Rapid test method for screening short stature in children

Abstract

This application discloses a rapid detection method for screening of dwarfism in children, which belongs to the field of intelligent medical technology. It includes the following steps: Step 1, accurately sampling through capillary effect, and automatically diluting the sample to ensure accurate concentration; Step 2, introducing the collected sample into a multi-channel microfluidic detection chip to achieve parallel quantitative detection of endocrine hormones; Step 3, real-time acquisition and amplification of weak biomolecule recognition signals to effectively improve the signal-to-noise ratio; Step 4, multi-dimensional analysis of the test results to achieve quantitative evaluation of growth and development, body mass index and sexual development indicators and risk prediction of endocrine diseases; Step 5, automatically integrating the analysis results and generating a standardized test report; Step 6, establishing a quality control mechanism throughout the entire detection process, including real-time monitoring of sample quality, tracking of detection process parameters, result consistency evaluation, and full traceability of detection data.

Description

Technical Field

[0001] The present application relates to the field of intelligent medical technology, and more specifically, to a rapid detection method for screening dwarfism in children. Background Art

[0002] Dwarfism, obesity and precocious puberty are common endocrine diseases, usually manifested as obvious abnormalities in height, weight or sexual development of children at a specific age. These conditions may be caused by a variety of factors, including genetics, malnutrition, chronic diseases and endocrine disorders. However, early detection and intervention are the key to improving the symptoms of these diseases and the overall health of children. Therefore, early screening and accurate diagnosis of these diseases are particularly important.

[0003] At present, traditional screening methods mainly rely on the experience of clinicians and physical examinations, which is not only time-consuming and laborious, but also may lead to misdiagnosis and missed diagnosis due to the lack of standardized testing methods. In addition, existing laboratory testing methods such as growth hormone stimulation test, bone age assessment and endocrine hormone level testing, although they can provide certain diagnostic basis, often require complex operating procedures and expensive equipment, and the testing cycle is long, which cannot meet the needs of rapid screening.

[0004] In recent years, with the rapid development of biotechnology and information technology, emerging technologies such as microfluidics, nanotechnology and artificial intelligence have shown great application potential in the field of medical testing. These technologies can achieve high-throughput and high-sensitivity detection of trace biological samples, thus providing an efficient and accurate solution for the early screening of these endocrine diseases.

[0005] In summary, how to develop a fast, accurate, economical and easy-to-operate method for screening children for dwarfism, obesity and precocious puberty has become a technical problem that needs to be solved urgently. Summary of the invention

[0006] In order to overcome a series of defects in the prior art, the purpose of the present application is to provide a rapid detection method for screening dwarfism in children in view of the above problems, including the following steps.

[0007] Step 1: Use a sampling device with integrated micro-blood collection and automatic dilution functions to accurately sample through the capillary effect and automatically dilute the sample to ensure accurate concentration.

[0008] Step 2: The collected samples are introduced into a multi-channel microfluidic detection chip, and parallel quantitative detection of endocrine hormones is achieved through a nano-biorecognition probe array.

[0009] Step 3: Based on high-sensitivity detectors and adaptive signal processing algorithms, weak biomolecule recognition signals are collected and amplified in real time, effectively improving the signal-to-noise ratio and ensuring the accuracy of the detection results.

[0010] Step 4: Based on the deep learning algorithm and combined with the dynamically updated reference database of children's growth and development and endocrine indicators, the test results are analyzed in multiple dimensions to achieve quantitative evaluation of growth and development, body mass index and sexual development indicators and prediction of endocrine disease risks.

[0011] Step 5: Automatically integrate the analysis results to generate a standardized test report, including quantitative test data, personalized growth curve, body mass index curve and sexual development assessment analysis, abnormal risk grading assessment and targeted intervention recommendations.

[0012] Step 6: Establish a quality control mechanism throughout the entire testing process, including real-time monitoring of sample quality, tracking of testing process parameters, result consistency assessment, and full traceability of testing data.

