A method of water power assisted autologous fat transplantation breast augmentation

Abstract

The present application relates to a kind of water power auxiliary autologous fat transplantation breast augmentation method, belong to plastic surgery technical field.The method is by multimodal sensor real-time sensing operation parameter, and dynamically predicts the survival rate of transplantation by prognosis prediction model, and water power parameter, injection parameter and biological active additive mixing proportion are adjusted by closed loop control module according to this linkage, form the adaptive closed loop control of physical parameter and biological active substance cooperation;While tissue electrical impedance sensor real-time identification injection needle tip the tissue type in and automatic interlock stop misinjection.The present application realizes the intelligent closed loop control of fat transplantation breast augmentation surgery, effectively improves the survival rate of fat transplantation, surgical safety and effect predictability.

Description

Technical Field

[0001] This invention relates to the field of plastic surgery technology, specifically to a water-assisted autologous fat transplantation method for breast augmentation. Background Technology

[0002] Autologous fat grafting for breast augmentation is a surgical procedure that involves collecting and purifying fat tissue from a donor area (such as the abdomen, inner thigh, or buttocks) and injecting it into the breasts to increase breast volume. Because the graft is derived from the patient's own tissue, it offers advantages such as no immune rejection, a natural feel, and no implant-related complications (such as capsular contracture or implant leakage). This method has become one of the important surgical procedures in the field of breast plastic surgery and breast reconstruction. Water-assisted liposuction is a key technology applied to fat harvesting in recent years. A representative system is the Body-jet system developed by Human Med in Germany, which uses a spiral water jet technique. It injects a pulsed, low-pressure water stream in a micro-fan shape into the body, precisely targeting the target tissue with pressurized water to selectively separate fat cells and gently break down fat tissue and surrounding connective tissue. Compared to traditional negative pressure liposuction, water-assisted liposuction reduces mechanical damage to fat particles, blood vessels, and nerve tissue, resulting in finer and more uniform fat particles. Clinical studies have shown that fat tissue obtained using water-assisted liposuction has higher cell activity and mesenchymal stem cell proliferation capacity. In addition, some studies have attempted to combine platelet-rich plasma or adipose-derived stem cell matrix vascular components with fat grafting to further improve the survival rate of fat grafts.

[0003] However, current water-assisted autologous fat transfer breast augmentation techniques still have significant limitations. First, the postoperative survival rate of fat grafts exhibits considerable individual variability and instability. Literature reports a wide range of survival rates after autologous fat transfer, from approximately 25% to 80%; a meta-analysis focusing on abdominal donor sites showed an average fat survival rate of approximately 56% at 12 months post-surgery. This instability makes the predictability of surgical outcomes poor, requiring some patients to undergo secondary or even multiple surgeries to achieve the desired breast augmentation results. Second, the operating parameters of existing water-assisted systems (such as water pressure, flow rate, injection speed, and single injection volume) are typically fixed after preoperative settings. There is a lack of real-time sensing of the local tissue condition during the procedure, and the system cannot adaptively adjust to dynamic changes in the tissue microenvironment (such as local tissue oxygen saturation and interstitial pressure). Excessive injection pressure or insufficient local tissue perfusion can lead to serious complications such as fat cell damage, ischemic necrosis, and even fat embolism. Furthermore, current technologies for adding bioactive additives (such as PRP and SVF) mostly involve preoperative one-time mixing at a fixed ratio, lacking a dynamic regulation mechanism that correlates with the real-time state of the tissue microenvironment during surgery. In addition, while some studies have proposed placing tissue impedance sensors at the injection needle tip to distinguish between adipose and muscle tissue, their function is limited to providing prompts to the operator and has not yet formed an automated safety interlock control with the injection execution system. In summary, current water-assisted autologous fat grafting for breast augmentation still has significant shortcomings in terms of real-time intraoperative sensing, adaptive closed-loop control, and the synergistic regulation of physical parameters and bioactive substance supply. There is an urgent need for a fat grafting system and method that can achieve intelligent, closed-loop control. Summary of the Invention

[0004] In order to solve the problems of the prior art, the present invention provides a water-assisted autologous fat transplantation method for breast augmentation.

[0005] To solve the above-mentioned technical problems, the present invention is achieved through the following technical solution: Firstly, a water-assisted autologous fat transfer breast augmentation method, comprising the following steps: Step S1: Obtain the patient's preoperative breast imaging data, including preoperative MRI images and / or preoperative CT images; Step S2: Input the preoperative imaging data into the trained prognostic prediction model to generate an initial surgical plan. The initial surgical plan includes donor site selection, expected amount of fat to be harvested, initial mixing ratio of bioactive additives, distribution of breast injection target area, and estimated injection volume distribution. Step S3: During the surgery, multi-dimensional surgical parameters are collected in real time by a multi-modal sensor group set on the water-assisted liposuction component and / or fat injection component. The multi-dimensional surgical parameters include water pressure parameters, flow rate parameters, injection depth parameters, injection pressure parameters, tissue electrical impedance parameters, local tissue oxygen saturation parameters, and temperature parameters. Step S4: Input the multidimensional surgical parameters collected in real time into the prognosis prediction model to update the predicted survival rate and / or complication risk level after fat grafting in real time; Step S5: Based on the prediction results of step S4 and the real-time changing trend of the multidimensional surgical parameters, adjust the water pressure parameters and / or flow rate parameters of the hydrodynamic liposuction component, the mixing ratio parameters of the bioactive additives, and the injection speed parameters and / or single injection volume parameters of the fat injection component through the closed-loop control module. Step S6: The prediction results, the real-time values ​​and trends of the multidimensional surgical parameters, and the adjustment information of the closed-loop control module are displayed to the operator in real time through the output module.

[0006] In one specific implementation of the first aspect, in step S5: When the real-time value of the injection pressure parameter exceeds 150% of the baseline injection pressure value and the real-time value of the local tissue oxygen saturation parameter is lower than 70% of the baseline value, the closed-loop control module reduces the injection speed of the fat injection component to 30%~50% of the current value, and at the same time increases the mixing ratio of the bioactive additive from the initial value to 15%~25% of the fat tissue volume; After the bioactive additive is injected, the injection pressure parameter and the local tissue oxygen saturation parameter are continuously monitored. When the injection pressure parameter falls below 150% of the baseline injection pressure value and the local tissue oxygen saturation parameter rises above 70% of the baseline value, the closed-loop control module gradually restores the injection speed to 80%~100% of the initial plan and adjusts the mixing ratio of the bioactive additive back to 5%~10% of the adipose tissue volume. The reference injection pressure value is the reference injection pressure value when the fat injection component is placed in the air outside the body and advanced at a set speed; the reference value is the local tissue oxygen saturation value of the breast target site in the patient's preoperative resting state.

[0007] In one specific embodiment of the first aspect, in step S3, the multimodal sensor group further includes a tissue electrical impedance sensor disposed at the tip of the injection needle, and the method further includes: The tissue electrical impedance sensor detects the tissue electrical impedance value at the location of the injection needle tip in real time. When the detected tissue impedance value falls within the range of 30Ω to 80Ω, it is determined that the injection needle tip is located in the pectoralis major muscle tissue, an alarm signal is issued and the injection is automatically stopped; When the detected tissue impedance value falls within the range of 100Ω to 300Ω, it is determined that the injection needle tip is located within the breast adipose tissue, and injection can continue.

[0008] In one specific implementation of the first aspect, the prognostic prediction model is a neural network model based on transfer learning. The neural network model includes an input layer, three hidden layers, and an output layer. The first hidden layer contains 128 neurons, the second hidden layer contains 64 neurons, and the third hidden layer contains 32 neurons. Each hidden layer uses the ReLU activation function. The input of the model includes at least two of the following: preoperative MRI image data, intraoperative injection pressure time-series data, intraoperative local tissue oxygen saturation time-series data, and postoperative infrared thermography data. The output includes a predicted fat survival rate value 3 months after surgery.

