Multi-parameter real-time physiological index monitoring system for tumor patient

Abstract

The invention relates to the technical field of biomedical engineering, in particular to a tumor patient multi-parameter real-time physiological index monitoring system which comprises a cell sensor detection unit, a signal acquisition and preprocessing unit, a data processing and intelligent analysis unit, a communication and interaction unit and an energy and communication unit. High-sensitivity parallel detection of the concentration of circulating tumor cells and lactic acid in body fluid is achieved through the gold nanorod modified flexible microelectrode array and the microfluidic technology, and signal preprocessing is completed through low-noise amplification, filtering and high-precision analog-to-digital conversion; multi-center data collaborative optimization and personalized anomaly detection are realized by means of a federal learning driven privacy protection type hybrid intelligent analysis framework, and a low-power-consumption dynamic power supply system is constructed through piezoelectric energy collection, magnetic resonance wireless charging and a solid-state lithium battery. Therefore, accurate, safe and long-endurance real-time physiological index monitoring and clinical decision support is provided for tumor patients.

Description

Technical Field

[0001] The present invention relates to the technical field of biomedical engineering, and in particular to a multi-parameter real-time physiological index monitoring system for tumor patients. Background Art

[0002] "Cancer patients" refer to people who have been diagnosed with various types of tumors (benign or malignant). A tumor is a disease caused by the uncontrolled growth and spread of cells that can interfere with the normal function of tissues and organs. According to the nature of the tumor, it can be divided into the following categories: Benign tumors: These tumors usually grow slowly and do not invade surrounding tissues or metastasize to other parts of the body. Although they are generally not life-threatening, in some cases, they may require treatment if they are located in critical positions or grow large enough to affect the function of surrounding organs; Malignant tumors (cancer): These are those that are invasive, can invade adjacent tissues, and have the potential to spread through A tumor that spreads (metastasizes) to other parts of the body through the blood or lymphatic system. Malignant tumors are one of the diseases that seriously threaten human health. Depending on the primary site, they can be divided into various types, such as lung cancer, breast cancer, and gastric cancer. Borderline tumors: These tumors are between benign and malignant, with certain invasiveness but a low risk of metastasis. Tumor patients are diagnosed through imaging examinations (such as CT scans, MRI), laboratory tests (such as blood tests), and pathological examinations (such as biopsies). Treatment methods may include surgical resection, radiotherapy, chemotherapy, immunotherapy, targeted therapy, etc. The specific plan depends on the type and stage of the tumor and the patient's overall health. Cancer patients require real-time monitoring of key physiological parameters to assess disease progression, treatment efficacy, and complication risks, depending on their tumor type (e.g., the aggressiveness of malignant tumors and the potential risk of borderline tumors) and treatment options (e.g., fluctuations in physiological indicators caused by chemotherapy). For example, circulating tumor cell (CTC) counts: CTC levels in the blood of malignant tumor patients are directly related to the risk of metastasis, and detection sensitivity must be below 1 cell / mL to meet early warning needs; lactate concentration: The high glycolysis characteristics of the tumor microenvironment lead to elevated lactate levels, and dynamic monitoring of its concentration (normal range 0.5-2.2 mM) can reflect tumor metabolic activity and tissue hypoxia; multi-parameter collaborative analysis: A single indicator is difficult to fully assess the disease, and a comprehensive judgment must be made based on multi-dimensional data such as CTCs, lactate, and immune markers.

[0003] Existing tumor monitoring technologies have the following core flaws, making it difficult to meet the needs of clinical precision medicine:

[0004] Insufficient detection sensitivity: The detection limit of traditional electrochemical sensors for CTCs is approximately 10 cells / mL, and they are unable to simultaneously detect metabolic markers such as lactate, resulting in missed detection of early-stage tiny lesions.

[0005] Difficulty in parallel analysis of multiple parameters: Most existing equipment is for single-parameter detection and lacks multi-channel integration technology, making it impossible to achieve joint analysis at the biomolecular and cellular levels.