[0013] Furthermore, step 1 includes the following steps.

[0014] Use 75% medical alcohol to thoroughly disinfect the sampling head and dilution chamber of the sampling device, and check the metering accuracy of the device with a standard solution to ensure that the device is in the best working condition.

[0015] Use alcohol cotton pads to disinfect the fingertips of children. After they are dried naturally, use a children's blood collection needle for puncture, and control the depth to 1.8-2.2 mm.

[0016] Relying on the capillary effect, the preset volume of blood sample is automatically aspirated.

[0017] When the blood reaches a preset volume, the buffer solution pre-installed in the dilution chamber is mixed with the blood sample in a precise proportion to form a sample solution with uniform concentration.

[0018] After the dilution process is completed, the mixed liquid sample automatically flows into the collection chamber and is fully mixed to avoid concentration gradient phenomenon.

[0019] Furthermore, step 2 includes the following steps.

[0020] The multi-channel microfluidic detection chip was preheated to 37°C and pretreated with a special buffer to ensure that the hydrophilicity of the inner wall of each microchannel reached the optimal state to prevent the generation of bubbles and ensure the uniform flow of the sample in the channel.

[0021] A microinjector is used to accurately introduce the mixed solution into the injection port of the chip. Under the action of microfluidics, the mixed solution is automatically distributed into multiple parallel detection channels.

[0022] The microfluidic control module on the chip is started to accurately adjust the flow rate and residence time of the sample in each detection channel to ensure that the sample fully contacts and reacts with the nanoprobes in each channel.

[0023] During the reaction, various endocrine hormones specifically bind to the corresponding nanoprobes, causing changes in the physical and chemical properties of the probes and capturing and recording them in real time.

[0024] After the reaction is completed, the detection channel is gently rinsed with a dedicated washing buffer to remove unbound interfering substances.

[0025] The signal acquisition module integrated in the chip can synchronously read the signal intensity of each detection channel and convert it into the specific concentration value of the corresponding hormone level according to the pre-established standard curve.

[0026] Furthermore, the signal intensity of each detection channel is synchronously read through the signal acquisition module integrated in the chip, and converted into a specific concentration value of the corresponding hormone level according to a pre-established standard curve, including the following steps.

[0027] The signal acquisition of all detection channels is synchronously controlled by a multiplexer, and the analog signal is converted into a digital signal using an analog-to-digital converter to eliminate the interference of random noise.

[0028] The relative signal strength value is calculated by comparing the signal amplitude of each channel with the signal difference of the reference channel.

[0029] According to the standard curve equation established by using samples of known concentration, the relative signal intensity value is substituted into the calculation to obtain the actual concentration of the target hormone in the sample to be tested.

[0030] The calculated actual concentration value is verified for validity to ensure the accuracy and reliability of the measurement results.

[0031] Furthermore, step 3 includes the following steps.

[0032] Under completely light-shielded conditions, the baseline noise level of the detector was recorded, and the background signal of the blank sample was collected for subsequent signal processing and correction.

[0033] Turn on the high-sensitivity photodetector and use a high-performance photomultiplier tube to collect the weak hormone-antibody binding signals of each detection channel of the chip in real time.

[0034] A method combining wavelet transform and adaptive Wiener filtering is used to automatically adjust the filtering parameters according to the signal characteristics, effectively removing high-frequency random noise and low-frequency drift, while retaining the effective information of the target signal to the greatest extent.

[0035] The amplification factor is automatically adjusted according to the change of signal strength to ensure that weak signals are amplified while avoiding saturation distortion of strong signals, so that the signal is always kept within the optimal detection range, which can be expressed as: G = G min +(G max -G min )·[A threshold / (A in +A threshold )], where G is the gain factor, i.e. the magnification; G min is the minimum gain, i.e. the lower limit of the gain coefficient; G max is the maximum gain, that is, the upper limit of the gain coefficient; A in is the input signal strength; A threshold It is a preset signal strength threshold used to determine whether a change in signal strength requires gain adjustment.

