Portable single gene expression rapid detection system and tumor risk assessment method thereof

Abstract

The invention relates to the technical field of gene detection and risk assessment, and particularly discloses a portable single gene expression rapid detection system and a tumor risk assessment method thereof. Calling a temperature control time sequence parameter and an amplification efficiency reference threshold value matched with the target gene from a preset database; the reaction cavity is controlled to be in different temperature intervals in sequence according to the temperature control time sequence parameters so as to execute an amplification reaction, in the reaction process, fluorescence intensity signals in the reaction cavity are collected according to the preset sampling frequency, and the fluorescence intensity change rate is calculated based on the time dimension. A closed-loop intelligent detection and evaluation system is constructed by integrating a sample physical property pre-detection module, an amplification efficiency real-time prediction module, a dynamic compensation module and the like; the system can realize multi-dimensional active monitoring and interference suppression of the whole process of gene amplification reaction on a portable platform.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of gene detection and risk assessment, in particular to a portable single-gene expression rapid detection system and a tumor risk assessment method thereof. BACKGROUND

[0002] In the prior art, tumor risk assessment based on single-gene expression is usually completed by large-scale quantitative PCR instruments and other equipment in a laboratory environment, and its detection process is fixed and requires a stable environment. When applied to portable and near-patient detection scenarios, sample states such as viscosity, light transmittance, amplification efficiency fluctuation during the reaction process, environmental temperature changes, and multi-channel optical crosstalk can introduce significant errors, resulting in inaccurate quantification of gene expression. Traditional systems lack online perception and dynamic compensation capabilities for the above interference factors, and the reliability and stability of their detection results are severely limited, making it difficult to meet the actual needs of rapid and accurate assessment of tumor risk on site.

[0003] The technical conflicts mainly manifest in the following aspects: on the one hand, to achieve rapid and portable detection, the system needs to be highly integrated and designed with small size using microfluidic chips, which will inevitably exacerbate the non-uniformity of the temperature field in the reaction chamber and the signal crosstalk between multiple channels;

[0004] On the other hand, the diversity of samples and the uncertainty of the environment, such as environmental temperature differences and sample bubbles, also pose higher requirements for the accuracy and robustness of the detection. Existing portable solutions often simplify the temperature control and signal processing logic, and cannot effectively balance small size and high accuracy, environmental adaptability, resulting in insufficient reliability of the evaluation results in clinical practice, making it difficult to be directly used for critical tumor risk classification decisions. SUMMARY

[0005] The present application aims to at least partially solve one of the technical problems in the related art. To this end, the purpose of the present application is to propose a portable single-gene expression rapid detection system and a tumor risk assessment method thereof to improve the reliability of tumor risk level determination.

[0006] To achieve the above-mentioned purpose, the first embodiment of the present application proposes a tumor risk assessment method of a portable single-gene expression rapid detection system, comprising the following steps:

[0007] S1, in response to a target gene confirmation instruction, retrieving temperature control time sequence parameters and amplification efficiency reference threshold values matched with the target gene from a pre-set database;

[0008] S2, controlling the reaction chamber to sequentially pass through different temperature intervals to perform amplification reaction according to the temperature control time sequence parameters, and collecting fluorescence intensity signals in the reaction chamber at a preset sampling frequency and calculating the fluorescence intensity change rate based on the time dimension during the reaction.

[0009] S3, the fluorescence intensity rate of change is substituted into an amplification efficiency prediction model to obtain a current amplification efficiency prediction value; the amplification efficiency prediction value is compared with the amplification efficiency reference threshold value, and if the amplification efficiency prediction value is lower than the amplification efficiency reference threshold value, an enzyme activation instruction is generated; the enzyme activation instruction is used to control the release of an enzyme activator into a reaction cavity and prolong the duration of a current reaction stage to compensate for the amplification rate of a reaction system;

[0010] S4, after the amplification reaction ends, a final reaction product fluorescence signal is collected; signal correction processing including gain normalization and background subtraction is performed on the reaction product fluorescence signal to obtain a target gene expression amount;

[0011] S5, the target gene expression amount is input into a multi-dimensional clinical risk assessment model to output a tumor risk grade.

[0012] To achieve the above object, the second aspect embodiment of the present application proposes a portable single-gene expression rapid detection system, which comprises:

[0013] A multi-channel microfluidic chip, which is integrated with a plurality of parallel reaction channels and bypass chambers in fluid communication with each reaction channel, and air isolation grooves are arranged between the reaction channels for heat insulation;

[0014] A regional temperature control unit, including a thin film heater corresponding to the central region of the reaction channel and a heat insulation material layer corresponding to the edge region, for controlling the reaction cavity to be in different temperature intervals in sequence to perform amplification reaction according to the preset temperature control time sequence parameters;

[0015] A fluorescence detection unit, provided with an optical readout window corresponding to the position of the reaction channel, for collecting fluorescence intensity signals at a preset sampling frequency during the amplification reaction, and collecting the final reaction product fluorescence signal after the reaction ends;

[0016] A main control module, electrically connected with the regional temperature control unit and the fluorescence detection unit, configured to perform the following operations:

[0017] In response to a target gene confirmation instruction, the temperature control time sequence parameters and the amplification efficiency reference threshold value matched with the target gene are called from a preset database;

[0018] The fluorescence intensity rate of change is calculated based on the time dimension, which is input into an amplification efficiency prediction model to obtain an amplification efficiency prediction value, and compared with the amplification efficiency reference threshold value, and if it is lower than the threshold value, the enzyme activator is released into the reaction cavity and the duration of the current reaction stage is prolonged;

[0019] The signal correction processing including gain normalization and background deduction is performed on the collected reaction product fluorescence signal to obtain a target gene expression amount;

[0020] The target gene expression amount is input into a multi-dimensional clinical risk assessment model to output a tumor risk grade;

[0021] The sample physical property pre-check unit includes a sensor for detecting the light transmittance parameter and the flow velocity parameter of the injected sample, and triggers an abnormality interception logic to prohibit the temperature control unit from starting if the parameters are abnormal;

[0022] The bubble interference detection and processing unit is configured to perform linear regression fitting on the fluorescence intensity signal and calculate a fluctuation characteristic value, and triggers a temperature shock bubble removal operation if the fluctuation characteristic value exceeds a stationarity threshold value;

[0023] The environmental temperature difference compensation unit is used to obtain the external environment temperature and calculate the difference with the set temperature, and starts an asymmetric temperature control mode if the temperature difference exceeds a temperature difference threshold value, so that the target temperature of the edge region is higher than that of the center region;

[0024] The optical crosstalk deconvolution unit is used to identify adjacent channels based on a channel mapping relationship, and performs deconvolution operation on the target channel signal when the crosstalk exceeds a threshold value;

[0025] The abnormal interruption and follow-up management unit is used to generate a log file and upload the server when a hardware failure is monitored, and calculates a review date and sends a reminder information to the user terminal after outputting the tumor risk grade.

[0026] To achieve the above object, the third aspect embodiment of the present application provides an electronic device, comprising a memory, a processor and a computer program stored in the memory, wherein the computer program is executed by the processor to realize the tumor risk assessment method of the portable single gene expression rapid detection system.

[0027] Compared with the prior art, the present application has the following beneficial effects:

[0028] The portable single gene expression rapid detection system and the tumor risk assessment method thereof can realize multi-dimensional active monitoring and interference suppression of the whole process of gene amplification reaction on a portable platform by integrating sample physical property pre-check, amplification efficiency real-time prediction and dynamic compensation modules, effectively guaranteeing the accuracy and repeatability of target gene expression quantification under different sample states and environmental conditions; further, based on the accurate expression data, multi-dimensional clinical risk assessment is performed, which significantly improves the reliability and practicability of tumor risk grade determination, and realizes the whole process service from detection to health management. BRIEF DESCRIPTION OF DRAWINGS

[0029] The disclosure of the present application will be described with reference to the accompanying drawings. It should be understood that the drawings are only for the purpose of illustration and are not intended to limit the scope of protection of the present application, in which the same reference signs are used to refer to the same parts throughout the drawings. Among them:

[0030] Figure 1 is a flowchart of the tumor risk assessment method of the portable single gene expression rapid detection system provided by the present application;

[0031] Figure 2 is a fluorescence change rate response curve diagram of the amplification efficiency prediction model in the tumor risk assessment method of the portable single gene expression rapid detection system provided by the present application;

[0032] Figure 3 is a schematic diagram of the comparison between the bubble interference signal and the fitting curve in the tumor risk assessment method of the portable single gene expression rapid detection system provided by the present application;

[0033] Figure 4 is a schematic diagram of the comparison between the fluorescence amplification curves before and after enzyme injection in the tumor risk assessment method of the portable single gene expression rapid detection system provided by the present application;

[0034] Figure 5 is a chip temperature field distribution simulation diagram under different environmental temperature differences in the tumor risk assessment method of the portable single gene expression rapid detection system provided by the present application;

[0035] Figure 6 is an implementation execution diagram of the portable single gene expression rapid detection system provided by the present application;

[0036] Figure 7 is a structural diagram of the electronic device provided by the present application. DETAILED DESCRIPTION

[0037] The embodiments of the present application will be described in detail below, and examples of the embodiments are shown in the accompanying drawings, in which the same or similar reference signs represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the drawings are exemplary and are intended to explain the present application, and cannot be understood as limiting the present application.