[0009] In one specific implementation of the first aspect, step S2, the generation of the initial surgical plan further includes: establishing a three-dimensional model of the patient's breast based on the preoperative image data, simulating the breast morphology effect after fat injection on the three-dimensional model, and optimizing the injection volume and injection distribution plan accordingly.

[0010] In one specific implementation of the first aspect, step S7 is further included: uploading the patient's preoperative imaging data, real-time multidimensional parameters during the operation, and postoperative follow-up data to a cloud data platform. The cloud data platform performs statistical analysis on the surgical data of multiple patients, identifies key characteristic parameters that affect the survival rate of fat grafting, and updates the prognostic prediction model accordingly.

[0011] In one specific embodiment of the first aspect, the bioactive additive comprises platelet-rich plasma and / or adipose-derived stem cell matrix vascular components; the platelet-rich plasma is prepared by a two-stage centrifugation method, with a first centrifugation speed of 2000 rpm and a centrifugation time of 10 minutes, and a second centrifugation speed of 3000 rpm and a centrifugation time of 15 minutes, resulting in a platelet concentration ≥1×10⁻⁶ in the prepared platelet-rich plasma. 6 The adipose-derived stem cell matrix vascular component was extracted by collagenase digestion, and the proportion of live cells in the obtained matrix vascular component was ≥85%.

[0012] In one specific embodiment of the first aspect, the hydrodynamic liposuction assembly includes a hydrodynamic liposuction needle, a pressure pump, and a flow regulating valve. The closed-loop control module controls the hydrodynamic parameters by adjusting the output pressure of the pressure pump and / or the opening degree of the flow regulating valve. The fat injection assembly includes an injection needle, an electric propeller, and a micro-injection pump. The closed-loop control module controls the injection speed and single injection volume by adjusting the propulsion speed of the electric propeller and / or the output flow rate of the micro-injection pump.

[0013] Secondly, a water-assisted autologous fat transfer breast augmentation intelligent system includes: The fat harvesting unit includes a hydrodynamic liposuction assembly, which is used to separate and harvest autologous fat tissue from the patient's donor area using pressurized water flow. A fat processing unit, connected to the fat collection unit, is used to purify the collected fat tissue. A fat injection unit, connected to the fat processing unit, is used to inject purified fat tissue into the target area of ​​the patient's breast. A bioactive substance supply unit, connected to the fat processing unit or the fat injection unit, is used to provide bioactive additives, the bioactive additives including platelet-rich plasma and / or fat-derived stem cell matrix vascular components. The sensing and monitoring module includes a multimodal sensor group disposed on the fat acquisition unit and / or the fat injection unit, for real-time acquisition of multidimensional parameters during the surgical process, including water pressure parameters, flow rate parameters, injection needle tip advancement depth parameters, injection pressure parameters, tissue electrical impedance parameters, local tissue oxygen saturation parameters, and temperature parameters. The data acquisition and storage module is connected to the sensing and monitoring module and is used to receive and store the multidimensional parameters, as well as the patient's preoperative imaging data and postoperative follow-up data. The intelligent analysis and decision-making module, connected to the data acquisition and storage module, includes a prognosis prediction submodule and a scheme optimization submodule. The prognosis prediction submodule is equipped with a trained machine learning model for predicting the survival rate and / or complication risk after fat grafting based on the multidimensional parameters and / or the preoperative imaging data. The scheme optimization submodule is used to generate an optimized surgical plan based on the prediction results of the prognosis prediction submodule, including parameters such as injection speed, single injection volume, and bioactive additive mixing ratio. A closed-loop control module is connected to the intelligent analysis and decision-making module, the fat collection unit, the fat processing unit, the bioactive substance supply unit, and the fat injection unit, respectively. It is used to adjust the water pressure parameters and / or flow rate parameters of the hydrodynamic liposuction component, the purification parameters of the fat processing unit, the mixing ratio parameters of the bioactive substance supply unit, and the injection speed parameters and / or single injection volume parameters of the fat injection unit according to the optimized surgical plan output by the intelligent analysis and decision-making module. And an output module for displaying the prediction results and / or the optimized surgical plan and / or real-time monitoring information to the operator.

[0014] In one specific implementation of the second aspect, when the injection pressure value collected by the sensing and monitoring module exceeds 150% of the baseline injection pressure value and the local tissue oxygen saturation value is lower than 70% of the baseline value, the intelligent analysis and decision-making module simultaneously outputs a first injection speed adjustment command and a first mixing ratio adjustment command. The first injection speed adjustment command causes the closed-loop control module to reduce the injection speed of the fat injection unit to 30%~50% of the current value; The first mixing ratio adjustment command causes the closed-loop control module to control the bioactive substance supply unit to increase the mixing ratio of the bioactive additive to 15%~25% of the adipose tissue volume; When the bioactive additive is injected into the breast target site, and the sensing and monitoring module detects that the injection pressure value drops below 150% of the baseline injection pressure value and the local tissue oxygen saturation value rises above 70% of the baseline value, the intelligent analysis and decision-making module outputs a second injection speed adjustment command, causing the closed-loop control module to gradually restore the injection speed of the fat injection unit to 80%~100% of the initial plan, and outputs a second mixing ratio adjustment command, causing the closed-loop control module to adjust the mixing ratio of the bioactive additive back to 5%~10% of the fat tissue volume; The reference injection pressure value is the reference injection pressure value of the fat injection unit when the injection needle tip is placed in the external air and advanced at a set speed; the reference value is the local tissue oxygen saturation value of the breast target site in the patient's preoperative resting state.

[0015] The beneficial effects of this invention are as follows: 1. This invention, by constructing a multi-level intelligent system architecture consisting of a fat harvesting unit, a fat processing unit, a fat injection unit, a bioactive substance supply unit, a sensing and monitoring module, a data acquisition and storage module, an intelligent analysis and decision-making module, a closed-loop control module, and an output module, elevates hydrodynamic-assisted fat grafting surgery from a traditional purely mechanical operation to an intelligent surgical paradigm integrating real-time perception, intelligent analysis, and automatic decision-making. The collaborative configuration of multimodal sensors in the sensing and monitoring module enables real-time, multi-dimensional perception of water pressure, flow rate, injection pressure, tissue impedance, local tissue oxygen saturation, and temperature parameters during the surgery, providing a multi-dimensional data foundation for subsequent intelligent decision-making. The prognostic prediction model based on transfer learning in the intelligent analysis and decision-making module can dynamically predict the survival rate and complication risk after fat grafting based on real-time collected multi-dimensional parameters, breaking the limitations of traditional surgery that relies on subjective judgment based on the surgeon's personal experience. Most importantly, the closed-loop control module combines the prognostic prediction results with the actual sensor data to form a real-time closed-loop control loop of "perception-analysis-decision-execution". This enables the system to adaptively and dynamically adjust the hydrodynamic parameters, injection speed parameters, single injection volume parameters and bioactive additive mixing ratio parameters during the surgical process, fundamentally solving the technical problem of fixed parameters and inability to dynamically adjust according to the real-time state of the tissue in traditional fat transplantation surgery. 2. The proposed control logic of this invention achieves synergistic linkage between physical parameter adjustment and bioactive substance supply: When the sensing module detects that the injection pressure exceeds the threshold and the local tissue oxygen saturation decreases, the system simultaneously executes dual commands to reduce the injection speed and increase the mixing ratio of bioactive additives (platelet-rich plasma and / or adipose-derived stem cell matrix vascular components), thus coupling surgical operation parameters with the tissue microenvironment repair capacity in real time. When the tissue microenvironment parameters recover to the safe threshold range, the system automatically executes reverse adjustment to restore the injection speed and adjust the bioactive additive ratio. This mechanism overcomes the technical limitations of existing technologies where physical operation parameters and bioactive substance addition are independent and lack dynamic correlation. Through a positive feedback regulation cycle of "physical reduction—biological repair—state perception—parameter correction," it effectively reduces the risk of fat cell damage and postoperative necrosis caused by excessive local injection pressure or tissue ischemia. On this basis, the real-time identification of the tissue type where the injection needle tip is located by the tissue impedance sensor and the safety interlock mechanism further enhance the safety boundary of the surgery and effectively prevent serious complications that may be caused by accidental injection into non-adipose tissue layers. The cloud-based data platform and big data analysis module aggregate and analyze surgical data from multiple patients and incrementally update the model, enabling the accuracy and universality of the prognostic prediction model to continuously improve with the accumulation of clinical data. This forms a data-driven self-optimization capability for surgical plans, providing systematic technical support for the safety, predictability of results, and long-term stability of fat grafting breast augmentation surgery. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the surgical method flow of the present invention.