[0006] Lack of data privacy protection: Traditional medical data is stored centrally, which faces the risk of patient privacy leakage during multi-center collaboration and does not comply with regulations such as GDPR / HIPAA.

[0007] Limited system endurance: Implantable or wearable devices rely on disposable batteries that require frequent replacement, and wireless charging is inefficient, affecting the continuity of long-term monitoring.

[0008] Based on this, the present invention provides a multi-parameter real-time physiological indicator monitoring system for tumor patients to solve the above-mentioned technical problems. Summary of the Invention

[0009] The purpose of the present invention is to provide a multi-parameter real-time physiological indicator monitoring system for tumor patients. The gold nanorod-modified flexible microelectrode array and microfluidic technology of the present invention realize high-sensitivity parallel detection of circulating tumor cells and lactate concentrations in body fluids, complete signal preprocessing through low-noise amplification, filtering and high-precision analog-to-digital conversion, and realize multi-center data collaborative optimization and personalized anomaly detection with the help of a privacy-protected hybrid intelligent analysis framework driven by federated learning. Reliable data transmission and human-computer interaction are achieved by combining multi-mode communication protocols with dynamic power regulation, and a low-power dynamic power supply system is constructed through piezoelectric energy harvesting, magnetic resonance wireless charging and solid-state lithium batteries, thereby providing accurate, safe and long-lasting real-time physiological indicator monitoring and clinical decision support for tumor patients.

[0010] To achieve the above object, the present invention provides the following technical solutions:

[0011] The present invention provides a multi-parameter real-time physiological indicator monitoring system for tumor patients, comprising a cell sensor detection unit, a signal acquisition and preprocessing unit, a data processing and intelligent analysis unit, a communication and interaction unit, and an energy and communication unit, wherein:

[0012] The cell sensor detection unit is used to perform parallel detection of circulating tumor cells and lactate concentration in body fluids using a flexible microelectrode array modified with gold nanorods and a microfluidic enrichment channel;

[0013] The signal acquisition and preprocessing unit is used to receive the weak analog signal output by the cell sensor detection unit, and amplify, filter and digitize the weak analog signal;

[0014] The data processing and intelligent analysis unit is used to build a privacy-preserving hybrid intelligent analysis framework, achieve model collaborative optimization of multi-center data through a federated learning architecture, integrate traditional statistical methods with deep learning technology for anomaly identification, and dynamically generate personalized baseline models based on individual physiological characteristics for anomaly detection;

[0015] The communication and interaction unit is used to implement data transmission using a multi-mode communication protocol, supports adaptive adjustment of transmission power according to data priority, and provides a touch screen and voice interaction interface;

[0016] The energy and communication unit is used to integrate piezoelectric energy harvesting, magnetic resonance coupling wireless charging technology and solid-state lithium batteries, and is combined with an intelligent power management chip to provide low-power dynamic power supply.

[0017] In the gold nanorod-modified flexible microelectrode array, the aspect ratio of the gold nanorods is 3:1 to 5:1, and the surface is modified with anti-EpCAM antibodies through a thiol self-assembled monolayer, and the fixed density of the antibodies is 5×10 9 molecules / mm 2 ~1×10 10 molecules / mm 2 The microfluidic enrichment channel adopts a spiral microchannel structure with a channel width of 50μm to 200μm and a depth of 30μm to 100μm. The immunomagnetic bead capture of circulating tumor cells is achieved through sheath flow technology, and the capture efficiency is ≥85%.

[0018] The cell sensor detection unit includes a nanomaterial sensing module, a microfluidic enrichment module, and a flexible substrate packaging module, wherein:

[0019] The nanomaterial sensing module is used to detect biomolecule signals with high sensitivity through a microelectrode array modified with gold nanorods;

[0020] The microfluidic enrichment module is used to separate and concentrate circulating tumor cells using fluid dynamics principles;

[0021] The flexible substrate packaging module is used to provide a biocompatible implantable package.