[0038] The portable single gene expression rapid detection system, the tumor risk assessment method thereof, and the electronic device of the embodiments of the present application will be described below with reference to the accompanying drawings.

[0039] Embodiment One:

[0040] Figure 1 is a flowchart of the tumor risk assessment method of the portable single gene expression rapid detection system of an embodiment of the present application, and the method of the present embodiment mainly includes the following steps:

[0041] S1, detecting parameter configuration and initialization step

[0042] At the beginning of the detection process, the system first needs to determine the object and environmental benchmark of the detection. This step is the cornerstone of the whole process, which ensures that the subsequent hardware actions are accurately matched with the biological characteristics of the sample to be tested.

[0043] For example, when the user inserts the microfluidic chip containing the sample to be tested into the device card slot, the system automatically activates the optical scanning module to read the pre-printed two-dimensional code or RFID electronic tag on the surface of the chip. The identification code records the types of primer probes and applicable tumor marker species pre-installed in the chip through encryption coding. In response to the target gene confirmation instruction generated by the reading operation, the system's main control unit immediately accesses the pre-installed database locally or in the cloud.

[0044] Optionally, the pre-installed database stores special parameter sets for lung cancer EGFR gene, breast cancer HER2 gene, colorectal cancer KRAS gene, and other common tumor driver genes. The system accurately retrieves the temperature control time sequence parameters and amplification efficiency benchmark threshold values matching the current target gene from the database according to the instruction. These parameters are not just simple temperature settings, but also include dynamic heating curves optimized for specific primer melting temperature Tm values, and minimum tolerance efficiency red lines calibrated for specific enzyme activity characteristics. For example, for EGFR mutation detection, the amplification efficiency benchmark threshold value loaded by the system may be set to 85%, which means that if the real-time monitored amplification efficiency is lower than this standard, the system will determine that the current reaction is blocked and needs to be intervened immediately.

[0045] In addition, before formally entering the heating reaction, in order to ensure that the physical state of the sample meets the requirements of the fluid dynamics of the microfluidic chip, and to avoid the blockage of the flow channel due to the excessive viscosity of the sample or the excessively high optical background due to severe hemolysis, the present embodiment introduces a sample physical property pre-checking step after step S1 and before step S2.

[0046] Specifically, the system utilizes optical sensors and pressure sensors integrated near the sample inlet to detect the light transmittance parameter and the flow speed parameter of the injected sample, respectively. The system internally presets a qualified light transmittance range and a speed interval corresponding to normal viscosity. For example, the qualified light transmittance range can be defined as an absorbance less than 0.5 OD at a wavelength of 520 nm, and the normal flow speed interval can be defined as a flow rate between 2 microliters per second and 5 microliters per second under a standard driving pressure. If the light transmittance parameter fed back by the sensors exceeds the preset qualified light transmittance range, or the flow speed parameter exceeds the preset speed interval corresponding to normal viscosity, the system will immediately trigger an abnormal interception logic. The abnormal interception logic is configured to lock the subsequent temperature control unit through software, forcibly prohibit the system from entering the heating operation of step S2, and generate prompt information of sample quality abnormality through the human-computer interaction interface, such as "detection of severe hemolysis of sample, please re-collect" or "sample viscosity is abnormal, please dilute and try again", so as to avoid wasting expensive microfluidic chip reagents and prevent pollution of the system pipeline.

[0047] S2, dynamic reaction control and environment self-adaptive step

[0048] After the parameters are loaded and the sample pre-check is passed, the system enters the core biochemical reaction control phase. The core of this step is to control the reaction cavity to be in different temperature intervals in sequence according to the retrieved temperature control time sequence parameters, so as to simulate the thermal cycling environment required by the polymerase chain reaction (PCR).

[0049] For example, the specific control time sequence of the multi-stage variable temperature amplification reaction in the embodiment is finely divided into three key stages:

[0050] Firstly, the denaturation stage: the system controls the temperature of the reaction cavity to rapidly rise and stably maintain in the first temperature interval for a first preset time length. For example, the temperature is controlled at 94-96 degrees Celsius for 30-60 seconds. The main purpose of this stage is to destroy the hydrogen bonds between double-stranded DNA molecules by high temperature, so that they are uncoiled into single-stranded structures, providing templates for subsequent primer binding. The length of the first preset time length depends on the length and GC content of the target gene fragment, which is determined by the parameters loaded in step S1.

[0051] Secondly, annealing-extension and monitoring phase: this is a critical composite function phase. The system reduces and controls the temperature of the reaction cavity in the second temperature interval for a second preset duration. For example, the temperature is controlled between 58 degrees Celsius and 62 degrees Celsius for 40 seconds to 90 seconds. At this temperature, the primers specifically bind to the single-stranded DNA template and extend to synthesize a new complementary strand by taking free nucleotides under the catalysis of DNA polymerase. It is particularly important to emphasize that the "acquisition of fluorescence intensity signal at a preset sampling frequency" mentioned in step S2 and the "execution of enzyme activation instruction" mentioned in subsequent step S3 are designed to be completed within the time window of this phase. This is because during the extension process, the fluorescence probe (such as TaqMan probe) is hydrolyzed or the dye (such as SYBR Green) is embedded, and the change of fluorescence signal can most directly reflect the real-time kinetic characteristics of the amplification reaction.

[0052] Finally, stable detection phase: in order to eliminate the influence of temperature fluctuations on the fluorescence quantum yield and ensure the accuracy of the final reading, the system adjusts the temperature of the reaction cavity to the third temperature interval and continues for a third preset duration until the process enters step S4. For example, the temperature is controlled between 72 degrees Celsius and 75 degrees Celsius. Among them, from the perspective of thermodynamics, the temperature value of the first temperature interval must be higher than that of the second temperature interval, and the temperature value of the second temperature interval is usually higher than or equal to that of the third temperature interval, but in some two-step PCR, the third temperature interval may be combined with the second temperature interval. In this embodiment, in order to pursue the ultimate signal stability, the third temperature interval is independently set as a constant temperature plateau.

[0053] During the reaction, the optical detection unit scans the reaction cavity and acquires the fluorescence intensity signal in the reaction cavity according to the preset sampling frequency, such as once every 2 seconds. The processor receives these discrete time series signals and calculates the fluorescence intensity change rate based on the time dimension. This change rate is not just a simple difference, but is obtained by smoothing the derivative of the data points in the sliding time window, which can sensitively capture the signal take-off trend in the early stage of exponential amplification.

[0054] S3, amplification efficiency closed-loop compensation and intelligent enzyme injection step

[0055] Step S3 constructs a monitoring-prediction-intervention closed-loop feedback mechanism to solve the drawbacks of traditional PCR that cannot be corrected.

[0056] Firstly, the system takes the fluorescence intensity change rate calculated in step S2 as an input variable and substitutes it into the pre-installed amplification efficiency prediction model. This model aims to infer the potential activity of the reaction system through the current reaction rate.

[0057] For example, in order to accurately describe the nonlinear characteristics of biochemical reactions, the construction logic of the amplification efficiency prediction model in this embodiment is designed to establish a functional mapping relationship between the amplification efficiency prediction value and the fluorescence intensity change rate. This functional mapping relationship is configured to conform to the principles of enzyme reaction kinetics: that is, the amplification efficiency prediction value increases with the increase of the fluorescence intensity change rate, and is limited by the maximum catalytic capacity of the enzyme, and approaches a preset maximum efficiency constant, such as 100% or the theoretical value 1.0.

[0058] In specific calculation, the processor performs the following mathematical operation: using the fluorescence intensity change rate Divided by the target specificity factor To obtain a dimensionless intermediate variable . Then, the power function value with the natural constant as the base and the negative number of the intermediate variable as the exponent is calculated . Finally, the power function value is weighted and deducted by the maximum efficiency constant to obtain the amplification efficiency prediction value . Its mathematical expression can be expressed as:

[0059] ;

[0060] Wherein, is the second model constant, used to adjust the intercept and steepness of the curve.