[0017] Figure 2 This is a schematic diagram of the closed-loop linkage control logic flow of the present invention. Detailed Implementation

[0018] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0019] like Figures 1 to 2 This illustrates a water-assisted autologous fat transfer method for breast augmentation.

[0020] I. System Overall Architecture like Figure 1 As shown, the intelligent system for hydrodynamic-assisted autologous fat transplantation breast augmentation of the present invention includes: a fat collection unit, a fat processing unit, a fat injection unit, a bioactive substance supply unit, a sensing and monitoring module, a data acquisition and storage module, an intelligent analysis and decision-making module, a closed-loop control module, and an output module.

[0021] The fat harvesting unit includes a water-assisted liposuction assembly, which comprises a water-assisted liposuction needle, a pressure pump, and a flow control valve. The water-assisted liposuction assembly separates and harvests autologous fat tissue from the patient's donor area (preferably the abdomen, inner thigh, or buttocks) using pressurized water flow. The water outlets of the water-assisted liposuction needle are arranged in a ring, and the pressurized water flow is sprayed in a fan shape, with the water jet direction forming a 25°–35° angle with the axis of the liposuction needle to facilitate gentle separation of the fat tissue from the surrounding connective tissue. The water pressure of the water-assisted liposuction assembly is adjustable from 20 bar to 100 bar, preferably from 30 bar to 90 bar; the water flow rate is adjustable from 50 mL / min to 300 mL / min.

[0022] The fat processing unit is connected to the fat collection unit and is used to purify the collected adipose tissue. The fat processing unit includes a centrifuge and / or a rinsing filter. The centrifuge speed is adjustable from 800 rpm to 3000 rpm, and the centrifugation time is adjustable from 1 minute to 10 minutes. The rinsing filter uses physiological saline to repeatedly rinse the adipose tissue to remove blood, swelling fluid, and free lipids.

[0023] The fat injection unit connects to the fat processing unit and is used to inject purified fat tissue into the target area of ​​the patient's breast. The fat injection unit includes an injection needle, a motorized actuator, and a microinfusion pump. The injection needle has an outer diameter of 1.2 mm to 2.5 mm and a blunt tip to reduce tissue damage. The motorized actuator's propulsion speed is adjustable from 0.1 mL / s to 2.0 mL / s, and the microinfusion pump's output flow rate is adjustable from 0.05 mL / min to 10 mL / min.

[0024] The bioactive substance supply unit is connected to the fat processing unit or fat injection unit to provide bioactive additives. These bioactive additives include platelet-rich plasma (PRP) and / or adipose-derived stem cell stromal vascular component (SVF).

[0025] The sensing and monitoring module includes a multimodal sensor array mounted on the fat harvesting unit and / or fat injection unit for real-time acquisition of multidimensional parameters during the surgical procedure. These multidimensional parameters include: water pressure (bar), flow rate (mL / min), needle tip depth (mm), injection pressure (mmHg or kPa), tissue electrical impedance (Ω), local tissue oxygen saturation (%), and temperature (°C).

[0026] The data acquisition and storage module is connected to the sensor monitoring module to receive and store multidimensional parameters, as well as the patient's preoperative imaging data and postoperative follow-up data.

[0027] The intelligent analysis and decision-making module is connected to the data acquisition and storage module, and includes a prognosis prediction submodule and a treatment plan optimization submodule. The prognosis prediction submodule is equipped with a trained machine learning model to predict postoperative survival rate and / or complication risk after fat grafting based on multidimensional parameters and / or preoperative imaging data. The treatment plan optimization submodule generates an optimized surgical plan based on the prediction results from the prognosis prediction submodule, including parameters such as injection rate, single injection volume, and the mixing ratio of bioactive additives.

[0028] The closed-loop control module is connected to the intelligent analysis and decision-making module, the fat collection unit, the fat processing unit, the bioactive substance supply unit, and the fat injection unit, respectively. It is used to adjust the water pressure parameters and / or flow parameters of the water-assisted liposuction component, the purification parameters of the fat processing unit, the mixing ratio parameters of the bioactive substance supply unit, and the injection speed parameters and / or single injection volume parameters of the fat injection unit according to the optimized surgical plan output by the intelligent analysis and decision-making module.

[0029] The output module is used to display predicted results and / or optimized surgical plans and / or real-time monitoring information to the operator. The output module includes a display interface that visually presents real-time surgical parameter curves, real-time identification of the tissue type where the injection needle tip is located, a three-dimensional simulation of fat distribution, a simulation of postoperative results, and predicted fat survival rates.

[0030] (II) Specific Implementation Methods of the Sensing and Monitoring Module The sensing and monitoring module includes a multimodal sensor group installed on the fat collection unit and / or fat injection unit. The multimodal sensor group includes: a water pressure sensor (installed at the pressure pump outlet of the water-assisted liposuction assembly, range 0~150 bar, accuracy ±0.5% FS), a flow sensor (installed in the tubing of the water-assisted liposuction assembly, range 0~500 mL / min, accuracy ±2%), a displacement sensor (installed on the electric actuator of the fat injection unit, used to detect the depth of needle tip insertion, accuracy ±0.5 mm), a pressure sensor (installed at the needle hub of the fat injection unit, used to detect real-time injection pressure during injection, range 0~200 mmHg, accuracy ±1 mmHg), and a tissue electrical impedance sensor (installed at the needle tip, used to detect the tissue electrical impedance at the needle tip location, frequency range 1 kHz~100 kHz). kHz, accuracy ±5%), pulse oximeter (set on the surface of the patient's breast target area to detect local tissue oxygen saturation, range 0~100%, accuracy ±2%), and temperature sensor (set at the injection needle tip and / or the surface of the breast target area, range 0~50℃, accuracy ±0.1℃).

[0031] In a preferred embodiment, the sensing and monitoring module further includes an image guidance submodule. The image guidance submodule includes an ultrasound probe and / or an infrared thermal imaging camera. The ultrasound probe has a frequency of 7.5 MHz to 15 MHz and is used to acquire real-time tissue image information of the target site during fat harvesting and / or fat injection, and to transmit the tissue image information to the intelligent analysis and decision-making module. The infrared thermal imaging camera has a temperature resolution of 0.05℃ and a spatial resolution of 320×240 pixels, and is used to monitor blood perfusion and temperature distribution at the breast target site in real time.