[0022] The signal acquisition and preprocessing unit includes a signal amplification module, a filtering and noise reduction module, and an analog-to-digital conversion module, wherein:

[0023] The signal amplification module is used to enhance the intensity of UV-level analog signals using a low-noise instrumentation amplifier;

[0024] The filtering and noise reduction module is used to eliminate environmental and motion artifact interference through an adaptive filter;

[0025] The analog-to-digital conversion module is used to convert analog signals into digital signals using a high-precision ADC.

[0026] The data processing and intelligent analysis unit includes a federated learning module, a hybrid analysis module, and a personalized modeling module, wherein:

[0027] The federated learning module is used to implement horizontal federated learning based on the PyTorch framework and protect the privacy of multi-center medical data through homomorphic encryption technology;

[0028] The hybrid analysis module is used to integrate LSTM deep learning and Bayesian statistics to perform dynamic anomaly detection;

[0029] The personalized modeling module is used to generate a dynamic baseline threshold using an extended Kalman filter algorithm to adapt to the patient's physiological fluctuations.

[0030] The hybrid analysis module is used to integrate LSTM deep learning and Bayesian statistics for dynamic anomaly detection. The specific formula is as follows:

[0031] P anomaly (t) = α·f LSTM (X T )+(1-α)·p(X t |θ Bayes )

[0032] Where, f LSTM (X T ) represents the abnormal score output by the LSTM network, p(X t |θ Bayes ) represents the probability density of the Bayesian statistical model, α=0.7 is the weight coefficient, when P anomaly When (t)>0.8, an abnormal alarm is triggered.

[0033] The dynamic baseline threshold is generated in the personalized modeling module. The specific formula is as follows:

[0034] ① Three-state variable modeling: using a three-dimensional state vector containing mean, variance, and mean change rate Iterative update through extended Kalman filter, the formula is as follows:

[0035] Status prediction:

[0036] Observation correction: The Kalman gain

[0037] ② Adaptive threshold generation: Calculate the dynamic baseline mean based on the updated state vector and standard deviation The kσ threshold range was generated to be k = 3 corresponding to a 99.7% confidence interval, and the k value was automatically adjusted according to the patient's activity status, with k = 3.5 at rest and k = 2.5 during exercise.

[0038] The communication and interaction unit includes a multi-protocol communication controller module, an adaptive power regulation module, a touch interaction interface module, and a voice recognition and feedback module, wherein:

[0039] The multi-protocol communication controller module is used to support switching of multiple communication modes such as Bluetooth, Wi-Fi, and 5G;

[0040] The adaptive power adjustment module is used to dynamically adjust the transmission power according to the data priority;

[0041] The touch interactive interface module is used to provide a graphical operation interface for users to input and view information;

[0042] The voice recognition and feedback module is used to operate through voice instructions and implement voice prompt feedback.

[0043] The piezoelectric energy harvester in the energy and communication unit uses a PVDF-TrFE nanofiber membrane with a thickness of 50 μm to 100 μm. The output power density is ≥ 0.2 mW / cm at a human motion frequency of 0.5 Hz to 3 Hz. 2 , energy collection efficiency ≥ 60%; the magnetic resonance coupling wireless charging operates in the 6.78MHz ISM frequency band, the coupling coefficient between the transmitting coil and the receiving coil is ≥ 0.3, when the charging distance is 2cm~5cm, the energy transmission efficiency is ≥ 70%, and supports the foreign object detection function of the Qi2.0 protocol.