[0061] It is particularly noted that the value of the target specificity factor is not fixed, but corresponds uniquely to the type of target gene identified in step S1. For example, for a gene fragment with high GC content and difficult amplification, the corresponding value is smaller, meaning that the same signal change rate corresponds to a relatively high amplification effort; and for an easy-to-amplify fragment, the value is larger. This parameterized design enables the same algorithm to adapt to the detection needs of different gene projects.

[0062] As Figure 2 shows the response relationship of the amplification efficiency prediction model to the fluorescence intensity change rate, wherein the horizontal value represents the fluorescence signal slope collected in real time during the annealing and extension stage, and the vertical value represents the amplification efficiency prediction result given by the system according to the slope.

[0063] Figure 2The different colors of the curves correspond to different target specificity parameters, representing the inherent kinetic differences of different gene fragments during amplification. The region with smaller slope usually appears in the early stage of amplification, when the three curves slowly rise from near zero, indicating that the amplification capacity of the system has not yet fully emerged. As the fluorescence rate gradually increases, the curves begin to rise at different speeds, with the curve of the larger target specificity parameter being darker in color and rising significantly faster, indicating that this type of gene fragment is more likely to be determined as having good amplification activity under the same fluorescence rate condition.

[0064] Figure 2 When the rate reaches between four and five, the three curves gradually approach the flat region at the top, indicating that the system's amplification capacity is approaching its limit. This graph reflects the non-linear growth feature adopted by the prediction model in Example One, which can effectively simulate the process of gradually saturating enzymatic reactions. It also illustrates the distinguishing ability of different gene types in the model, allowing the system to automatically select the appropriate specificity parameter based on different target genes, making the amplification efficiency prediction more adaptable and reliable. The curve can directly prove that the amplification efficiency prediction model can sensitively reflect the development trend of the fluorescence signal and provide accurate basis for whether to perform enzyme compensation.

[0065] Next, the system compares the calculated amplification efficiency prediction value with the pre-set amplification efficiency threshold in real time. If the comparison result shows that the amplification efficiency prediction value is lower than the amplification efficiency threshold, it usually means that there may be inhibition of enzyme activity in the reaction system (such as the presence of trace inhibitors in the sample) or low primer binding efficiency. If no intervention is made at this time, the reaction may not reach the plateau phase, resulting in false negative results.

[0066] Therefore, once the efficiency is determined to be insufficient, the system immediately generates an enzyme activation instruction. The instruction is transmitted to the control interface of the microfluidic chip through the circuit, driving the micro-pump or micro-valve to act, controlling the pre-existing enzyme activator in the bypass chamber, such as a high-concentration Taq enzyme supplement or a magnesium ion cofactor solution, to release and flow into the main reaction cavity. At the same time, in order to provide sufficient action time for the newly added activator, the instruction also modifies the current temperature control program, prolonging the duration of the current reaction stage (i.e., the annealing and extension stage), such as 30 seconds. This series of combined actions constitutes a compensation for the amplification rate of the reaction system, ensuring that the weak reaction that may have failed can be revived, thereby reaching a detectable substrate concentration level before entering the final detection step.

[0067] As Figure 3 The figure shows the change trend of the fluorescence amplification curve before and after the intervention operation performed by the amplification efficiency closed-loop compensation and intelligent enzyme injection step, with the horizontal axis representing the reaction time and the vertical axis representing the real-time collected fluorescence intensity signal.

[0068] Figure 3 The blue solid line represents the signal linearly rising with time in the initial stage of the reaction, but the slope obviously decreases and quickly tends to be flat after thirty seconds, indicating that the amplification system may fail prematurely due to insufficient enzyme activity or inhibitor interference, making it difficult to enter the effective exponential amplification period. The red dotted line is the modified curve obtained by injecting enzyme activator and appropriately extending the reaction stage after the system identifies that the amplification efficiency is lower than the benchmark threshold. From the curve, it can be observed that the signal behaves consistently in the same initial stage, and after thirty seconds, it still maintains a high rising slope, and the final fluorescence intensity is much higher than that of the uncompensated group, fully demonstrating that the compensation measure successfully extends the system amplification activity.

[0069] Alternatively, to prevent data misjudgment caused by bubble interference, the embodiment also performs a bubble interference rejection sub-step in parallel while collecting the signal in step S2. The system performs linear regression fitting on the fluorescence intensity signal collected in the preset time window to obtain an ideal signal baseline. Then, the residual standard deviation of the original signal point relative to the fitted curve is calculated, and the ratio of the residual standard deviation to the current signal mean is defined as the fluctuation characteristic value.

[0070] If the fluctuation characteristic value exceeds the preset stationarity threshold, such as 5%, it indicates that the signal has a severe non-biological jump, which is most likely caused by a bubble passing through the detection window. At this time, the system determines that the current reaction chamber is disturbed by a bubble, and in order to prevent false data from entering the prediction model, the system temporarily suspends the execution of step S3, i.e., does not perform efficiency prediction and enzyme injection judgment, but preferentially triggers the temperature shock bubble removal operation. This operation is configured to control the temperature of the reaction chamber to superimpose an alternating temperature signal on the basis of the current set temperature, and the alternating temperature signal has a preset shock frequency (such as 0.5 Hz) and a preset shock amplitude (such as plus or minus 0.5 degrees Celsius). This rapid micro-thermal disturbance will cause thermal expansion and contraction pulsation of the fluid, effectively stripping and removing the small bubbles attached to the wall and out of the light path area. After the bubbles are removed, the system resumes the normal data collection and efficiency prediction process.

[0071] As Figure 4 The difference between the raw signal and the fitted curve collected by the fluorescence detection unit at a frequency of one per second during the amplification reaction is shown.

[0072] Figure 4The middle blue solid line represents the original fluorescence signal with mutation jump characteristics. Obvious fluctuations can be observed in the 20th to 25th seconds and the 45th to 48th seconds in the figure, such as a sudden increase of 30 to 50 units higher than the normal trend or a sudden decrease of more than 40 units. Such discontinuous jumps are difficult to explain by enzyme reaction behavior in amplification kinetics, which is consistent with the characteristics of bubble interference. The red dotted line is the smooth curve automatically generated by the system through polynomial fitting method, representing the signal change trend in the ideal amplification state, which is used to estimate the normal kinetic behavior.

[0073] In this figure, it can be seen that a plurality of original data points deviate from the fitting curve, indicating that the reaction cavity is most likely to encounter bubble shielding or interference during these periods. Combined with the fluctuation characteristic value calculation method of the present application, that is, the residual standard deviation and mean value ratio of the original signal and the fitting curve, the system can determine that the current signal has exceeded the preset stability threshold, triggering the temperature shock bubble removal operation. Figure 4 The appearance of the middle jump waveform and its difference from the smooth curve support the necessity and timeliness of the dynamic intervention mechanism, and thus provide a clear trigger basis for the stable recovery of amplification efficiency, significantly improving the automation and intelligence level of the detection process.

[0074] S4, end data acquisition and correction step

[0075] When the amplification reaction is completed according to the preset time sequence (or the extended time sequence), the reaction system has theoretically accumulated a sufficient amount of fluorescence product. At this time, the system enters the end detection link:

[0076] First, the system controls the optical readout device to collect the final reaction product fluorescence signal. This is usually a raw photoelectric conversion voltage value or a digital quantity after analog-to-digital conversion (Raw Data). However, due to the differences in optical devices of portable devices and the background fluorescence of sample matrix, the raw data cannot be directly used for horizontal comparison. Therefore, signal correction processing including gain normalization and background subtraction must be performed on the reaction product fluorescence signal to obtain the standardized target gene expression quantity.

[0077] For example, the specific execution logic of the signal correction processing is as follows: First, the system reads the current gain setting parameter (Gain) of the fluorescence detection unit. Different detection projects may require different photomultiplier tube voltages or photodiode amplification factors. The system uses the gain setting parameter to convert the collected original reaction product fluorescence signal into a standard signal value under a unified reference, eliminating the numerical difference caused by different hardware amplification factors.

[0078] Subsequently, the system needs to solve the problem of background noise brought by sample matrix. Different types of samples such as blood, saliva, tissue homogenate, etc. have huge differences in their own spontaneous fluorescence intensity. The system identifies the biological matrix category to which the sample to be tested belongs according to user input or kit type. For example:

[0079] If the biological matrix category is the first matrix type, such as a whole blood sample, the hemoglobin in the sample has strong light absorption and spontaneous fluorescence characteristics. The system will subtract the background noise value defined by the first proportion coefficient from the standard signal value.

[0080] If the biological matrix category is the second matrix type, such as a saliva sample, the matrix is relatively clear, and the background noise value defined by the second proportion coefficient is subtracted from the standard signal value.

