore, descriptions of well-known structures and techniques are omitted in the following description to avoid unnecessarily obscuring the concepts of the present disclosure.

[0026] Figure 1 A schematic block diagram of a high-frequency instrument power regulation system based on image recognition and impedance recognition according to an embodiment of the present disclosure is shown.

[0027] like Figure 1 As shown, the high-frequency instrument power regulation system based on image recognition and impedance recognition in this embodiment may include: a main control module, a power control module, an impedance recognition module, and an image recognition module.

[0028] The image recognition module is used to acquire images of the tissue in response to acquisition commands, and to perform feature recognition on the images to obtain image features.

[0029] The impedance identification module is used to acquire electrical signals from the tissue in response to acquisition commands, and to process the electrical signals to obtain impedance information.

[0030] The main control module is used to respond to control signals, generate acquisition commands and send them to the image recognition module and impedance recognition module, and perform weighted fusion calculations on image features and impedance information to obtain power transmission parameters.

[0031] The power control module is used to adjust the power of high-frequency instruments based on power-downloaded parameters.

[0032] For example, after receiving the control signal to start the doctor's action, the main control module sends acquisition instructions to the impedance identification module and the image recognition module at a preset frequency. The impedance identification module collects tissue impedance information through high-frequency instruments, calculates the results, and uploads them to the main control module. The image recognition module collects tissue image information through the lens, analyzes the results, and uploads them to the main control module. The main control module weights and fuses the impedance identification results and image recognition results, and then outputs power parameters and sends them to the power control module. The power control module supplies the mains power to the high-frequency instruments after frequency conversion, voltage transformation, and power amplification according to the parameters sent by the main control module to complete the output.

[0033] In some embodiments, the high-frequency instrument may include, but is not limited to, a high-frequency electrosurgical unit.

[0034] In some embodiments, the image recognition module performs feature recognition on the image to obtain image features, which may include: preprocessing the image to remove noise and obtain a preprocessed image; adjusting the contrast and brightness of the preprocessed image and performing image segmentation to obtain a region of interest, the region of interest being able to highlight the features of the image and the tissue features of the high-frequency instrument contact area; inputting the region of interest into a lightweight pre-trained model for feature recognition and outputting image features, the image features including image features, tissue type and tissue morphology.

[0035] Figure 2 A flowchart illustrating the operation of an image recognition module according to an embodiment of the present disclosure is shown.

[0036] like Figure 2 As illustrated, the image recognition module, for example, uses the endoscope used during surgery to acquire images. Upon receiving an acquisition command, it acquires the current frame of the image. First, the image is preprocessed to remove noise, adjust contrast and brightness, segment the Region of Interest (ROI), and highlight image features and tissue features in the area contacted by the high-frequency electrosurgical unit. Next, a deep learning model deployed on the image recognition module performs feature recognition. Finally, the image recognition results are output to the main control module, including image features (smoke, pollution, etc.), tissue type (such as blood vessels, nerves, muscles, fat, etc.), and morphology (such as edema, bleeding, normal, carbonization, etc.).

[0037] In some embodiments, the impedance identification module processes electrical signals to obtain impedance information, which may include: filtering and amplifying the acquired voltage and current signals to obtain amplified voltage and current signals; converting the amplified voltage and current signals into analog and digital signals respectively to obtain a first digital signal and a second digital signal; and calculating the first impedance information and the second impedance information based on the biological ternary impedance model using the first and second digital signals, wherein the first impedance information is used to characterize the conductivity of extracellular fluid and the second impedance information is used to characterize the conductivity of intracellular fluid.

[0038] Figure 3 A schematic diagram illustrating the operation of an impedance identification module according to an embodiment of the present disclosure is provided.

[0039] like Figure 3 As shown, for example, after receiving the acquisition command, the impedance identification module acquires tissue voltage and current signals through a high-frequency instrument. The high-voltage signal is attenuated to a safe range through a voltage divider circuit, then processed by signal filtering and amplification. Finally, after analog-to-digital conversion by an A / D chip, the tissue impedance value is calculated using a biological ternary impedance model, and the impedance calculation result is uploaded to the main control module.

[0040] In some embodiments, the main control module performs weighted fusion calculation on image features and impedance information to obtain power transmission parameters, which may include: calculating a first output power based on image features using a morphology-output power model; calculating a second output power based on impedance information using an impedance-output power model; performing weighted fusion calculation on the first output power and the second output power based on a first static weight and a second static weight assigned to the morphology-output power model and the impedance-output power model respectively based on prior verified knowledge to obtain a target power; and determining the power transmission parameters based on the target power.

[0041] Furthermore, the main control module performs weighted fusion calculations on image features and impedance information to obtain power transmission parameters, which may include: calculating a first output power based on the morphology-output power model using image features; calculating a second output power based on the impedance-output power model using impedance information; assigning a first dynamic weight and a second dynamic weight to the morphology-output power model and the impedance-output power model respectively based on the first confidence level output by the morphology-output power model and the second confidence level output by the impedance-output power model; performing weighted fusion calculations on the first output power and the second output power based on the first dynamic weight and the second dynamic weight to obtain the target power; and determining the power transmission parameters based on the target power.