[0044] The solid-state lithium battery has an energy density of ≥400Wh / L, a capacity of 500mAh to 1500mAh, supports more than 1000 charge and discharge cycles, realizes dynamic voltage scaling through the power management chip LTC3330, and the overall standby power consumption of the system is ≤1mW.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] The present invention uses gold nanorod-modified flexible microelectrode arrays and microfluidic technology to achieve high-sensitivity parallel detection of circulating tumor cells and lactate concentrations in body fluids, completes signal preprocessing through low-noise amplification, filtering and high-precision analog-to-digital conversion, and uses a privacy-protected hybrid intelligent analysis framework driven by federated learning to achieve multi-center data collaborative optimization and personalized anomaly detection. It uses a multi-mode communication protocol combined with dynamic power regulation to achieve reliable data transmission and human-computer interaction, and constructs a low-power dynamic power supply system through piezoelectric energy harvesting, magnetic resonance wireless charging and solid-state lithium batteries, thereby providing cancer patients with accurate, safe and long-lasting real-time physiological indicator monitoring and clinical decision support. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 This is a system diagram of a multi-parameter real-time physiological indicator monitoring system for tumor patients according to the present invention.

[0048] Figure 2 This is a schematic diagram of the microfluidic enrichment channel structure in a multi-parameter real-time physiological indicator monitoring system for tumor patients of the present invention.

[0049] Figure 3 This is a state machine diagram of energy management in a multi-parameter real-time physiological indicator monitoring system for tumor patients according to the present invention.

[0050] Description of Figure Numbers:

[0051] 100. Cell sensor detection unit; 101. Nanomaterial sensing module; 102. Microfluidic enrichment module; 103. Flexible substrate packaging module; 200. Signal acquisition and preprocessing unit; 201. Signal amplification module; 202. Filtering and noise reduction module; 203. Analog-to-digital conversion module; 300. Data processing and intelligent analysis unit; 301. Federated learning module; 302. Hybrid analysis module; 303. Personalized modeling module; 400. Communication and interaction unit; 401. Multi-protocol communication controller module; 402. Adaptive power regulation module; 403. Touch interaction interface module; 404. Speech recognition and feedback module; 500. Energy and communication unit. DETAILED DESCRIPTION

[0052] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0053] Example:

[0054] like Figure 1-Figure 3As shown, this embodiment provides a multi-parameter real-time physiological indicator monitoring system for tumor patients, including a cell sensor detection unit 100, a signal acquisition and preprocessing unit 200, a data processing and intelligent analysis unit 300, a communication and interaction unit 400, and an energy and communication unit 500, wherein: the cell sensor detection unit 100 is used to use a flexible microelectrode array modified with gold nanorods and a microfluidic enrichment channel to perform parallel detection of circulating tumor cells and lactate concentration in body fluids; the signal acquisition and preprocessing unit 200 is used to receive the weak analog signal output by the cell sensor detection unit 100, and amplify, filter and digitize the weak analog signal; the data Data processing and intelligent analysis unit 300: used to build a privacy-preserving hybrid intelligent analysis framework, realize model collaborative optimization of multi-center data through a federated learning architecture, integrate traditional statistical methods and deep learning technology for anomaly identification, and dynamically generate personalized baseline models based on individual physiological characteristics for anomaly detection; Communication and interaction unit 400: used to realize data transmission through multi-mode communication protocols, support adaptive adjustment of transmission power according to data priority, and provide touch screen and voice interaction interface; Energy and communication unit 500: used to integrate piezoelectric energy harvesting, magnetic resonance coupling wireless charging technology and solid-state lithium batteries, combined with intelligent power management chips for low-power dynamic power supply.

[0055] In this embodiment, it should be noted that: the cell sensor detection unit 100 detects circulating tumor cells and lactate concentrations, which are processed by the signal acquisition and preprocessing unit 200, and then the data processing and intelligent analysis unit 300 performs federated learning-driven intelligent analysis, and then realizes data transmission and human-computer interaction through the communication and interaction unit 400, and is dynamically powered by the energy and communication unit 500.