[0032] In a preferred embodiment, the sensing and monitoring module further includes a tissue type identification submodule disposed at the injection needle tip of the fat injection unit. The tissue type identification submodule distinguishes between adipose tissue and muscle tissue by detecting tissue impedance. The detection frequency of the tissue impedance sensor is preferably 10 kHz to 50 kHz. Within this frequency range, the impedance value of adipose tissue is approximately 100 Ω to 300 Ω, and the impedance value of muscle tissue is approximately 30 Ω to 80 Ω. When the detected tissue impedance value falls within the 30 Ω to 80 Ω range, it is determined that the injection needle tip is located within muscle tissue; when the detected tissue impedance value falls within the 100 Ω to 300 Ω range, it is determined that the injection needle tip is located within adipose tissue. When the injection needle tip is detected to have entered non-adipose tissue, the tissue type identification submodule issues an alarm signal (including a visual alarm and / or an audible alarm), and the closed-loop control module automatically stops the injection.

[0033] The signals collected by the sensors are processed by the signal conditioning circuit (including amplification, filtering and analog-to-digital conversion) and then transmitted to the data acquisition and storage module at a sampling frequency of 100Hz~1000Hz.

[0034] (III) Specific Implementation Methods of the Intelligent Analysis and Decision-Making Module The intelligent analysis and decision-making module includes a prognostic prediction submodule and a scheme optimization submodule.

[0035] The prognostic prediction submodule is configured with a trained machine learning model to predict postoperative survival rate and / or complication risk after fat grafting based on multidimensional parameters and / or preoperative imaging data. In a preferred embodiment, the machine learning model is a neural network model based on transfer learning. The neural network model includes an input layer, three hidden layers, and an output layer. The first hidden layer contains 128 neurons, the second hidden layer contains 64 neurons, and the third hidden layer contains 32 neurons. Each hidden layer uses the ReLU activation function (f(x)=max(0,x)). The output layer uses the Sigmoid activation function (f(x)=1 / (1+e^(-x))) to output the predicted survival rate (a probability value between 0 and 1), or uses the Softmax activation function to output the complication risk level (low risk, medium risk, and high risk).

[0036] The input data for the model includes at least two of the following: (1) preoperative MRI image data (features such as breast volume, fat layer thickness, and breast tissue distribution extracted after three-dimensional reconstruction); (2) intraoperative injection pressure time series data (the time window is the injection pressure change curve within the past 30 seconds); (3) intraoperative local tissue oxygen saturation time series data (the time window is the oxygen saturation change curve within the past 30 seconds); (4) postoperative infrared thermal imaging data (temperature distribution changes from immediately after surgery to within 72 hours after surgery).

[0037] The model was trained using a transfer learning strategy: first, pre-training was performed using animal experimental data (such as postoperative survival data from rabbit and pig fat transplantation models), followed by fine-tuning using a small amount of human clinical data. The learning rate during pre-training was 1×10^(-3), and the learning rate during fine-tuning was 1×10^(-5). The optimizer used was the Adam optimizer (β1=0.9, β2=0.999), and the loss function used was either mean squared error loss (MSE Loss) for the survival prediction task or cross-entropy loss for the risk grading task. The ratio of the training set, validation set, and test set was 7:1.5:1.5. The model output included the predicted fat survival rate at 3 months post-surgery (expressed as a percentage) and / or the risk level of postoperative complications.

[0038] The surgical plan optimization submodule generates an optimized surgical plan based on the prediction results from the prognosis prediction submodule. The optimized surgical plan includes parameters for injection rate (mL / s), single injection volume (mL), and the mixing ratio of bioactive additives (%). The optimization submodule employs a multi-objective optimization algorithm, aiming to maximize fat survival rate and minimize complication risk. The optimization algorithm uses a non-dominated sorting genetic algorithm (NSGA-II) with an elitist strategy, a population size of 50, 100 generations, a crossover probability of 0.8, and a mutation probability of 0.1.

[0039] (iv) Specific implementation of the closed-loop control module The closed-loop control module is connected to the intelligent analysis and decision-making module, the fat acquisition unit, the fat processing unit, the bioactive substance supply unit, and the fat injection unit. The closed-loop control module uses a PID control algorithm (proportional-integral-derivative control) to precisely adjust various parameters.

[0040] Taking closed-loop control of injection speed as an example: The closed-loop control module receives the target injection speed value (setpoint SP) output by the intelligent analysis and decision-making module, and simultaneously receives the real-time injection speed value (process variable PV) fed back by the sensor monitoring module. It calculates the deviation e(t) = SP - PV, and then outputs the control signal u(t) = Kp·e(t) + Ki·∫e(t)dt + Kd·de(t) / dt according to the PID control law. Wherein, the proportional coefficient Kp = 0.8~1.2, the integral coefficient Ki = 0.1~0.5, and the derivative coefficient Kd = 0.01~0.05. The control signal u(t) is output to the electric actuator of the fat injection unit. Precise adjustment of the injection speed is achieved by regulating the actuator's drive voltage or pulse width modulation (PWM) duty cycle. The control cycle is 100 ms.

[0041] Taking closed-loop control of hydrodynamics as an example: the closed-loop control module controls hydrodynamic parameters by adjusting the output pressure of the pressure pump (adjustment range 20~100 bar) and / or the opening degree of the flow regulating valve (adjustment range 0~100%). The PID control parameters for water pressure are: Kp=1.0~1.5, Ki=0.2~0.8, Kd=0.02~0.08. The control cycle is 200 ms.

[0042] In a preferred embodiment, when the injection pressure value collected by the sensing monitoring module exceeds 150% of the baseline injection pressure value and the local tissue oxygen saturation value is lower than 70% of the baseline value, the intelligent analysis and decision-making module simultaneously outputs a first injection speed adjustment command and a first mixing ratio adjustment command. The baseline injection pressure value is the baseline injection pressure value (typically 5-15 mmHg) of the fat injection unit when the injection needle tip is placed in external air and advanced at a set speed. The baseline value is the local tissue oxygen saturation value of the breast target site in the patient's preoperative resting state (typically 95%-99%). The first injection speed adjustment command causes the closed-loop control module to reduce the injection speed of the fat injection unit to 30%-50% of the current value. The first mixing ratio adjustment command causes the closed-loop control module to control the bioactive substance supply unit to increase the mixing ratio of bioactive additives to 15%-25% of the fat tissue volume. When the bioactive additive is injected into the breast target site, the sensing and monitoring module detects that the injection pressure value has dropped to below 150% of the baseline injection pressure value and the local tissue oxygen saturation value has risen to above 70% of the baseline value. At this time, the intelligent analysis and decision-making module outputs a second injection speed adjustment command, which causes the closed-loop control module to gradually restore the injection speed of the fat injection unit to 80% to 100% of the initial plan, and outputs a second mixing ratio adjustment command, which causes the closed-loop control module to adjust the mixing ratio of the bioactive additive back to 5% to 10% of the fat tissue volume.

[0043] (v) Specific implementation method of bioactive substance supply unit The bioactive substance supply unit is connected to the fat processing unit or fat injection unit to provide bioactive additives. These bioactive additives include platelet-rich plasma (PRP) and / or adipose-derived stem cell stromal vascular component (SVF).