[0081] According to the transmission principle of physical optics, the optical transmittance of the first matrix type (whole blood) is significantly lower than that of the second matrix type (saliva), and its background noise contribution is greater, so the value of the first proportion coefficient is set to be greater than the value of the second proportion coefficient. For example, a whole blood sample may deduct 20% of the baseline background, while a saliva sample only deducts 5%. This differential deduction strategy ensures that the detection results of different sample sources are comparable.

[0082] S5, risk classification evaluation and health management step

[0083] After a series of precise physical and algorithmic processing, the system finally obtains accurate and reliable target gene expression, usually expressed in copies per milliliter or relative expression fold. The last step is to convert this biological data into a clinically understandable risk assessment conclusion.

[0084] For example, the system inputs the target gene expression into a pre-set multi-dimensional clinical risk assessment model. This model is no longer a single indicator judgment, but a comprehensive scoring system combining gene level and patient individual characteristics. The judgment logic of the multi-dimensional clinical risk assessment model is as follows:

[0085] The system first sets two key clinical thresholds: a low risk determination threshold (such as 10 copies per milliliter) and a high risk determination threshold (such as 1000 copies per milliliter).

[0086] Case one: when the target gene expression is lower than the low risk determination threshold, and there is no clinical risk factor in the patient's medical record, the system determines it as a low risk level. This means that the subject has very low tumor burden in the body and does not need special intervention.

[0087] Case two: when the target gene expression is between the low risk determination threshold and the high risk determination threshold, or although the expression is low but there is at least one clinical risk factor, the system determines it as a medium risk level. This suggests that there may be early lesions or recurrence risk, and it is recommended to shorten the review period.

[0088] Scenario 3: When the expression level of the target gene is higher than the high-risk threshold, or when multiple clinical risk factors are present, the system classifies it as high-risk. This usually indicates a high tumor burden or the presence of significant driver gene mutations, and immediate imaging confirmation or drug intervention is recommended.

[0089] Clinical risk factors are an important dimension of the model, including but not limited to the patient's medical history records (such as whether there is a history of cancer, radiotherapy or chemotherapy) and family genetic characteristics (such as whether immediate family members have a history of related cancers). By introducing these factors, the model can more comprehensively assess the patient's true health threat and avoid missed diagnoses caused by relying solely on genetic data.

[0090] Finally, in order to achieve a closed loop from detection to service, the method in this embodiment also includes abnormal interruption handling and follow-up management steps.

[0091] On the one hand, regarding the stability of equipment operation, the system continuously monitors the hardware status throughout the execution of steps S2 to S4. If a hardware fault signal is detected, such as a temperature control sensor open circuit or a sudden drop in battery voltage, the system will automatically capture the temperature control data curve and raw fluorescence data within a preset time period before the fault signal occurs, generate an encrypted log file, and transmit it to the remote operation and maintenance server via wireless network. This helps technicians quickly locate the cause of the fault, whether it is insufficient equipment consumables, harsh environment, or hardware damage.

[0092] On the other hand, regarding patient health management, after the tumor risk level is output in step S5, the system will automatically calculate the recommended next follow-up date based on that level. For example, a high-risk patient is advised to have a follow-up examination in one month, while a low-risk patient is advised to have a follow-up examination in one year. The system will send follow-up reminder information, including the follow-up date and precautions, to the user's mobile device, such as a mobile app or SMS, and simultaneously synchronize it with the hospital's electronic medical record system to ensure that patients receive timely follow-up management, truly realizing an integrated service of portable testing, risk assessment, and health management.

[0093] In summary, this embodiment successfully ported the sophisticated and complex gene testing process to a portable platform by introducing environmental awareness, dynamic compensation, physical interference elimination, and multidimensional assessment strategies at each stage, and provided clinically valuable tumor risk assessment results while ensuring data accuracy.

[0094] Example 2:

[0095] This embodiment focuses on how to ensure the uniformity and accuracy of the temperature field within a microfluidic reaction system through an innovative environmental temperature difference compensation sub-step in tumor risk assessment methods, especially when portable devices are used in complex and variable external environments.

[0096] Considering that portable devices are often used outdoors, in ambulances, or in primary clinics lacking constant temperature conditions, drastic changes in external ambient temperature can cause significant thermal interference to the miniature reaction chamber, especially the temperature gradient caused by edge effects. Therefore, before controlling the reaction chamber according to temperature control timing parameters, i.e., before entering the denaturation stage, this embodiment also creatively performs an environmental temperature difference compensation sub-step. This step, as a key pre-step in step S2, aims to solve the edge thermal collapse effect that is prone to occur in miniaturized reaction chambers under low-temperature environments.

[0097] In summary, the system first acquires the current external ambient temperature using an external ambient temperature sensor. Then, the main control processor calculates the difference between the current set baseline value for the first temperature range (e.g., 95 degrees Celsius) and the external ambient temperature (e.g., 10 degrees Celsius), recording this difference as the ambient temperature difference. The system internally presets a temperature difference threshold, for example, 15 degrees Celsius. If the calculated ambient temperature difference exceeds this preset threshold, it indicates that the device is in a low-temperature environment with high heat loss. Conventional uniform heating strategies would result in excessively low temperatures at the chip edges, affecting reaction efficiency. In this case, the system automatically activates an asymmetric temperature control mode. In asymmetric temperature control mode, the regional temperature control units no longer apply uniform power to all heating films; instead, they set the target temperature at the edge of the reaction chamber to be higher than the target temperature at the center. The increase in the target temperature at the edge relative to the target temperature at the center is not arbitrarily set but is positively correlated with the logarithmic function of the ambient temperature difference. This non-linear compensation strategy conforms to the physical laws of heat conduction, effectively offsetting the attenuation of edge heat effects caused by the low-temperature environment, ensuring a highly uniform fluid temperature throughout the reaction chamber.

[0098] For example, after the entire detection process is started, the system first completes the parameter configuration in step S1. Then, before officially entering the denaturation stage in step S2, the control program automatically inserts and executes an environmental temperature difference compensation sub-step. The timing of this step is crucial because the denaturation stage typically requires temperatures as high as around 95 degrees Celsius, at which point the temperature difference between the reaction chamber and the external environment is greatest, and heat loss is most severe. If compensation is not planned in advance, a significant temperature gradient can easily form inside the reaction solution, leading to a decrease in amplification efficiency or even failure.

[0099] Specifically, the first step in this sub-step is to acquire the ambient temperature. The system calls upon a high-precision thermistor or digital temperature sensor integrated into the device casing or air inlet to read the current ambient temperature data in real time. For example, when the device is used outdoors in winter, the reading may be as low as 5 degrees Celsius; while when used indoors in summer, the reading may be 25 degrees Celsius. Next, the system processor reads the temperature control parameters loaded in step S1 and locks in the target temperature for the upcoming denaturation stage, i.e., the baseline value of the currently set first temperature range, such as 95 degrees Celsius. The processor calculates the difference between the currently set baseline value of the first temperature range and the ambient temperature through subtraction, and records this result as the ambient temperature difference. This ambient temperature difference directly reflects the intensity of the heat load the system will soon face.

[0100] Next, the system enters the judgment logic stage. The system internally stores a preset temperature difference threshold. This threshold is a critical point calibrated based on extensive thermodynamic simulations and measured data, for example, set to 50 degrees Celsius. If the calculated ambient temperature difference is less than or equal to this threshold, it indicates that the external environment is relatively mild, and conventional PID control is sufficient to maintain temperature uniformity. However, if the ambient temperature difference exceeds the preset threshold—for example, heating from 5 degrees Celsius to 95 degrees Celsius, resulting in a temperature difference of 90 degrees Celsius, far exceeding 50 degrees Celsius—this means that simply relying on uniform heating will not be able to offset the edge heat dissipation. At this point, the system determines that intervention is necessary and immediately activates the asymmetric temperature control mode.

[0101] In asymmetric temperature control mode, the system breaks away from the traditional control strategy of treating all heating areas equally. The heating array beneath the microfluidic chip is logically divided into a central region and an edge region. The control algorithm sets the target temperature of the edge region of the reaction chamber to be higher than that of the central region. The physical basis for this strategy is that the central region of the reaction chamber is surrounded by the surrounding fluid and substrate, resulting in a longer heat dissipation path and less heat loss; while the edge region is directly adjacent to the chip's casing or air interface, resulting in rapid lateral heat conduction. By artificially increasing the target temperature of the edge region—for example, setting it to 95 degrees Celsius at the center and 96.5 degrees Celsius at the edge—an additional "thermal protection ring" can be established at the edge to compensate for the heat loss due to lateral heat dissipation, thus ensuring that the temperature field ultimately transferred to the interior of the reaction liquid exhibits an ideally flat distribution.