[0042] Figure 4 A schematic flowchart of the main control module according to an embodiment of the present disclosure is shown.

[0043] like Figure 4 As illustrated, the main control module serves as the computational and control center of the electrosurgical system. Upon receiving external operational commands from the doctor, it analyzes and calculates the parameters based on the commands and the acquired impedance and image signals, then sends them to control the system to complete the output. The module incorporates an "impedance-output power model," a "morphology-output power model," and a "morphology-impedance correspondence model." After the doctor issues an operation command (start, electrosurgical resection, electrocoagulation, mixed resection, stop, etc.), it first sends acquisition commands to the impedance recognition module and the image recognition module. Based on the feedback information, it determines the patient's initial state, calibrates the "morphology-impedance correspondence model," and adjusts the "impedance-output power model" and "morphology-output power model" accordingly before starting output. After the operation begins, the main control module sends acquisition commands to the impedance recognition module and the image recognition module according to the preset impedance acquisition frequency and image acquisition frequency, respectively. After receiving the acquisition results, it assigns weights to the "impedance-output power model" and the "morphology-output power model," and then calculates the power transmission parameters. This acquisition-calculation loop is executed, continuously calculating and adjusting the transmission parameters until the feedback result indicates operation completion or abnormal termination.

[0044] The built-in "morphology-impedance correspondence model" is derived from average human parameters. However, due to individual patient differences, the tissue electrochemical properties will vary accordingly. Therefore, the model needs to be calibrated based on individual patient differences to improve the system's adaptability. The "impedance-output power model" and "morphology-output power model" can be trained based on surgical experimental data, expert experience data, or database models. They need to be adjusted accordingly based on the calibration results of the "morphology-impedance correspondence model." That is, the output power value corresponding to a set of morphological and impedance characteristics should be equal in their respective power output correspondence models.

[0045] For example, the weights of the "impedance-output power model" and the "morphology-output power model" are assigned, with the sum of the two being 100%. The assignment can be static: initially, both impedance and image weights are 50%. When bleeding occurs, the image weight decreases to 20%; when the surgical field smoke concentration is too high, the image weight decreases to 30%; when a blood vessel is closed, the image weight increases to 70%; when the tissue is charred or carbonized, the image weight increases to 80%. Based on prior knowledge or expert experience, the weights can be set for different tissue types or states, and the weights can be dynamically adjusted during surgery according to changes in the state. Alternatively, the weights can be dynamic, using a lightweight deep learning algorithm. Based on preset "impedance-power curves" and "morphology-power curves," the model is trained using surgical procedures from a database to establish a model that dynamically adjusts the weights based on the confidence level of single-modal predictions. For example, the weight of an image modality is reduced when its quality is low, or when the impedance change rate is too high.

[0046] In some embodiments, the main control module is also used to control the image recognition module and the impedance recognition module to synchronously acquire images and electrical signals of tissues.

[0047] Furthermore, the main control module generates trigger pulses through hardware to simultaneously trigger the image recognition module and the impedance recognition module to synchronously acquire images and tissue electrical signals; and / or, the main control module uses software time alignment to control the synchronous acquisition of images and tissue electrical signals by the image recognition module and the impedance recognition module.

[0048] For example, in the dual-modal fusion of impedance data and image data, data synchronization must be ensured. This can be achieved through hardware trigger pin alignment, such as using a control chip to generate precise trigger pulses to simultaneously trigger image and impedance acquisition, eliminating differences in acquisition timing. Alternatively, software timestamp alignment can be used, matching the nearest-time data in the impedance data based on the image timestamp. A combination of both methods can also be used to achieve real-time data processing.

[0049] In some embodiments, the power control module includes: a rectifier-filter circuit for rectifying and filtering the electrical signal input to the power supply; a frequency conversion modulation circuit for frequency conversion modulation of the rectified and filtered electrical signal; a power amplifier circuit for power amplifying the frequency-modulated electrical signal and outputting the amplified signal to a high-frequency instrument to adjust the power of the high-frequency instrument; and a feedback circuit for correcting the pulse width modulation duty cycle of the rectified and filtered electrical signal input to the frequency conversion modulation circuit based on the amplified signal and power output parameters.

[0050] Furthermore, the power control module has a response speed of less than 1ms, and the output frequency to the high-frequency instrument is 300kHz~5MHz, with a power of 10W~300W.

[0051] Figure 5 An architectural diagram of a power control module according to an embodiment of the present disclosure is shown schematically.

[0052] like Figure 5 As shown, the power control module is the output terminal of the electrosurgical system. Based on parameters (including frequency, waveform, voltage, etc.) issued by the main control module, it uses pulse width modulation (PWM) technology to rectify, filter, modulate, and amplify the input power before outputting it. By monitoring the output current and voltage in real time, and then performing A / D conversion and error amplification, the PWM duty cycle is dynamically corrected to maintain accurate output power. Using PWM output regulation, a response speed of <1ms can be achieved, and any waveform within the frequency range of 300kHz to 5MHz and power range of 10W to 300W can be output.