[0044] Preparation method of PRP: Collect peripheral venous blood from the patient (preferably 30 to 60 minutes before surgery, with a blood volume of 30 mL to 60 mL) and place it in a centrifuge tube containing sodium citrate anticoagulant (final concentration 0.31% to 0.38%). PRP is prepared using a two-stage centrifugation method. First centrifugation: Centrifugation force 800 g (approximately 2000 rpm), centrifugation time 10 minutes. After centrifugation, the blood separates into three layers: the upper layer is platelet-poor plasma (PPP), the middle layer is the white membrane layer (rich in platelets), and the lower layer is the erythrocyte layer. Carefully aspirate the upper and middle layers (white membrane layer) of plasma using a pipette and place them in another sterile centrifuge tube. Second centrifugation: Centrifugation force 1400 g (approximately 3000 rpm), centrifugation time 15 minutes. After centrifugation, discard the supernatant (approximately 3 / 4 volume), retaining approximately 4 mL to 6 mL of precipitate at the bottom, which is platelet-rich plasma (PRP). The platelet concentration in the prepared PRP is ≥1×10⁻⁶. 6 cells / μL (i.e. ≥1000×10⁻⁶) 9 / L). Platelet recovery rate in PRP >85%. Platelet concentration in PRP >5-fold (relative to whole blood platelet concentration).

[0045] SVF extraction method: Take purified adipose tissue (approximately 20 mL–50 mL) and place it in a sterile container. Rinse repeatedly three times with sterile phosphate-buffered saline (PBS) to remove residual blood and swelling fluid. Cut the adipose tissue into particles of 1 mm–2 mm. Add type I collagenase solution (concentration 0.1%–0.2%, dissolved in DMEM medium containing 2% bovine serum albumin), with the volume of collagenase solution being 1–2 times the volume of the adipose tissue. Digest in a 37°C constant temperature shaker for 30–60 minutes at a shaking speed of 100–150 rpm. After digestion, add an equal volume of DMEM medium containing 10% fetal bovine serum to terminate the digestion reaction. Centrifuge the digested mixture at 400 g for 5–10 minutes. After centrifugation, discard the supernatant (containing free lipids and collagenase solution); the precipitate is the stromal vascular component (SVF). The SVF precipitate was resuspended in sterile saline and filtered through a 100 μm cell sieve to remove undigested tissue debris. The resulting SVF contained ≥85% viable cells. SVF is a heterogeneous cell population containing various cell types, including adipose-derived mesenchymal stem cells, endothelial progenitor cells, and macrophages.

[0046] Method for mixing PRP and / or SVF with adipose tissue: Mix the prepared PRP or SVF with purified adipose tissue according to a predetermined ratio. Mixing is performed under sterile conditions using a two-way syringe connected to a three-way valve, injecting repeatedly (20-30 times) to ensure thorough mixing of the PRP / SVF with the adipose tissue. The mixed adipose-PRP / SVF complex is stored at 4°C and used within 2 hours of preparation.

[0047] (vi) Specific implementation methods of the surgical procedure The water-assisted autologous fat transplantation breast augmentation method of the present invention includes the following steps: Step S1: Acquire preoperative imaging data Preoperative breast imaging data was acquired from the patient. This included preoperative MRI and / or preoperative CT images. MRI scans used T1-weighted and T2-weighted sequences with a slice thickness of 1 mm–2 mm and an interslice spacing of 0 mm. CT scans used spiral CT with a slice thickness of 1 mm–1.5 mm. The image data was exported in DICOM format and imported into the intelligent analysis and decision-making module for 3D reconstruction. 3D reconstruction employed volume rendering technology to reconstruct a 3D model of the patient's breast, including skin contours, breast tissue, pectoralis major muscle, and ribs.

[0048] Step S2: Generate the initial surgical plan Preoperative imaging data is input into a trained prognostic prediction model to generate an initial surgical plan. The initial surgical plan includes donor site selection, estimated amount of fat to be harvested, initial mixing ratio of bioactive additives, breast injection target area distribution, and estimated injection volume distribution.

[0049] Donor site selection is based on a comprehensive assessment of the fat layer thickness and uniformity of fat distribution in various areas of the patient (abdomen, inner thigh, buttocks, etc.) from preoperative imaging data, with sites having a fat layer thickness ≥1.5 cm preferred as donor sites. The expected fat harvest volume is calculated as 1.5–2.0 times the patient's target breast volume (considering the absorption rate after fat grafting, typically 30%–50%). The initial mixing ratio of bioactive additives is 5%–10% of the fat tissue volume. Breast injection target area distribution is planned based on a three-dimensional breast model, including three layers: the subcutaneous layer, the retromammary space, and the inner layer of the pectoralis major muscle. The estimated injection volume distribution is optimized based on the tissue volume and vascular distribution of each target area.

[0050] In a preferred embodiment, the generation of the initial surgical plan further includes: establishing a three-dimensional model of the patient's breast based on preoperative imaging data, simulating the breast morphology effect after fat injection on the three-dimensional model, and optimizing the injection volume and distribution plan accordingly. The simulated injection uses the finite element analysis method to simulate the morphological changes of the breast after fat injection, with the goal of achieving a preset fullness and symmetry in the breast.

[0051] Step S3: Real-time acquisition of multidimensional surgical parameters During the procedure, multi-modal sensor arrays mounted on the water-assisted liposuction and / or fat injection components collect multi-dimensional surgical parameters in real time. These parameters include water pressure (bar), flow rate (mL / min), injection depth (mm), injection pressure (mmHg), tissue electrical impedance (Ω), local tissue oxygen saturation (%), and temperature (°C). Each sensor operates at a sampling frequency ranging from 100 Hz to 1000 Hz. The collected data is transmitted in real time to the data acquisition and storage module.

[0052] In a preferred embodiment, step S3 further includes real-time detection of the tissue impedance value at the injection needle tip location using a tissue impedance sensor disposed at the injection needle tip. When the detected tissue impedance value falls within the 30 Ω to 80 Ω range, it is determined that the injection needle tip is located within the pectoralis major muscle tissue; when the detected tissue impedance value falls within the 100 Ω to 300 Ω range, it is determined that the injection needle tip is located within the breast adipose tissue. The determination result is displayed on the output module in real time, and an alarm signal is issued and the injection is automatically stopped when the needle tip enters the muscle tissue.

[0053] Step S4: Update prediction results in real time Real-time acquired multidimensional surgical parameters are input into the prognostic prediction model to update the predicted survival rate and / or complication risk level after fat grafting in real time. The prediction update frequency is 1 Hz to 10 Hz (i.e., updated once every 0.1 seconds to 1 second). The prediction results are displayed on the output module in real time.

[0054] Step S5: Closed-loop control adjustment Based on the prediction results of step S4 and the real-time changing trends of multidimensional surgical parameters, the water pressure parameters and / or flow parameters of the hydrodynamic liposuction component, the mixing ratio parameters of bioactive additives, and the injection speed parameters and / or single injection volume parameters of the fat injection component are adjusted through the closed-loop control module.

[0055] Specifically, the closed-loop control module executes the following control logic: When the real-time value of the injection pressure parameter exceeds 150% of the baseline injection pressure value and the real-time value of the local tissue oxygen saturation parameter is lower than 70% of the baseline value, the closed-loop control module reduces the injection speed of the fat injection component to 30%–50% of the current value, while simultaneously increasing the mixing ratio of bioactive additives from the initial value to 15%–25% of the fat tissue volume. The baseline injection pressure value is the baseline injection pressure value of the fat injection component when the injection needle tip is placed in external air and advanced at a set speed (typically 5–15 mmHg). The baseline value is the local tissue oxygen saturation value of the breast target site in the patient's preoperative resting state (typically 95%–99%).

[0056] After the bioactive additive is injected, the injection pressure parameters and local tissue oxygen saturation parameters are continuously monitored.

[0057] When the injection pressure parameter drops to below 150% of the baseline injection pressure value and the local tissue oxygen saturation parameter rises to above 70% of the baseline value, the closed-loop control module will gradually restore the injection speed to 80%~100% of the initial plan and adjust the mixing ratio of bioactive additives back to 5%~10% of the adipose tissue volume.