[0102] To achieve precise compensation and prevent over-compensation leading to boiling of the edge reaction liquid or under-compensation causing edge amplification failure, this embodiment rigorously quantifies the temperature increase. Specifically, the increase in target temperature in the edge region relative to the target temperature in the center region is positively correlated with the logarithmic function of the ambient temperature difference. This mathematical model is chosen based on the nonlinear characteristics of heat conduction: as the temperature difference increases, the heat dissipation rate does not simply increase linearly, but is affected by changes in boundary layer thermal resistance. The growth characteristic of the logarithmic function—rapid initial growth followed by a gradual plateau—perfectly matches this physical process, ensuring sufficient compensation under drastic temperature differences while preventing the compensation value from diverging infinitely under extreme temperature variations and damaging the heating device.

[0103] For example, this mathematical relationship can be implemented using the following specific engineering formula: Let the target temperature of the edge region be... The target temperature (i.e., the baseline value) in the central area is The ambient temperature difference is The formula for setting the target temperature in the edge region is:

[0104] ;

[0105] in, Represents the natural logarithm operation; This is the thermal compensation coefficient, whose value depends on the thermal conductivity and thickness of the microfluidic chip substrate material. For example, for a 2 mm thick PMMA chip, It may take the value 0.8; To correct the bias, the domain of the logarithmic function is adjusted to ensure smooth calculations near the temperature difference threshold.

[0106] Calculated using this formula Always greater than And with The increase is not linear. For example, when the ambient temperature difference increases from 60 degrees Celsius to 80 degrees Celsius, the compensation value may increase from 1.5 degrees Celsius to 1.8 degrees Celsius. This fine adjustment ensures that no matter how cold the external wind is, the biochemical reactions inside the microfluidic chip are always in a gentle breeze and in the optimal thermodynamic environment.

[0107] Furthermore, this asymmetric temperature control mode is not static but dynamically adjustable. In the subsequent stages of step S2, i.e., when switching from the denaturation stage (e.g., 95 degrees Celsius) to the annealing stage (e.g., 60 degrees Celsius), the new ambient temperature difference decreases as the target temperature drops. The system then re-executes the above calculation process: acquiring the external ambient temperature and recalculating the difference between the current set baseline value for the second temperature range and the external ambient temperature. If the new temperature difference falls below the preset temperature difference threshold, the system automatically exits the asymmetric temperature control mode and reverts to the lower-energy-consumption conventional symmetric control mode; if it still exceeds the threshold, the system recalculates and updates the boost value for the edge region based on the new temperature difference value. This dynamic following strategy not only ensures temperature accuracy throughout the process but also maximizes the battery life of portable devices.

[0108] like Figure 5 The simulation results of the environmental temperature difference compensation mechanism are presented, demonstrating the difference in two-dimensional temperature field distribution of the microfluidic chip under conventional temperature control mode and asymmetric temperature control compensation mode.

[0109] Figure 5 Darker colors indicate lower temperatures, while lighter colors indicate higher temperatures. Figure 5 The middle left figure shows the chip thermal distribution under the standard symmetrical heating mode. The center temperature is set to 95 degrees. However, due to severe heat dissipation in the edge area, a clear cooling trend from the center to the edge can be observed in the figure. The edge temperature is nearly 10 degrees lower than the center, forming a clear temperature gradient. This non-uniformity can easily affect the efficiency and consistency of the amplification reaction.

[0110] Figure 5 The right-middle figure shows the temperature field distribution under the asymmetric temperature control compensation strategy introduced in this embodiment. Under the same heating target, the system introduces a heating strategy based on the ambient temperature difference for the chip edge region, which automatically increases the target temperature of the edge region by one to two degrees. Figure 5 The color at the edges of the display becomes noticeably lighter, and the temperature rises closer to the center area, making the temperature field of the entire chip area tend to be flatter.

[0111] This figure visually illustrates the technical approach of compensating for heat loss in low-temperature environments by increasing the temperature in the edge region. This significantly reduces the temperature gradient, ensuring that the amplification system is within the appropriate reaction range at different locations, and effectively improving the detection stability of portable devices in outdoor or low-temperature scenarios.

[0112] Furthermore, to ensure the effective implementation of the environmental temperature difference compensation sub-step, the system also incorporates corresponding hardware design. The heating module of the microfluidic reaction chamber is physically divided into an independent central heating zone and an edge heating zone, each controlled by an independent power drive circuit (PWM driver). Simultaneously, to prevent unnecessary heat penetration from the edge high temperature to the central region, tiny air insulation channels or anisotropic thermally conductive materials are physically integrated between the two heating zones. When the asymmetric temperature control mode is activated, the PWM duty cycle of the edge heating zone is significantly higher than that of the central heating zone, and the increment of this duty cycle is directly converted from the temperature difference calculated by the aforementioned logarithmic function through a PID algorithm. This deep hardware-software coupling design ensures that algorithm commands can be converted into physical heat energy without delay, achieving millisecond-level precise control of the microscopic reaction environment.

[0113] In summary, this embodiment addresses the edge effect problem faced by portable nucleic acid testing devices in non-laboratory environments, especially in low-temperature outdoor environments, by introducing an environmental temperature difference compensation mechanism based on a logarithmic model. It no longer relies on a bulky external insulation layer but actively reconstructs the thermal field distribution through intelligent algorithms, ensuring that every microliter of reaction solution can be efficiently amplified at the accurate temperature, thereby greatly improving the reliability and reproducibility of tumor risk assessment results.

[0114] Example 3:

[0115] This embodiment aims to elaborate on how to solve the signal crosstalk problem, which is difficult to avoid in physical optics, through software algorithm innovation in the high-density parallel detection scenario of microfluidic chips. Because the distance between adjacent channels on high-throughput microfluidic chips is extremely close, scattered light from strong positive channels can easily contaminate neighboring weak signal channels, causing false positives. Therefore, before obtaining the final target gene expression level, this embodiment also introduces a deconvolution step to address optical crosstalk between adjacent channels.

[0116] In summary, based on the internally stored chip layout and channel mapping, the system determines the target channel being processed and its adjacent channels in spatial geometry, i.e., physically adjacent channels. The system acquires the signal strength values ​​of the adjacent channels and the target channel, respectively, and calculates their ratio. If the ratio exceeds a preset crosstalk threshold (e.g., the neighboring signal is more than 10 times stronger than the target signal), a significant risk of optical crosstalk is identified. At this point, the system performs a deconvolution operation on the signal strength value of the target channel. This operation is not a complex Fourier transform, but a correction based on the principle of linear superposition: subtracting the product of the adjacent channel's signal strength value and a preset optical coupling coefficient from the target channel's signal strength value. This optical coupling coefficient is a constant pre-calibrated at the factory based on the physical properties of the microfluidic chip material (such as PMMA or PDMS), channel spacing, and the depth of the isolation trenches. Through this step, the system can extract the true biological signal from the mixed signal, further improving the accuracy of quantification.

[0117] The application of this step in this embodiment enables the system to maintain clinical-grade detection specificity while pursuing the ultimate in chip miniaturization and channel density, effectively preventing false positives caused by the NIMBY effect.

[0118] For example, in portable rapid single-gene expression detection systems, to increase throughput per detection, the reaction channels on microfluidic chips are typically designed to be extremely compact, with physical spacing between adjacent channels potentially only a few hundred micrometers or even tens of micrometers. Although air-isolation channels or light-shielding coatings have been incorporated into the hardware design, in high-sensitivity fluorescence detection, scattered light emitted from strong positive channels can still penetrate the transparent chip substrate and partially spill into adjacent channels. When an adjacent channel happens to be a weak positive or negative sample, this spilled stray light (i.e., crosstalk signal) will superimpose on the true signal, leading to artificially high readings and potentially misleading subsequent risk assessment models. To mathematically eliminate this interference, the system must enforce an adjacent channel optical crosstalk deconvolution step before entering the final data output stage, i.e., before obtaining the target gene expression level.

[0119] Specifically, the first stage of this sub-step is building topology awareness. The system's main control module first reads the chip's model identifier and loads the corresponding geometric layout file. Based on the system's channel mapping relationship, the algorithm can accurately identify the spatial coordinates of each physical channel. Accordingly, the system can determine the target channel being processed and its spatially adjacent channels. The so-called "adjacent channels" not only refer to channels with adjacent physical numbers, such as channel N and channel N+1, but more importantly, channels that have an optically visible path to the target channel in the two-dimensional spatial layout. For example, in a serpentine chip arrangement, two channels with significantly different physical numbers may be spatially adjacent. The system traverses the adjacency matrix to lock down a list of all neighboring objects that pose a potential source of light pollution to the current target channel.