[0056] In the present invention, the cell sensor detection unit 100 includes a nanomaterial sensing module 101, a microfluidic enrichment module 102, and a flexible substrate packaging module 103. The nanomaterial sensing module 101 is used to detect biomolecular signals with high sensitivity through a gold nanorod-modified microelectrode array; the microfluidic enrichment module 102 is used to separate and concentrate circulating tumor cells using fluid dynamics principles; and the flexible substrate packaging module 103 is used to provide a biocompatible, implantable package. In the gold nanorod-modified flexible microelectrode array, the gold nanorods have an aspect ratio of 3:1 to 5:1, and the surface is modified with anti-EpCAM antibodies via a thiol self-assembled monolayer, with an antibody immobilization density of 5×10 9 molecules / mm 2 ~1×10 10 molecules / mm 2The microfluidic enrichment channel adopts a spiral microchannel structure with a channel width of 50μm to 200μm and a depth of 30μm to 100μm. The immunomagnetic bead capture of circulating tumor cells is achieved through sheath flow technology, with a capture efficiency of ≥85%.

[0057] In this embodiment, it should be noted that: the nanomaterial sensing module 101 detects biomolecule signals with high sensitivity through a microelectrode array modified with gold nanorods, the microfluidic enrichment module 102 combines a spiral microfluidic enrichment channel to achieve efficient capture of circulating tumor cells (≥85% efficiency), and the flexible substrate packaging module 103 provides biocompatible implantable packaging.

[0058] In addition, it should be noted that the lactate concentration was detected using a lactate oxidase-modified working electrode, the potential difference between the working electrode and the reference electrode was +0.6V to +0.8V (vs Ag / AgCl), the detection linear range was 0.1mM to 20mM, and the detection lower limit was ≤0.05mM.

[0059] In the present invention, the signal acquisition and preprocessing unit 200 includes a signal amplification module 201, a filtering and noise reduction module 202, and an analog-to-digital conversion module 203, wherein: the signal amplification module 201 is used to use a low-noise instrumentation amplifier to enhance the UV-level analog signal strength; the filtering and noise reduction module 202 is used to eliminate environmental and motion artifact interference through an adaptive filter; the analog-to-digital conversion module 203 is used to convert the analog signal into a digital signal using a high-precision ADC.

[0060] In this embodiment, it should be noted that: the signal amplification module 201 enhances the weak signal strength through a low-noise instrumentation amplifier, the filtering and noise reduction module 202 eliminates environmental and motion artifact interference through an adaptive filter, and the analog-to-digital conversion module 203 completes signal digitization processing through a high-precision ADC, thereby realizing accurate collection and preprocessing of bioelectric signals.

[0061] In addition, it should be noted that: the low-noise instrumentation amplifier is used to amplify the weak ±10mV signal output by the microelectrode by 1000 times; the filtering and noise reduction module 202 uses a fourth-order Butterworth low-pass filter with a cutoff frequency set to 100Hz to filter out high-frequency noise; the analog-to-digital conversion module 203 uses a 24-bit Σ-Δ analog-to-digital converter ADS1256 with an adjustable sampling rate of 200SPS to 10kSPS.

[0062] In the present invention, the data processing and intelligent analysis unit 300 includes a federated learning module 301, a hybrid analysis module 302, and a personalized modeling module 303. The federated learning module 301 is used to implement horizontal federated learning based on the PyTorch framework and protect the privacy of multi-center medical data through homomorphic encryption technology. The hybrid analysis module 302 is used to integrate LSTM deep learning and Bayesian statistics for dynamic anomaly detection. The specific formula is as follows:

[0063] P anomaly (t) = α·f LSTM (X T )+(1-α)·p(X t |θ Bayes )

[0064] Where, f LSTM (X T ) represents the abnormal score output by the LSTM network, p(X t |θ Bayes ) represents the probability density of the Bayesian statistical model, α=0.7 is the weight coefficient, when P anomaly When (t)>0.8, an abnormal alarm is triggered. The personalized modeling module 303 is used to generate a dynamic baseline threshold using the extended Kalman filter algorithm to adapt to the patient's physiological fluctuations. The specific formula is as follows: ① Three-state variable modeling: using a three-dimensional state vector containing mean, variance, and mean change rate By iteratively updating the extended Kalman filter, the formula is as follows: State prediction: Observation correction: The Kalman gain ② Adaptive threshold generation: Calculate the dynamic baseline mean based on the updated state vector and standard deviation The kσ threshold range was generated to be k = 3 corresponding to a 99.7% confidence interval, and the k value was automatically adjusted according to the patient's activity status, with k = 3.5 at rest and k = 2.5 during exercise.