[0036] Taking into account the signal strength, time characteristics and spatial distribution characteristics, the effective signal is accurately extracted through pattern recognition methods to further improve the signal-to-noise ratio, which is expressed as: SNR optimized =[I signal (T) F signal (ω)·S~ effective (k x ,k y ,k z )·M(S effective ,N)] / P noise , where SNR optimized is the optimized signal-to-noise ratio, that is, the signal-to-noise ratio that is finally improved after pattern recognition and multi-dimensional feature analysis; P noise is the noise power, which means the noise signal in the entire time window The average power within signal (T) is the signal strength, that is, the signal in a specific time window The average intensity within signal (ω) is the spectrum of the signal, that is, the frequency distribution of the signal obtained by Fourier transform, reflecting the distribution of the signal in different frequency components; S~ effective (k x ,k y ,k z） is the spatial spectrum of the effective signal, that is, the frequency component of the effective signal in the spatial domain; M(S effective ,N) is the pattern recognition function, which represents the relationship between the effective signal and the noise signal processed by the pattern recognition method; N is the noise signal; ω is the representation of the signal in the frequency domain; k x ,k y ,k z Represent the frequencies in the X, Y and Z directions respectively.

[0037] Furthermore, step 4 includes the following steps.

[0038] Establish digital files for children, collect growth and development indicators, body mass index, developmental stages and bone age assessment information, and perform standardized processing to ensure that the data is suitable for deep learning model analysis.

[0039] The standardized data is input into the deep learning model, and the key features are extracted through the multi-layer convolution structure and attention mechanism, focusing on analyzing the dynamic changes of endocrine hormone levels and growth and development indicators.

[0040] Utilizing a dynamically updated reference database of children's growth and development and endocrine diseases, transfer learning technology is used to conduct multi-dimensional comparisons of children's data with similar cases to generate quantitative evaluation indicators.

[0041] Use ensemble learning methods to predict the risks of dwarfism, obesity, and precocious puberty, and calculate the risk probability of children developing related diseases in the future.

[0042] Furthermore, the standardized data are input into the deep learning model, and the key features are extracted through the multi-layer convolution structure and attention mechanism, focusing on analyzing the dynamic changes of endocrine hormone levels and growth and development indicators, including the following steps.

[0043] A multi-layer convolutional neural network is designed as the backbone network, which includes multiple layers of convolutional layers, pooling layers and fully connected layers. The convolutional layers are responsible for extracting local features, the pooling layers are responsible for feature dimensionality reduction and main information extraction, and the fully connected layers are responsible for integrating high-dimensional features.

[0044] Attention layers are embedded in key positions of multi-layer convolutional neural networks to automatically learn the importance weights between different features, so as to better focus on the key correspondence between endocrine hormones and various growth and development indicators.

[0045] Recurrent neural networks are used to process time series data to capture the patterns of changes in various indicators over time, with special attention paid to the lagged effects of changes in hormone levels on growth and development.

[0046] The extracted static features and temporal features are multimodally fused through a multi-layer perceptron to deeply analyze the nonlinear relationship between endocrine hormones and growth and development indicators, and a prediction model is established.

[0047] The performance of the prediction model was evaluated through cross-validation, and the prediction results were displayed using visualization technology to analyze the impact of different hormone indicators on growth and development.

[0048] Furthermore, an integrated learning method is used to predict the risks of dwarfism, obesity and precocious puberty, and calculate the risk probability of children developing related diseases in the future, including the following steps.

[0049] Multiple types of machine learning algorithms were selected as base classifiers, and each base classifier was modeled for dwarfism, obesity, and precocious puberty respectively.

[0050] The cross-validation method is used to generate the prediction results of the base classifiers, and logistic regression is used as a meta-learner to optimize the combination of the prediction outputs of each base classifier.