[0120] Next, the system enters the data acquisition and preliminary evaluation phase. At the same time point, the system controls the photoelectric detection unit to acquire the signal strength values ​​of adjacent channels and the target channel. The signals acquired here are typically relative fluorescence units (RFU) values ​​that have already undergone pre-gain adjustment and baseline subtraction. To determine whether crosstalk exists and whether it is severe enough to require intervention, the algorithm performs a crucial comparison operation: calculating the ratio of the signal strength value of the adjacent channel to the signal strength value of the target channel.

[0121] Optionally, this ratio not only measures the difference in absolute intensity but also reflects the risk of signal-to-noise ratio degradation. Assuming the target channel is negative (extremely low signal) and the adjacent channel is strong positive (extremely high signal), this ratio will become extremely large. The system internally has a preset crosstalk threshold. This threshold is an empirical value set based on the optical attenuation characteristics of the chip material, for example, set to 10. If the ratio exceeds the preset crosstalk threshold, i.e., the signal strength of the adjacent channel is more than 10 times that of the target channel, the system determines that the current target channel is subject to significant and overwhelming optical crosstalk. In this case, any tiny light leakage may constitute a major component of the target channel signal and must be corrected. Conversely, if the ratio is below the threshold, it indicates that the signal strengths of the two channels are comparable, or the target channel's signal is strong enough. In this case, the weak crosstalk can be ignored, and the algorithm chooses to skip the correction to save computational resources and preserve the original data characteristics.

[0122] Once a correction is determined, the system immediately activates its core algorithm engine to perform deconvolution on the signal strength value of the target channel. This deconvolution is not traditional image restoration, but rather a reverse solution process based on the principle of linear superposition. Its physical assumption is that the total signal received by the detector equals the true fluorescence signal of the target channel itself plus the leakage signal from adjacent channels. Therefore, to restore the true signal, the leakage component must be subtracted from the total signal.

[0123] For example, the deconvolution operation is configured to perform the following mathematical formula operation:

[0124] ;

[0125] This can be understood as follows: for a single adjacent interference source, the product of the signal strength value of the adjacent channel and the preset optical coupling coefficient needs to be subtracted from the signal strength value of the target channel.

[0126] in, : Represents the corrected, crosstalk-free true signal strength of the target channel; : Represents the original signal strength of the target channel containing interference, measured before correction; : Represents the signal strength value of the adjacent channel that is the source of interference; : Represents the preset optical coupling coefficient.

[0127] Regarding the preset optical coupling coefficient This coefficient is the cornerstone of achieving high-precision correction in this embodiment. It is a dimensionless physical quantity that characterizes the lateral penetration capability of light under specific chip materials and geometries. For example, if... A value of 0.05 means that for every 100 units of fluorescence generated in an adjacent channel, 5 units leak into the target channel. This coefficient is typically not calculated in real-time at the testing site, but rather pre-determined and stored in the system firmware through a rigorous factory calibration process. The calibration process is as follows: a high concentration of fluorescent dye (strong positive source) is injected into one channel of the chip, while pure water (zero signal source) is injected into the adjacent channel. The ratio of the signal intensity measured in the pure water channel to the signal intensity measured in the dye channel is defined as the optical coupling coefficient. For chips made of polymethyl methacrylate (PMMA) with a wall thickness of 200 micrometers and no blackening treatment on the surface, this coefficient may be between 0.03 and 0.08. The system automatically calls the corresponding [chip batch number] based on the chip batch number. The value is used in the calculation.

[0128] Optionally, when performing deconvolution operations, the algorithm also needs to consider the complex situation of multi-source interference superposition. If a target channel is sandwiched between two strong positive channels, for example, channel 2 is simultaneously interfered with by channels 1 and 3, the above operation logic will be extended to an accumulation subtraction mode. That is, the crosstalk components contributed by the left neighbor and the crosstalk components contributed by the right neighbor are subtracted from the target channel signal respectively. At this time, the subtraction term in the formula becomes the sum of the products of all adjacent channel signals that meet the judgment threshold and their respective coupling coefficients. This comprehensive coverage strategy ensures that even under the most unfavorable sample arrangement, the weak signal sample located in the center can still be restored to its true low expression or no expression state.

[0129] Furthermore, to enhance the robustness of the algorithm, this embodiment also introduces a boundary constraint mechanism. Mathematically, deconvolution is essentially a subtraction operation. In extreme cases, such as when the calibrated coupling coefficient is slightly larger than the actual value, or when the sensor contains random thermal noise, the calculation result... Negative values ​​may occur. This is physically unreasonable because fluorescence intensity cannot be negative. Therefore, the deconvolution operation also includes a post-processing logic: if the calculation result is less than zero, it is forced to zero or the system's background noise level. This constraint prevents invalid data caused by overcorrection and ensures that the output target gene expression level always falls within the legal physical domain.

[0130] Furthermore, the effectiveness of the deconvolution sub-step for adjacent channel optical crosstalk can be intuitively demonstrated through comparison of actual detection data. For example, in a screening for EGFR gene mutations in lung cancer, channel A is a strongly positive mutation sample with a set signal value of 50,000, while the adjacent channel B is a wild-type negative sample with a set true signal value of 200. Due to optical crosstalk, before executing this sub-step, the measured signal of channel B is as high as 2200, which includes approximately 2000 of crosstalk light. This exceeds the positive determination threshold, such as 1000, causing channel B to be misjudged as a false positive. After performing the deconvolution operation of this embodiment, the system identifies that the interference ratio of channel A to channel B is as high as 250 (exceeding the threshold of 10), and calculates using a preset coupling coefficient of 0.04: 2200 minus 50,000 multiplied by 0.04, resulting in a corrected value of 200. The corrected data accurately falls back to the negative range, thus avoiding a serious medical misdiagnosis.

[0131] In summary, this embodiment successfully constructs a virtual optical isolation wall at the software level by building a precise channel topology mapping, intelligently identifying crosstalk risks using a ratio judgment method, and performing mathematical deconvolution based on a physical optics coupling model. This technique significantly reduces the complexity and cost of microfluidic chip hardware fabrication without requiring expensive physical light-shielding processes. Simultaneously, it significantly improves the detection specificity and reliability of the system when processing mixed samples of high and low concentrations, making it an indispensable key technical feature of portable high-throughput gene detection systems.

[0132] Example 4:

[0133] The portable single-gene expression rapid detection system of this embodiment is physically constructed as a handheld or portable host device, internally integrating four core hardware modules: a multi-channel microfluidic chip, a regional temperature control unit, a fluorescence detection unit, and a main control module. These four modules work collaboratively through an internal high-speed data bus and a precision mechanical interface, jointly forming the hardware foundation for performing dynamic reaction control, amplification efficiency closed-loop compensation, and multi-dimensional risk assessment. The system's workflow is as follows: Figure 6 As shown:

[0134] First, as the physical carrier for biochemical reactions, the multichannel microfluidic chip employs a disposable cartridge design, allowing for convenient plug-and-play use on-site. The chip's substrate material is preferably a polymer with high optical transparency and low autofluorescence background, such as polymethyl methacrylate (PMMA) or cyclic olefin copolymer (COC). The chip's microstructure integrates several parallel reaction channels. These channels are typically arranged in an array, with the volume of each channel precisely controlled in the microliter range, for example, from 10 μL to 30 μL, to accommodate precious clinical samples.

[0135] Specifically, to achieve the dynamic enzyme injection compensation function in this embodiment, the chip features an innovative improvement in its fluid design: each location fluidly connected to a reaction channel is equipped with an independent bypass chamber. This bypass chamber is isolated from the main reaction channel by a micro-valve or burst valve controlled by a micro-actuator. Before the chip leaves the factory, the bypass chamber is pre-lyophilized or encapsulated with a high concentration of enzyme activator, such as a highly active DNA polymerase supplement or a magnesium ion cofactor solution. This design allows the system to physically inject the replenishment packet into the reaction system during detection, based on real-time feedback of the amplification efficiency.

[0136] Furthermore, to address the thermal crosstalk issue caused by the high-density channel arrangement and ensure the temperature independence of each reaction channel, this embodiment incorporates air isolation grooves between adjacent reaction channels for thermal insulation. These air isolation grooves are deep trench structures formed on the substrate using laser etching or injection molding processes. Utilizing the extremely low thermal conductivity of still air, they create a physical thermal barrier, effectively cutting off the lateral heat conduction path between adjacent heating areas.

[0137] Secondly, the system's thermal cycling control is performed by a regional temperature control unit. To adapt to the complex outdoor ambient temperature and address the temperature field distortion caused by the edge effect of the microfluidic chip, this unit employs a non-uniform heating hardware layout. Specifically, it includes a thin-film heater corresponding to the central region of the reaction channel and a thermal insulation material layer corresponding to the edge region. The thin-film heater can use an indium tin oxide (ITO) transparent conductive film or metal photolithography resistive circuitry, tightly attached to the bottom of the chip's reaction area, providing rapid and precise thermal energy input. The thermal insulation material layer surrounds the non-reaction area around the chip, using materials such as aerogel or vacuum insulation panels to minimize heat loss to the surrounding environment.