[0065] In this embodiment, it should be noted that: the federated learning module 301 implements privacy-protected multi-center data collaborative training, uses the hybrid analysis module 302 to fuse LSTM and Bayesian methods for dynamic anomaly detection, and uses the extended Kalman filter algorithm of the personalized modeling module 303 to generate an adaptive baseline threshold.

[0066] In addition, it should be noted that the federated learning architecture uses a horizontal federated learning model. Each medical institution's local model is built based on the PyTorch framework. Model parameter updates are transmitted using homomorphic encryption, and the global model aggregation cycle is 1 to 24 hours. In the hybrid intelligent analysis framework, traditional statistical methods use CUSUM control charts to detect step changes in lactate concentration. The deep learning model uses a bidirectional LSTM network to analyze the time series trend of circulating tumor cell counts, with an anomaly identification accuracy of ≥95%. A personalized baseline model is generated using the extended Kalman filter algorithm, with a dynamic update cycle of 5 to 30 minutes. The baseline threshold is the mean of the real-time data ±3 times the standard deviation.

[0067] In the present invention, the communication and interaction unit 400 includes a multi-protocol communication controller module 401, an adaptive power adjustment module 402, a touch interaction interface module 403, and a voice recognition and feedback module 404, wherein: the multi-protocol communication controller module 401 is used to support switching of multiple communication modes such as Bluetooth, Wi-Fi, and 5G; the adaptive power adjustment module 402 is used to dynamically adjust the transmission power according to data priority; the touch interaction interface module 403 is used to provide a graphical operation interface for users to input and view information; the voice recognition and feedback module 404 is used to operate through voice commands and realize voice prompt feedback.

[0068] In this embodiment, it should be noted that: a multi-protocol communication controller module 401 is used to implement switching between multiple wireless communication modes, an adaptive power regulation module 402 is combined to perform dynamic power optimization, and a touch interaction interface module 403 and a voice recognition and feedback module 404 are integrated.

[0069] Additionally, it's worth noting that the communication mode switching logic automatically switches when signal strength is less than -85dBm, with a switching delay of less than 200ms. Bluetooth 5.2 is used for real-time data transmission within a 10-meter range, with transmit power adaptively adjusting between -20dBm and +4dBm. The touchscreen interface integrates tumor marker trend graphs, abnormal event alert pop-ups, and historical data query functionality. Voice interaction supports natural language command recognition.

[0070] In the present invention, the piezoelectric energy harvester in the energy and communication unit 500 uses a PVDF-TrFE nanofiber membrane with a thickness of 50 μm to 100 μm and an output power density of ≥ 0.2 mW / cm at a human motion frequency of 0.5 Hz to 3 Hz. 2The device boasts an energy collection efficiency of ≥60%. The magnetic resonance coupling wireless charging system operates in the 6.78MHz ISM band, with a coupling coefficient of ≥0.3 between the transmitting and receiving coils. At a charging distance of 2cm to 5cm, it achieves an energy transfer efficiency of ≥70%. It also supports Qi 2.0 foreign object detection. The solid-state lithium battery boasts an energy density of ≥400Wh / L, a capacity of 500mAh to 1500mAh, and supports over 1000 charge and discharge cycles. Dynamic voltage scaling is achieved through the LTC3330 power management chip, ensuring overall system standby power consumption of ≤1mW.

[0071] In this embodiment, it should be noted that: motion energy harvesting is achieved through PVDF-TrFE nanofiber membrane, a hybrid power supply system is constructed by combining 6.78MHz magnetic resonance wireless charging and solid-state lithium batteries, and intelligent control is achieved by the LTC3330 power management chip.