[0058] Step S6: Real-time display The output module displays the predicted results, real-time values ​​and trends of multidimensional surgical parameters, and adjustment information of the closed-loop control module to the operator in real time. The display interface includes: real-time surgical parameter curves (horizontal axis is time, vertical axis is the value of each parameter), real-time identification of the tissue type where the injection needle tip is located (adipose tissue is marked in green, muscle tissue is marked in red and accompanied by an alarm), a three-dimensional simulation map of fat distribution (updating the fat distribution in the injection area in real time), a simulation map of postoperative effect (predicted postoperative breast shape based on the current injection volume), and predicted fat survival rate (displayed as a percentage with a confidence interval).

[0059] Step S7: Data Upload and Model Update Patients' preoperative imaging data, real-time multidimensional parameters during surgery, and postoperative follow-up data are uploaded to a cloud-based data platform. The cloud platform performs statistical analysis on the surgical data of multiple patients to identify key characteristic parameters affecting fat graft survival rates and updates the prognostic prediction model accordingly. Data uploads use encrypted transmission protocols (such as HTTPS or SFTP) to ensure patient data security. Model updates employ incremental learning, fine-tuning the model every 50 new data points, with an update cycle of 1-3 months.

[0060] IV. Examples Example 1 1. Patient Information Patient Zhang, female, 32 years old, height 165 cm, weight 58 kg. Chief complaint: congenital bilateral micromastia, desiring breast enlargement. Preoperative MRI showed: bilateral breast volumes of 180 mL (left) and 175 mL (right), with small breast tissue volume and a subcutaneous fat layer thickness of approximately 0.8 cm to 1.2 cm. The baseline oxygen saturation of the target breast tissue at rest was 97% preoperatively. The patient had no history of systemic diseases such as hypertension or diabetes, and no history of breast surgery.

[0061] 2. Preoperative preparation Thirty minutes before the procedure, 50 mL of peripheral venous blood was collected from the patient and placed in a centrifuge tube containing sodium citrate anticoagulant (final concentration 0.35%) for PRP preparation. The installation and debugging of the water-assisted autologous fat transfer breast augmentation intelligent system were completed preoperatively. The system includes: a water-assisted liposuction component (including a pressure pump, flow control valve, and disposable sterile liposuction treatment kit), a fat processing unit (including a centrifuge), a fat injection unit (including an electric propeller, micro-injection pump, and disposable sterile injection filling kit), a bioactive substance supply unit, a multimodal sensor group, a data acquisition and storage module, an intelligent analysis and decision-making module, a closed-loop control module, and an output module.

[0062] 3. PRP Preparation PRP was prepared using a two-stage centrifugation method. First centrifugation: 800 g, 10 min. The upper and middle layers of plasma were transferred to another sterile centrifuge tube. Second centrifugation: 1400 g, 15 min. The supernatant was discarded, retaining approximately 5 mL of the bottom precipitate. Hematology analysis showed that the platelet concentration in the PRP was 1.2 × 10⁻⁶. 6 Platelet count / μL was reduced to 88%, and the concentration was 6.2 times that of whole blood. PRP was stored at 4°C for later use.

[0063] 4. Preoperative image acquisition and surgical planning One day prior to surgery, the patient underwent a breast MRI scan. A 1.5T MRI scanner was used, employing T1-weighted and T2-weighted sequences with a slice thickness of 1.5 mm and an interslice spacing of 0 mm. The image data was exported in DICOM format and imported into the intelligent analysis and decision-making module for 3D reconstruction. After 3D reconstruction, the breast volumes were measured as follows: left breast 183 mL, right breast 178 mL (MRI measurements consistent with clinical assessment).

[0064] Preoperative MRI image data was input into a trained prognostic prediction model (a transfer learning-based neural network model containing an input layer, three hidden layers, and an output layer; the first hidden layer has 128 neurons, the second hidden layer has 64 neurons, and the third hidden layer has 32 neurons; each hidden layer uses the ReLU activation function) to generate an initial surgical plan. Donor sites: Abdomen (fat layer thickness approximately 2.0 cm) and inner thigh (fat layer thickness approximately 1.8 cm). Target volume increase: 150 mL increase in volume for each breast; Expected fat harvest: 225 mL per side (1.5 times the target volume gain, considering a postoperative absorption rate of approximately 33%). Initial mixing ratio of bioactive additives: 8% of adipose tissue volume; Target area distribution for breast injection: subcutaneous layer 30%, retromammary space 50%, inner layer of pectoralis major muscle 20%; Estimated injection volume distribution: 150 mL for the left breast and 150 mL for the right breast. After simulated injection on a 3D model, the predicted postoperative breast volume is 320 mL on the left and 315 mL on the right, with a full and natural breast shape and good symmetry.

[0065] 5. Surgical Procedure The patient was placed in a supine position with both upper limbs abducted. Routine disinfection and draping were performed. Local anesthesia combined with sedation was administered.

[0066] Fat Collection Stage: The abdomen was selected as the donor area. The water pressure of the water-assisted liposuction system was set to 50 bar (level 2 intensity), and the flow rate was set to 150 mL / min. Two 3 mm incisions were made in the abdomen, and the water-assisted liposuction cannula was inserted into the subcutaneous fat layer. The water outlets of the water-assisted liposuction cannula were arranged in a ring, and the pressurized water flow was sprayed in a fan shape, with the water flow direction at a 30° angle to the axis of the liposuction cannula. The water-assisted liposuction system was activated, and the gentle pressurized water flow separated the fat cells from the surrounding connective tissue, while the negative pressure drew the fat tissue into the collection container. During the liposuction process, the water pressure and flow rate sensors collected water pressure and flow rate data in real time and transmitted them to the data acquisition and storage module. The abdominal liposuction lasted approximately 15 minutes, and approximately 230 mL of fat tissue was collected. Subsequently, supplementary liposuction was performed on the inner thigh, with the water pressure set to 40 bar and the flow rate set to 120 mL / min. The liposuction lasted approximately 10 minutes, and approximately 50 mL of fat tissue was collected. A total of approximately 280 mL of fat tissue was collected.

[0067] Fat processing stage: The collected adipose tissue was transferred to the fat processing unit. It was rinsed three times repeatedly with physiological saline to remove blood, swelling fluid, and free lipids. Then, it was centrifuged at 1200 rpm for 3 minutes. After centrifugation, the adipose tissue separated into three layers: the upper layer was free lipids (oil layer), the middle layer was purified fat particles, and the lower layer was an aqueous layer containing blood and swelling fluid. The upper and lower layers were discarded, retaining approximately 240 mL of the middle layer of purified fat particles. The purified fat particles were divided into two portions: one portion (approximately 200 mL) was used for mixing with PRP, and the other portion (approximately 40 mL) was reserved for later use.

[0068] PRP Mixing Stage: Mix the prepared PRP (5 mL) with purified fat particles (approximately 200 mL) according to the preset ratio. Use a two-way syringe connected to a three-way valve to repeatedly inject 25 times to ensure thorough mixing of the PRP and adipose tissue. The resulting fat-PRP complex should contain 2.5% PRP (i.e., 2.5% of the adipose tissue volume). According to the initial protocol, the bioactive additive (PRP) should be mixed at 8% of the adipose tissue volume; therefore, additional PRP needs to be added to reach the target ratio. Prepare an additional 8 mL of PRP, add it in portions and mix thoroughly, resulting in a final fat-PRP complex containing 8% PRP of the adipose tissue volume (approximately 16 mL PRP / 200 mL adipose tissue). Store the fat-PRP complex at 4°C and use within 1.5 hours of preparation.

[0069] Fat injection stage: The fat-PRP complex is loaded into the syringe of the fat injection unit. A blunt-tipped needle with an outer diameter of 1.8 mm is used for injection. Following the "multi-point, multi-tunnel, multi-layer" injection method, the fat-PRP complex is evenly injected into both breasts.