[0138] More importantly, this zoned temperature control unit does not simply perform a constant temperature operation, but rather controls the reaction chamber to sequentially operate within different temperature zones according to preset temperature control timing parameters to execute the amplification reaction. Specifically, it is configured to have environmental adaptability, i.e., to perform environmental temperature difference compensation logic. For example, the system has a built-in high-precision environmental temperature sensor to acquire the external ambient temperature. The main control module reads the sensor data in real time and calculates the difference between the current set reference value of the first temperature zone and the external ambient temperature, recording this difference as the environmental temperature difference. The system internally presets a temperature difference threshold (e.g., 15 degrees Celsius). If the environmental temperature difference exceeds the preset threshold, it indicates that the device is in a low-temperature environment where heat loss is highly likely. In this case, the zoned temperature control unit automatically activates the asymmetric temperature control mode. In asymmetric temperature control mode, the system no longer applies a uniform voltage to the entire heating surface, but instead uses independent drive circuits to set the target temperature of the reaction chamber's edge regions to be higher than the target temperature of the center regions. This gradient temperature setting aims to compensate for the physical difference between faster heat dissipation at the edges and slower heat dissipation at the center. Specifically, the increase in target temperature at the edge region relative to the target temperature at the center region is not arbitrarily set, but is positively correlated with the logarithmic function value of the ambient temperature difference. This specific mathematical relationship ensures that the compensation intensity exhibits a non-linear growth in accordance with thermodynamic laws as the severity of the environment increases, thereby maintaining extremely high temperature uniformity at the fluid level inside the chip.

[0139] Furthermore, signal acquisition is handled by the fluorescence detection unit. This unit has an optical readout window corresponding to the reaction channel location, typically located above or to the side of the chip, avoiding obstruction from the heater. It includes an excitation source (such as an LED array) and a photodetector (such as a photodiode or CMOS sensor) to acquire fluorescence intensity signals at a preset sampling frequency during the amplification reaction, providing a data stream for real-time calculation of amplification efficiency. Simultaneously, it is also responsible for acquiring the final fluorescence signal of the reaction product after the reaction, for final quantitative analysis. To eliminate differences in photoelectric conversion between different batches of hardware, this unit also features a programmable gain amplification function, capable of adjusting the signal amplification factor in response to instructions from the main control module.

[0140] Finally, as the core brain of the system, the main control module integrates a high-performance microprocessor and embedded algorithm logic. It is electrically connected to the regional temperature control unit and fluorescence detection unit through wires or flexible circuit boards on the PCB board, and is configured to execute a series of complex logical operations to achieve deep integration of software and hardware.

[0141] During the detection initiation phase, the main control module responds to the target gene confirmation command by retrieving the temperature control timing parameters and amplification efficiency baseline threshold that match the target gene from the pre-set database. This step ensures that subsequent physical actions (heating temperature and duration) and software judgment criteria (efficiency threshold) are specifically optimized for the current gene to be tested (such as EGFR or KRAS).

[0142] During the reaction phase, the main control module processes sensor data in parallel and calculates the fluorescence intensity change rate based on the time dimension. Subsequently, it inputs this change rate into the amplification efficiency prediction model to obtain the predicted amplification efficiency value. As described in Example 1, the calculation logic of this model is stored in the main control module's memory and satisfies the following relationship:

[0143] ;

[0144] in, This is a predicted value for amplification efficiency. The rate of change of fluorescence intensity This is a target-specific factor. The main control module will calculate the... The amplification efficiency is compared to a baseline threshold. If the result is below the threshold, the main control module immediately sends an electrical pulse signal to drive the microvalve on the microfluidic chip, controlling the release of the enzyme activator into the reaction chamber. Simultaneously, the timer in the temperature control program is modified to extend the duration of the current reaction stage. This series of actions enables proactive intervention and salvage of the biochemical reaction process.

[0145] During the data processing phase, after the reaction is complete, the main control module performs signal correction processing, including gain normalization and background subtraction, on the collected fluorescence signals of the reaction products to obtain the target gene expression level. Specifically, it selects a preset first or second proportionality coefficient for background subtraction based on the sample matrix type (e.g., whole blood or saliva). Subsequently, the target gene expression level is input into a multidimensional clinical risk assessment model, which outputs the tumor risk level and displays the result on the system screen or sends it to the cloud.

[0146] Optionally, to improve the robustness of the system, the main control module also integrates several auxiliary function sub-units:

[0147] Sample physical property pre-detection unit: This unit includes sensors, such as optical transmission sensors and flow rate sensors, used to detect the transmittance and flow rate parameters of the injected sample. The main control module compares these physical parameters with preset acceptable ranges. If the parameters are abnormal, an anomaly interception logic is triggered to prevent the temperature control unit from starting, thereby avoiding detection failures due to sample quality issues (such as severe hemolysis or coagulation).

[0148] Bubble Interference Detection and Processing Unit: This unit is configured to perform linear regression fitting on the fluorescence intensity signal and calculate fluctuation characteristic values. If the fluctuation characteristic value exceeds the stability threshold, the main control module determines that bubble interference exists, suspends the efficiency prediction process, and controls the temperature control unit to trigger a temperature oscillation debubbling operation. This operation removes microbubbles in the optical path through rapid and minute temperature fluctuations, utilizing thermo-fluid pulsation physics.

[0149] Optical crosstalk deconvolution unit: To address optical crosstalk that may occur during multi-channel parallel detection, this unit is used to identify adjacent channels based on channel mapping relationships. It monitors the signal strength ratio of adjacent channels in real time and performs deconvolution operations on the target channel signal when crosstalk exceeds a threshold. The specific operation logic is as follows: subtract the product of the adjacent channel signal and the preset optical coupling coefficient from the target channel signal to reconstruct the true biological signal.

[0150] Abnormal Interruption and Follow-up Management Unit: This unit is responsible for the full lifecycle management of the system. It generates log files and uploads them to the server when hardware failures are detected to ensure the maintainability of the equipment. At the same time, it calculates the follow-up date after outputting the tumor risk level and sends reminder information to the user terminal, realizing a closed loop from detection to health management services.

[0151] In summary, the portable single-gene expression rapid detection system of this embodiment, through the fine flow channel design of the microfluidic chip, the asymmetric compensation strategy of the temperature control unit, and the multiple intelligent algorithm units integrated in the main control module, successfully overcomes the challenges faced by portable devices in complex environments, such as temperature unevenness, signal interference, and sample differences, providing a solid hardware guarantee for achieving accurate tumor risk assessment.

[0152] Example 5:

[0153] Corresponding to the above embodiments, the present invention also proposes an electronic device.

[0154] like Figure 7 The diagram shows a structural schematic of an electronic device according to the present invention. The electronic device 100 includes a processor 101 and a memory 103. The processor 101 and the memory 103 are connected, for example, via a bus 102. Optionally, the electronic device 100 may further include a transceiver 104. It should be noted that in practical applications, the transceiver 104 is not limited to one unit, and the structure of this electronic device 100 does not constitute a limitation on the embodiments of the present invention.

[0155] Processor 101 may be a CPU, a general-purpose processor, a DSP, an ASIC, an FPGA, or other programmable logic device, transistor logic device, hardware component, or any combination thereof. It may implement or execute the various exemplary logic blocks, modules, and circuits described in connection with this disclosure. Processor 101 may also be a combination that implements computational functions, such as including one or more microprocessor combinations, a combination of a DSP and a microprocessor, etc.

[0156] Bus 102 may include a pathway for transmitting information between the aforementioned components. Bus 102 may be a PCI bus or an EISA bus, etc. Bus 102 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 7 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0157] The memory 103 stores a computer program corresponding to the tumor risk assessment method of the portable single-gene expression rapid detection system of the above embodiments of the present invention. This computer program is executed under the control of the processor 101. The processor 101 executes the computer program stored in the memory 103 to implement the content shown in the aforementioned method embodiments.

[0158] Among them, electronic devices 100 include, but are not limited to: mobile terminals such as laptops and PADs (tablet computers) and fixed terminals such as desktop computers. Figure 7 The electronic device 100 shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of the present invention.