[0072] In addition, it should be noted that: the switching conditions between piezoelectric energy harvesting and wireless charging (for example, when the battery power is less than 20% and human motion is detected, piezoelectric charging is enabled first, and wireless charging is turned off when the charging power is ≥0.5mW), and the sleep mode power consumption of the power management chip (for example, the current of LTC3330 is ≤50nA in deep sleep).

[0073] Throughout this specification, references to terms such as "one embodiment," "example," or "specific example" indicate that the specific features, structures, materials, or characteristics described in conjunction with that embodiment or example are included in at least one embodiment or example of the present invention. In this specification, schematic representations of these terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0074] The preferred embodiments of the present invention disclosed above are intended only to help illustrate the present invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the present invention to the specific embodiments described. Obviously, many modifications and variations are possible based on the content of this specification. These embodiments are selected and described in detail in this specification to better explain the principles and practical applications of the present invention, thereby enabling those skilled in the art to better understand and utilize the present invention. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A multi-parameter real-time physiological index monitoring system for tumor patients, characterized by: The system comprises a cell sensor detection unit (100), a signal acquisition and preprocessing unit (200), a data processing and intelligent analysis unit (300), a communication and interaction unit (400), and an energy and communication unit (500), wherein: The cell sensor detection unit (100) is used to perform parallel detection of circulating tumor cells and lactate concentration in body fluids using a flexible microelectrode array modified with gold nanorods and a microfluidic enrichment channel; The signal acquisition and preprocessing unit (200) is used to receive the weak analog signal output by the cell sensor detection unit (100), and amplify, filter and digitize the weak analog signal; The data processing and intelligent analysis unit (300) is used to construct a privacy-preserving hybrid intelligent analysis framework, realize model collaborative optimization of multi-center data through a federated learning architecture, integrate traditional statistical methods with deep learning technology to perform anomaly identification, and dynamically generate a personalized baseline model based on individual physiological characteristics for anomaly detection; The communication and interaction unit (400) is used for implementing data transmission with a multi-mode communication protocol, supports adaptive adjustment of transmission power according to data priority, and provides a touch screen and voice interaction interface; The energy and communication unit (500) is used to integrate piezoelectric energy collection, magnetic resonance coupling wireless charging technology and solid-state lithium batteries, and is combined with an intelligent power management chip to perform low-power dynamic power supply.

2. A multi-parameter real-time physiological indicator monitoring system for tumor patients according to claim 1, characterized in that: In the gold nanorod-modified flexible microelectrode array, the aspect ratio of the gold nanorods is 3:1 to 5:1, and the surface is modified with anti-EpCAM antibodies through a thiol self-assembled monolayer, and the fixed density of the antibodies is 5×10 9 molecules / mm 2 ~1×10 10 molecules / mm 2 The microfluidic enrichment channel adopts a spiral microchannel structure with a channel width of 50μm to 200μm and a depth of 30μm to 100μm. The immunomagnetic bead capture of circulating tumor cells is achieved through sheath flow technology, and the capture efficiency is ≥85%.

3. A multi-parameter real-time physiological indicator monitoring system for tumor patients according to claim 2, characterized in that: The cell sensor detection unit (100) comprises a nanomaterial sensing module (101), a microfluidic enrichment module (102), and a flexible substrate packaging module (103), wherein: The nanomaterial sensing module (101) is used to detect biomolecule signals with high sensitivity through a microelectrode array modified with gold nanorods; The microfluidic enrichment module (102) is used to separate and concentrate circulating tumor cells using fluid dynamics principles; The flexible substrate packaging module (103) is used to provide a biocompatible implantable package.

4. A multi-parameter real-time physiological indicator monitoring system for tumor patients according to claim 1, characterized in that: The signal acquisition and preprocessing unit (200) comprises a signal amplification module (201), a filtering and noise reduction module (202), and an analog-to-digital conversion module (203), wherein: The signal amplification module (201) is used to enhance the intensity of UV-level analog signals using a low-noise instrumentation amplifier; The filtering and noise reduction module (202) is used to eliminate environmental and motion artifact interference through an adaptive filter; The analog-to-digital conversion module (203) is used to convert analog signals into digital signals using a high-precision ADC.