[0070] During the injection, the multimodal sensor array collects various parameters in real time. The baseline value for the injection pressure sensor is 8 mmHg (the baseline injection pressure in extracorporeal air at a rate of 0.5 mL / s). The baseline value for the tissue oxygen saturation sensor is 97% (preoperative resting state).

[0071] After injection begins, administer injections at a rate of 0.5 mL / s, with a single injection volume of 0.8 mL. Injection depths are controlled in the subcutaneous layer (5–10 mm), the retromammary space (15–25 mm), and the inner layer of the pectoralis major muscle (25–35 mm). A tissue impedance sensor continuously monitors the tissue type at the needle tip location: in the subcutaneous layer and retromammary space, the tissue impedance value remains stable at 150 Ω–220 Ω (100 Ω–300 Ω in adipose tissue); when injecting into the inner layer of the pectoralis major muscle, the tissue impedance value drops to 50 Ω–70 Ω (30 Ω–80 Ω in muscle tissue), and the system issues a warning sound, but since the injection has not penetrated deep into the muscle tissue, the injection continues.

[0072] When approximately 80 mL was injected into the left breast, the real-time value of the injection pressure parameter rose to 13 mmHg (exceeding 150% of the baseline value of 8 mmHg, i.e., 12 mmHg), while the real-time value of the local tissue oxygen saturation parameter dropped to 65% (below 70% of the baseline value of 97%, i.e., 67.9%). The intelligent analysis and decision-making module detected the above dual triggering conditions and simultaneously output the first injection speed adjustment command and the first mixing ratio adjustment command.

[0073] The closed-loop control module executes the first injection speed adjustment command: reducing the injection speed of the fat injection unit from 0.5 mL / s to 0.2 mL / s (40% of the current value). Simultaneously, it executes the first mixing ratio adjustment command: controlling the bioactive substance supply unit to increase the PRP mixing ratio from 8% to 20% of the adipose tissue volume (i.e., supplementing 12 mL of PRP per 100 mL of adipose tissue). Additional PRP (approximately 10 mL) is injected into the mixing container through the bioactive substance supply unit, thoroughly mixed with the remaining fat-PRP complex, and then injected further.

[0074] Approximately 2 minutes after PRP injection, the sensor monitoring module detected that the injection pressure parameter had dropped to 10 mmHg (150% below the baseline value, i.e., 12 mmHg), and the local tissue oxygen saturation parameter had risen to 72% (above the baseline value of 70%, i.e., 67.9%). The intelligent analysis and decision-making module then output a second injection rate adjustment command and a second mixing ratio adjustment command.

[0075] The closed-loop control module executes the second injection speed adjustment command: gradually restoring the injection speed to 0.45 mL / s (90% of the initial protocol's 0.5 mL / s). It also executes the second mixing ratio adjustment command: reverting the PRP mixing ratio back to 8% of the adipose tissue volume (i.e., restoring the initial protocol ratio).

[0076] Continue injecting until a total of 150 mL has been injected into the left breast and 150 mL into the right breast, then stop the injection. The entire injection process takes approximately 45 minutes.

[0077] 6. Postoperative management After the injection, gently massage both breasts to distribute the fat evenly. Cover the surgical area with sterile dressings and wear an elastic bra. Post-operatively, routine antibiotics are given for 3 days to prevent infection.

[0078] 7. Intraoperative data recording Throughout the entire surgical procedure, the data acquisition and storage module recorded complete time-series parameter data:

[0079] 8. Postoperative follow-up One week post-surgery: Mild swelling in both breasts, with no obvious bruising or hematoma, and no fever, pain, or other discomfort.

[0080] One month post-surgery: The volume of both breasts was 295 mL on the left and 290 mL on the right. They looked natural, felt soft, and had no complications such as lumps or cysts.

[0081] Three months post-surgery: The volumes of both breasts were 310 mL on the left and 308 mL on the right, with a fat survival rate of approximately 85% (relative to the immediate post-operative volume). The patient was satisfied with the surgical results, with full and natural breast shape and good bilateral symmetry.

[0082] Six months post-surgery: The volume of both breasts remained stable at 315 mL on the left and 312 mL on the right, with a fat survival rate of approximately 88%. Follow-up MRI showed that the transplanted fat was evenly distributed and there were no complications such as cysts or calcification.

[0083] 9. Comparison of effects with existing technologies To verify the technical effectiveness of this invention, 30 control patients who underwent traditional water-assisted fat grafting breast augmentation (without intelligent sensing and closed-loop control) during the same period were selected for comparison. There were no statistically significant differences in baseline data such as age, BMI, and preoperative breast volume between the control group and the patients in this embodiment (P>0.05).

[0084] The comparison results are as follows:

[0085] The above comparative results show that the present invention effectively improves the survival rate of fat grafting, reduces the risk of complications, and improves surgical efficiency through the organic combination of intelligent sensing, real-time analysis, and closed-loop control.

[0086] 10. Data Upload and Model Update On the 7th day post-surgery, the complete surgical data of the patients in this embodiment (including preoperative MRI images, intraoperative real-time multidimensional parameter time-series data, PRP preparation parameters, surgical plan, postoperative follow-up data, etc.) was uploaded to the cloud data platform via encrypted transmission. The cloud data platform performed statistical analysis on the data of a total of 52 patients, including those in this embodiment, and identified key characteristic parameters affecting the survival rate of fat grafting, including: peak injection pressure (weight 0.25), minimum tissue oxygen saturation (weight 0.30), PRP mixing ratio (weight 0.20), single injection volume (weight 0.15), and injection speed (weight 0.10). Based on the above analysis results, the cloud data platform incrementally learned and updated the prognostic prediction model, with a model update cycle of 2 months.

[0087] The water-assisted autologous fat grafting method and its intelligent system provided by this invention can be widely applied in autologous fat grafting surgeries in the field of plastic surgery, especially suitable for breast augmentation, breast reconstruction, and facial fat grafting. This invention, through intelligent sensing and closed-loop control technology, effectively improves the survival rate of transplanted fat and reduces the risk of postoperative complications, demonstrating promising clinical application prospects and industrialization value.

[0088] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A water-assisted autologous fat transfer breast augmentation method, characterized in that, Includes the following steps: Step S1: Obtain the patient's preoperative breast imaging data, including preoperative MRI images and / or preoperative CT images; Step S2: Input the preoperative imaging data into the trained prognostic prediction model to generate an initial surgical plan. The initial surgical plan includes donor site selection, expected amount of fat to be harvested, initial mixing ratio of bioactive additives, distribution of breast injection target area, and estimated injection volume distribution. Step S3: During the surgery, multi-dimensional surgical parameters are collected in real time by a multi-modal sensor group set on the water-assisted liposuction component and / or fat injection component. The multi-dimensional surgical parameters include water pressure parameters, flow rate parameters, injection depth parameters, injection pressure parameters, tissue electrical impedance parameters, local tissue oxygen saturation parameters, and temperature parameters. Step S4: Input the multidimensional surgical parameters collected in real time into the prognosis prediction model to update the predicted survival rate and / or complication risk level after fat grafting in real time; Step S5: Based on the prediction results of step S4 and the real-time changing trend of the multidimensional surgical parameters, adjust the water pressure parameters and / or flow rate parameters of the hydrodynamic liposuction component, the mixing ratio parameters of the bioactive additives, and the injection speed parameters and / or single injection volume parameters of the fat injection component through the closed-loop control module. Step S6: The prediction results, the real-time values ​​and trends of the multidimensional surgical parameters, and the adjustment information of the closed-loop control module are displayed to the operator in real time through the output module.