[0053] Based on the aforementioned high-frequency instrument power adjustment system using image recognition and impedance recognition, precise adjustment of the electrosurgical system's output power is achieved. Furthermore, the fusion of impedance and image dual-mode can significantly increase surgical safety. If the camera malfunctions during surgery, it can switch to pure impedance mode, or utilize the safety redundancy of image recognition when impedance data is abnormal; there are also dual safety threshold criteria for both image recognition and impedance recognition states.

[0054] The embodiments of this disclosure have been described above. However, these embodiments are for illustrative purposes only and are not intended to limit the scope of this disclosure. Although various embodiments have been described above, this does not mean that the measures in the various embodiments cannot be used advantageously in combination. The scope of this disclosure is defined by the appended claims and their equivalents. Various substitutions and modifications can be made by those skilled in the art without departing from the scope of this disclosure, and all such substitutions and modifications should fall within the scope of this disclosure.

Claims

1. A high-frequency instrument power regulation system based on image recognition and impedance recognition, characterized in that, include: An image recognition module is used to acquire an image of the tissue in response to an acquisition command, and to perform feature recognition on the image to obtain image features; An impedance identification module is used to acquire electrical signals of the tissue in response to the acquisition command, and process the electrical signals to obtain impedance information; The main control module is used to respond to the control signal, generate the acquisition command and send it to the image recognition module and the impedance recognition module, and perform weighted fusion calculation on the image features and the impedance information to obtain the power transmission parameters; A power control module is used to adjust the power of the high-frequency instrument based on the power output parameters.

2. The system according to claim 1, characterized in that, The image recognition module performs feature recognition on the image to obtain image features, including: The image is preprocessed to remove noise, resulting in a preprocessed image. The contrast and brightness of the preprocessed image are adjusted, and image segmentation is performed to obtain the region of interest, which can highlight the image features and the tissue features of the high-frequency instrument contact area; The region of interest is input into a lightweight pre-trained model for feature recognition, and the image features are output. The image features include image features, tissue type, and tissue morphology.

3. The system according to claim 1, characterized in that, The impedance identification module processes the electrical signal to obtain impedance information, including: The acquired voltage and current signals are filtered and amplified to obtain amplified voltage and current signals. The amplified voltage signal and the amplified current signal; The amplified voltage signal and the amplified current signal are converted from analog to digital respectively to obtain the first digital signal and the second digital signal; Based on the biological ternary impedance model, first impedance information and second impedance information are calculated using the first digital signal and the second digital signal. The first impedance information is used to characterize the conductivity of extracellular fluid, and the second impedance information is used to characterize the conductivity of intracellular fluid.

4. The system according to any one of claims 1 to 3, characterized in that, The main control module performs weighted fusion calculations on the image features and the impedance information to obtain power transmission parameters, including: The first output power is calculated based on the morphology-output power model using the image features. The second output power is calculated based on the impedance-output power model and using the impedance information. Based on the first static weight and the second static weight assigned to the morphology-output power model and the impedance-output power model respectively by prior verified knowledge, the first output power and the second output power are weighted and fused to obtain the target power. The power distribution parameters are determined based on the target power.

5. The system according to any one of claims 1 to 3, characterized in that, The main control module performs weighted fusion calculations on the image features and the impedance information to obtain power transmission parameters, and also includes: The first output power is calculated based on the morphology-output power model using the image features. The second output power is calculated based on the impedance-output power model and using the impedance information. The first confidence level output by the morphology-output power model and the second confidence level output by the impedance-output power model are the first dynamic weights and the second dynamic weights assigned to the morphology-output power model and the impedance-output power model, respectively. The first output power and the second output power are weighted and fused based on the first dynamic weight and the second dynamic weight to obtain the target power; The power distribution parameters are determined based on the target power.

6. The system according to claim 5, characterized in that, The main control module is also used to control the synchronous acquisition of the image and the electrical signals of the tissue by the image recognition module and the impedance recognition module.

7. The system according to claim 6, characterized in that, The main control module generates a trigger pulse through hardware to simultaneously trigger the image recognition module and the impedance recognition module to acquire the electrical signals of the image and the tissue. and / or; The main control module uses software time alignment to control the synchronous acquisition of the image and the electrical signals of the tissue by the image recognition module and the impedance recognition module.

8. The system according to claim 1, characterized in that, The power control module includes: A rectifier and filter circuit is used to rectify and filter the electrical signal input to the power supply. A frequency conversion modulation circuit is used to perform frequency conversion modulation on the rectified and filtered electrical signal. The power amplifier circuit is used to amplify the power of the frequency-modulated electrical signal and output the amplified signal to the high-frequency instrument to adjust the power of the high-frequency instrument. The feedback circuit is used to correct the pulse width modulation duty cycle of the rectified and filtered electrical signal input to the frequency converter modulation circuit based on the amplified signal and the power output parameters.

9. The system according to claim 8, characterized in that, The power control module has a response speed of less than 1ms, and the output frequency to the high-frequency instrument is 300kHz~5MHz, with a power of 10W~300W.

10. The system according to claim 1, characterized in that, The high-frequency instrument includes a high-frequency electrosurgical unit.

Images