5. A multi-parameter real-time physiological index monitoring system for tumor patients according to claim 4, characterized in that: The data processing and intelligent analysis unit (300) includes a federated learning module (301), a hybrid analysis module (302), and a personalized modeling module (303), wherein: The federated learning module (301) is used to implement horizontal federated learning based on the PyTorch framework and protect the privacy of multi-center medical data through homomorphic encryption technology; The hybrid analysis module (302) is used to integrate LSTM deep learning and Bayesian statistics to perform dynamic anomaly detection; The personalized modeling module (303) is used to generate a dynamic baseline threshold using an extended Kalman filter algorithm to adapt to the patient's physiological fluctuations.

6. A multi-parameter real-time physiological index monitoring system for cancer patients according to claim 5, characterized in that: The hybrid analysis module (302) is used to integrate LSTM deep learning and Bayesian statistics to perform dynamic anomaly detection. The specific formula is as follows: P anomaly (t)=α·f LSTM (X T )+(1-α)·p(X t |θ Bayes ) Where, f LSTM (X T ) represents the abnormal score output by the LSTM network, p(X t |θ Bayes ) represents the probability density of the Bayesian statistical model, α=0.7 is the weight coefficient, when P anomaly When (t)>0.8, an abnormal alarm is triggered.

7. A multi-parameter real-time physiological index monitoring system for cancer patients according to claim 5, characterized in that: The dynamic baseline threshold is generated in the personalized modeling module (303), and the specific formula is as follows: ① Three-state variable modeling: using a three-dimensional state vector containing mean, variance, and mean change rate Iterative update through extended Kalman filter, the formula is as follows: Status prediction: Observation correction: The Kalman gain ② Adaptive threshold generation: Calculate the dynamic baseline mean based on the updated state vector and standard deviation The kσ threshold range was generated to be k = 3 corresponding to a 99.7% confidence interval, and the k value was automatically adjusted according to the patient's activity status, with k = 3.5 at rest and k = 2.5 during exercise.

8. The multi-parameter real-time physiological index monitoring system for cancer patients according to claim 1, characterized in that: The communication and interaction unit (400) comprises a multi-protocol communication controller module (401), an adaptive power regulation module (402), a touch interaction interface module (403), and a speech recognition and feedback module (404), wherein: The multi-protocol communication controller module (401) is used to support switching of multiple communication modes such as Bluetooth, Wi-Fi, and 5G; The adaptive power adjustment module (402) is used to dynamically adjust the transmission power according to the data priority; The touch interaction interface module (403) is used to provide a graphical operation interface for users to input and view information; The voice recognition and feedback module (404) is used to operate through voice instructions and implement voice prompt feedback.

9. A multi-parameter real-time physiological index monitoring system for cancer patients according to claim 1, characterized in that: The piezoelectric energy collection in the energy and communication unit (500) adopts a PVDF-TrFE nanofiber membrane with a thickness of 50 μm to 100 μm, and an output power density of ≥0.2 mW / cm at a human body movement frequency of 0.5 Hz to 3 Hz. 2 , energy collection efficiency ≥ 60%; the magnetic resonance coupling wireless charging operates in the 6.78MHz ISM frequency band, the coupling coefficient between the transmitting coil and the receiving coil is ≥ 0.3, when the charging distance is 2cm~5cm, the energy transmission efficiency is ≥ 70%, and supports the foreign object detection function of the Qi2.0 protocol.

10. A multi-parameter real-time physiological indicator monitoring system for cancer patients according to claim 9, characterized in that: The solid-state lithium battery has an energy density of ≥400Wh / L, a capacity of 500mAh to 1500mAh, supports more than 1000 charge and discharge cycles, realizes dynamic voltage scaling through the power management chip LTC3330, and the overall standby power consumption of the system is ≤1mW.

Images