2. The method according to claim 1, characterized in that, In step S5: When the real-time value of the injection pressure parameter exceeds 150% of the baseline injection pressure value and the real-time value of the local tissue oxygen saturation parameter is lower than 70% of the baseline value, the closed-loop control module reduces the injection speed of the fat injection component to 30%~50% of the current value, and at the same time increases the mixing ratio of the bioactive additive from the initial value to 15%~25% of the fat tissue volume; After the bioactive additive is injected, the injection pressure parameter and the local tissue oxygen saturation parameter are continuously monitored. When the injection pressure parameter falls below 150% of the baseline injection pressure value and the local tissue oxygen saturation parameter rises above 70% of the baseline value, the closed-loop control module gradually restores the injection speed to 80%~100% of the initial plan and adjusts the mixing ratio of the bioactive additive back to 5%~10% of the adipose tissue volume. The reference injection pressure value is the reference injection pressure value when the fat injection component is placed in the air outside the body and advanced at a set speed; the reference value is the local tissue oxygen saturation value of the breast target site in the patient's preoperative resting state.

3. The method according to claim 1, characterized in that, In step S3, the multimodal sensor group further includes a tissue electrical impedance sensor disposed at the tip of the injection needle, and the method further includes: The tissue electrical impedance sensor detects the tissue electrical impedance value at the location of the injection needle tip in real time. When the detected tissue impedance value falls within the range of 30Ω to 80Ω, it is determined that the injection needle tip is located in the pectoralis major muscle tissue, an alarm signal is issued and the injection is automatically stopped; When the detected tissue impedance value falls within the range of 100Ω to 300Ω, it is determined that the injection needle tip is located within the breast adipose tissue, and injection can continue.

4. The method according to claim 1, characterized in that, The prognostic prediction model is a neural network model based on transfer learning. The neural network model includes an input layer, three hidden layers, and an output layer. The first hidden layer contains 128 neurons, the second hidden layer contains 64 neurons, and the third hidden layer contains 32 neurons. Each hidden layer uses the ReLU activation function. The input of the model includes at least two of the following: preoperative MRI image data, intraoperative injection pressure time-series data, intraoperative local tissue oxygen saturation time-series data, and postoperative infrared thermography data. The output includes the predicted fat survival rate at 3 months postoperatively.

5. The method according to claim 1, characterized in that, In step S2, the generation of the initial surgical plan further includes: establishing a three-dimensional model of the patient's breast based on the preoperative image data, simulating the breast morphology effect after fat injection on the three-dimensional model, and optimizing the injection volume and injection distribution plan accordingly.

6. The method according to claim 1, characterized in that, The procedure also includes step S7: uploading the patient's preoperative imaging data, real-time multidimensional parameters during the operation, and postoperative follow-up data to a cloud data platform. The cloud data platform performs statistical analysis on the surgical data of multiple patients, identifies key characteristic parameters that affect the survival rate of fat grafting, and updates the prognostic prediction model accordingly.

7. The method according to claim 1, characterized in that, The bioactive additives include platelet-rich plasma and / or adipose-derived stem cell matrix vascular components; the platelet-rich plasma is prepared by a two-stage centrifugation method, with the first centrifugation speed at 2000 rpm and a centrifugation time of 10 minutes, and the second centrifugation speed at 3000 rpm and a centrifugation time of 15 minutes, resulting in a platelet concentration ≥1×10⁻⁶ in the prepared platelet-rich plasma. 6 The adipose-derived stem cell matrix vascular component was extracted by collagenase digestion, and the proportion of live cells in the obtained matrix vascular component was ≥85%.

8. The method according to claim 1, characterized in that, The hydrodynamic liposuction assembly includes a hydrodynamic liposuction needle, a pressure pump, and a flow regulating valve. The closed-loop control module controls the hydrodynamic parameters by adjusting the output pressure of the pressure pump and / or the opening of the flow regulating valve. The fat injection assembly includes an injection needle, an electric propeller, and a micro-injection pump. The closed-loop control module controls the injection speed and single injection volume by adjusting the propulsion speed of the electric propeller and / or the output flow rate of the micro-injection pump.

9. A water-assisted autologous fat transfer breast augmentation intelligent system, characterized in that, include: The fat harvesting unit includes a hydrodynamic liposuction assembly, which is used to separate and harvest autologous fat tissue from the patient's donor area using pressurized water flow. A fat processing unit, connected to the fat collection unit, is used to purify the collected fat tissue. A fat injection unit, connected to the fat processing unit, is used to inject purified fat tissue into the target area of ​​the patient's breast. A bioactive substance supply unit, connected to the fat processing unit or the fat injection unit, is used to provide bioactive additives, the bioactive additives including platelet-rich plasma and / or fat-derived stem cell matrix vascular components. The sensing and monitoring module includes a multimodal sensor group disposed on the fat acquisition unit and / or the fat injection unit, for real-time acquisition of multidimensional parameters during the surgical process, including water pressure parameters, flow rate parameters, injection needle tip advancement depth parameters, injection pressure parameters, tissue electrical impedance parameters, local tissue oxygen saturation parameters, and temperature parameters. The data acquisition and storage module is connected to the sensing and monitoring module and is used to receive and store the multidimensional parameters, as well as the patient's preoperative imaging data and postoperative follow-up data. The intelligent analysis and decision-making module, connected to the data acquisition and storage module, includes a prognosis prediction submodule and a scheme optimization submodule. The prognosis prediction submodule is equipped with a trained machine learning model for predicting the survival rate and / or complication risk after fat grafting based on the multidimensional parameters and / or the preoperative imaging data. The scheme optimization submodule is used to generate an optimized surgical plan based on the prediction results of the prognosis prediction submodule, including parameters such as injection speed, single injection volume, and bioactive additive mixing ratio. A closed-loop control module is connected to the intelligent analysis and decision-making module, the fat collection unit, the fat processing unit, the bioactive substance supply unit, and the fat injection unit, respectively. It is used to adjust the water pressure parameters and / or flow rate parameters of the hydrodynamic liposuction component, the purification parameters of the fat processing unit, the mixing ratio parameters of the bioactive substance supply unit, and the injection speed parameters and / or single injection volume parameters of the fat injection unit according to the optimized surgical plan output by the intelligent analysis and decision-making module. And an output module for displaying the prediction results and / or the optimized surgical plan and / or real-time monitoring information to the operator.

10. The system according to claim 9, characterized in that, When the injection pressure value collected by the sensing and monitoring module exceeds 150% of the baseline injection pressure value and the local tissue oxygen saturation value is lower than 70% of the baseline value, the intelligent analysis and decision-making module simultaneously outputs a first injection speed adjustment command and a first mixing ratio adjustment command. The first injection speed adjustment command causes the closed-loop control module to reduce the injection speed of the fat injection unit to 30%~50% of the current value; The first mixing ratio adjustment command causes the closed-loop control module to control the bioactive substance supply unit to increase the mixing ratio of the bioactive additive to 15%~25% of the adipose tissue volume; When the bioactive additive is injected into the breast target site, and the sensing and monitoring module detects that the injection pressure value drops below 150% of the baseline injection pressure value and the local tissue oxygen saturation value rises above 70% of the baseline value, the intelligent analysis and decision-making module outputs a second injection speed adjustment command, causing the closed-loop control module to gradually restore the injection speed of the fat injection unit to 80%~100% of the initial plan, and outputs a second mixing ratio adjustment command, causing the closed-loop control module to adjust the mixing ratio of the bioactive additive back to 5%~10% of the fat tissue volume; The reference injection pressure value is the reference injection pressure value of the fat injection unit when the injection needle tip is placed in the external air and advanced at a set speed; the reference value is the local tissue oxygen saturation value of the breast target site in the patient's preoperative resting state.

Images