[0159] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A tumor risk assessment method using a portable single-gene expression rapid detection system, characterized in that, Includes the following steps: S1. In response to the target gene confirmation command, retrieve the temperature control timing parameters and amplification efficiency benchmark threshold that match the target gene from the preset database; S2. According to the temperature control timing parameters, the reaction chamber is controlled to be in different temperature ranges in sequence to perform the amplification reaction. During the reaction, the fluorescence intensity signal in the reaction chamber is collected according to the preset sampling frequency, and the fluorescence intensity change rate is calculated based on the time dimension. S3. Substitute the fluorescence intensity change rate into the amplification efficiency prediction model to obtain the current amplification efficiency prediction value; compare the amplification efficiency prediction value with the amplification efficiency benchmark threshold; if the amplification efficiency prediction value is lower than the amplification efficiency benchmark threshold, generate an enzyme activation command; the enzyme activation command is used to control the release of enzyme activator into the reaction chamber and extend the duration of the current reaction stage to compensate for the amplification rate of the reaction system. S4. After the amplification reaction is completed, collect the fluorescence signal of the final reaction product; The fluorescence signal of the reaction product is subjected to signal correction processing including gain normalization and background subtraction to obtain the expression level of the target gene. S5. Input the target gene expression level into the multidimensional clinical risk assessment model and output the tumor risk level.

2. The method according to claim 1, characterized in that, In step S2, the step of controlling the reaction chamber to be in different temperature ranges sequentially according to the temperature control timing parameters specifically includes: Denaturation stage: The temperature of the reaction chamber is controlled within the first temperature range for a first preset duration; Annealing extension and monitoring stage: The temperature of the reaction chamber is controlled in the second temperature range for a second preset duration; the acquisition of fluorescence intensity signal in step S2 and the execution of enzyme activation command in step S3 are both completed in this stage; Stable detection phase: Adjust the temperature of the reaction chamber to the third temperature range and continue for the third preset time until step S4 is entered; The temperature value in the first temperature range is higher than the temperature value in the second temperature range, and the temperature value in the second temperature range is higher than the temperature value in the third temperature range.

3. The method according to claim 1, characterized in that, In step S4, the specific execution logic of the signal correction process is as follows: Obtain the current gain setting parameters of the fluorescence detection unit, and use these gain setting parameters to convert the collected reaction product fluorescence signal into a standard signal value; Identify the biological matrix category to which the sample to be tested belongs; If the biological matrix category is the first matrix type, then the background noise value defined by the first scaling factor is subtracted from the standard signal value; If the biological matrix category is the second matrix type, then the background noise value defined by the second scaling factor is subtracted from the standard signal value; Wherein, the optical transmittance of the first matrix type is lower than that of the second matrix type, and the value of the first proportionality coefficient is greater than that of the second proportionality coefficient.

4. The method according to claim 1, characterized in that, In step S5, the judgment logic of the multidimensional clinical risk assessment model is as follows: Set low-risk and high-risk assessment thresholds; When the expression level of the target gene is lower than the low-risk threshold and there are no clinical risk factors, it is judged to be of low risk level; When the expression level of the target gene is between the low-risk threshold and the high-risk threshold, or when at least one clinical risk factor is present, it is determined to be of medium risk level. When the expression level of the target gene is higher than the high-risk threshold, or when multiple clinical risk factors are present, it is determined to be of a high-risk level. The clinical risk factors include past medical history records and family genetic characteristics data.

5. The method according to claim 1, characterized in that, Before step S2, a sample physical characteristic pre-inspection step is also included: Sensors are used to detect the transmittance and flow rate parameters of the injected sample; If the light transmittance parameter exceeds the preset acceptable light transmittance range, or the flow velocity parameter exceeds the preset velocity range corresponding to normal viscosity, then the abnormal interception logic is triggered. The anomaly interception logic is configured to: lock the temperature control unit to prevent the execution of step S2, and generate a prompt message indicating abnormal sample quality.

6. The method according to claim 1, characterized in that, The method also includes abnormal interruption handling and follow-up management steps: During the execution of steps S2 to S4, if a hardware fault signal is detected, the temperature control data and fluorescence data within a preset time period before the fault signal is generated are automatically captured, a log file is generated, and transmitted to the server. After outputting the tumor risk level in step S5, the next follow-up examination date is calculated based on the tumor risk level, and the follow-up examination reminder information is sent to the user terminal.

7. The method according to claim 1, characterized in that, In step S2, while acquiring the fluorescence intensity signal, a bubble interference removal sub-step is also performed: Linear regression fitting was performed on the fluorescence intensity signals collected within the preset time window; Calculate the residual standard deviation of the original signal points relative to the fitted curve, and define the ratio of the residual standard deviation to the signal mean as the fluctuation characteristic value; If the fluctuation characteristic value exceeds the preset stability threshold, it is determined that the current reaction chamber is disturbed by bubbles, the execution of step S3 is temporarily suspended, and a temperature oscillation defoaming operation is triggered. The temperature oscillation defoaming operation is configured to: control the temperature of the reaction chamber to be superimposed with an alternating temperature signal on the basis of the current set temperature, wherein the alternating temperature signal has a preset oscillation frequency and a preset oscillation amplitude, so as to remove bubbles by using thermal fluid pulsation.

8. The method according to claim 3, characterized in that, In step S2, before entering the denaturation stage, an environmental temperature difference compensation sub-step is also performed: Obtain the ambient temperature and calculate the difference between the current set reference value of the first temperature range and the ambient temperature, which is denoted as the ambient temperature difference. If the ambient temperature difference exceeds the preset temperature difference threshold, the asymmetric temperature control mode will be activated. In the asymmetric temperature control mode, the target temperature of the edge region of the reaction chamber is set to be higher than the target temperature of the center region. The increase in target temperature in the edge region relative to target temperature in the center region is positively correlated with the logarithmic function value of the ambient temperature difference.

9. The method according to claim 4, characterized in that, In step S4, before obtaining the target gene expression level, an adjacent channel optical crosstalk deconvolution step is also performed: Based on the system's channel mapping relationship, determine the target channel being processed and its adjacent channels in space; Obtain the signal strength value of the adjacent channel and the signal strength value of the target channel; Calculate the ratio of the signal strength value of the adjacent channel to the signal strength value of the target channel; If the ratio exceeds a preset crosstalk determination threshold, then a deconvolution operation is performed on the signal strength value of the target channel; The deconvolution operation is configured to subtract the product of the signal strength value of the adjacent channel and a preset optical coupling coefficient from the signal strength value of the target channel.

10. A portable rapid detection system for single-gene expression, characterized in that, include: A multi-channel microfluidic chip integrates several parallel reaction channels and a bypass chamber fluidly connected to each reaction channel. Air isolation grooves are provided between the reaction channels for heat insulation. The zoned temperature control unit includes a thin-film heater corresponding to the central region of the reaction channel and a heat insulation material layer corresponding to the edge region. It is used to control the reaction chamber to be in different temperature ranges in sequence according to the preset temperature control timing parameters in order to perform the amplification reaction. The fluorescence detection unit is provided with an optical readout window corresponding to the position of the reaction channel, which is used to collect fluorescence intensity signals at a preset sampling frequency during the amplification reaction and to collect the fluorescence signal of the final reaction product after the reaction is completed. The main control module, electrically connected to the regional temperature control unit and the fluorescence detection unit, is configured to perform the following operations: In response to the target gene confirmation command, the temperature control timing parameters and amplification efficiency benchmark threshold that match the target gene are retrieved from the preset database; The fluorescence intensity change rate is calculated based on the time dimension, and then input into the amplification efficiency prediction model to obtain the amplification efficiency prediction value. This value is then compared with the amplification efficiency benchmark threshold. If the value is lower than the threshold, the enzyme activator is controlled to be released into the reaction chamber and the duration of the current reaction stage is extended. The collected fluorescence signals of the reaction products were subjected to signal correction processing including gain normalization and background subtraction to obtain the expression level of the target gene. The target gene expression level is input into a multidimensional clinical risk assessment model, which outputs the tumor risk level. The sample physical property pre-detection unit includes sensors to detect the transmittance and flow rate parameters of the injected sample. If the parameters are abnormal, the abnormal interception logic is triggered to prevent the temperature control unit from starting. The bubble interference detection and processing unit is configured to perform linear regression fitting on the fluorescence intensity signal and calculate the fluctuation characteristic value. If the stability threshold is exceeded, a temperature oscillation debubbling operation is triggered. The ambient temperature difference compensation unit is used to acquire the external ambient temperature and calculate the difference between it and the set temperature. If the difference exceeds the temperature difference threshold, the asymmetric temperature control mode is activated to make the target temperature in the edge area higher than the target temperature in the center area. The optical crosstalk deconvolution unit is used to identify adjacent channels based on channel mapping relationships and to perform deconvolution operations on the target channel signal when crosstalk exceeds a threshold. The abnormal interruption and follow-up management unit is used to generate log files and upload them to the server when hardware failure is detected, calculate the re-examination date after outputting the tumor risk level, and send reminder information to the user terminal.

Images