[0051] According to the individual characteristics of children, the incidence probabilities of dwarfism, obesity and precocious puberty were calculated, and a comprehensive evaluation was conducted based on the disease correlation, which was expressed as: R total =∑ j=1 3 α j ·P(d j |X)+λ·∑ i≠j P(d i ,d j |X), where R total It is a comprehensive risk assessment result, which indicates the overall risk of children developing dwarfism, obesity and precocious puberty, taking into account the correlation between individual characteristics and diseases; α j is the weight factor of the jth disease, which is used to measure the relative importance of the disease in the comprehensive risk assessment; P(d j |X) is the disease d under the given child individual characteristics X j The probability of disease occurrence; λ is the adjustment factor, which indicates the strength of the correlation between diseases; P(d i ,d j |X) is the disease d under the given child individual characteristics X i and d j The joint probability of simultaneous occurrence.

[0052] Furthermore, step 5 includes the following steps.

[0053] In accordance with the clinical test report specifications, endocrine hormone quantitative data, growth and sexual development assessment results, and risk prediction analysis information are classified and formatted, and abnormal indicators are automatically marked and displayed in the report in a striking manner.

[0054] Based on the children's previous test data, combined with the standard growth curve, body mass index curve and developmental assessment standards, a personalized developmental trajectory chart is generated.

[0055] Comprehensively analyze various testing indicators, combine with the diagnostic standards for children's endocrine diseases, conduct multi-dimensional risk grading assessment, and graphically display the assessment results and explanations.

[0056] Personalized intervention recommendations are automatically generated based on risk assessment results, prioritized according to the child’s specific circumstances, and accompanied by scientific evidence and expected results.

[0057] The information is integrated into a hierarchical electronic report, in which the doctor version provides professional data and the parent version uses plain language and illustrations to ensure that the information is prominent and easy to understand.

[0058] Furthermore, based on the children's previous test data, combined with the standard growth curve, body mass index curve and sexual development assessment standards, a personalized development trajectory chart is generated, including the following steps.

[0059] Collect and organize all the children's test data, arrange all the test data in chronological order, and record the specific age at the time of the test.

[0060] Select the latest version of the standard reference curve, including the height standard curve, weight standard curve and BMI standard curve for the same gender and age group.

[0061] The children’s actual test data points are plotted onto the standard curve and connected to form the child’s individual growth and development trajectory, while also marking important developmental events and turning points.

[0062] The development of secondary sexual characteristics was assessed according to the Tanner staging criteria, and the progression of sexual development was correlated with height and weight development to fully understand the overall development status of the child.

[0063] Calculate development trends based on historical data, and calculate changing trends in height growth rate and weight growth rate.

[0064] Calculate the prediction interval of the child in the future and mark the upper and lower limits of the prediction interval on the trajectory graph. The prediction interval is expressed as: Y^(t f )±t α / 2 ·√[MSE·(1+1 / n+(t f -tˉ) 2 / ∑ r=1 n (t r -tˉ) 2 )], where Y^(t f） is predicted by the regression model at a certain time in the future f Estimated value of developmental index; t α / 2 is the critical value determined according to the selected confidence level; MSE represents the average of the squares of the differences between the model prediction value and the actual observed value; n is the number of sample data points, that is, the number of known historical test data; tˉ is the average value of the time data, which represents the average value of the age values ​​corresponding to all historical test data; t r Represents the age value corresponding to the rth data point.

[0065] Furthermore, step 6 includes the following steps.

[0066] The volume, concentration and uniformity parameters of blood samples are monitored in real time. When abnormalities are found, an alarm is automatically issued and treatment suggestions are provided to ensure that the sample quality meets the standards.

[0067] The key parameters of the detection process are tracked and recorded throughout the process, and the detection channel of the microfluidic chip is monitored through multi-point sensors to ensure the optimal state of the detection process.

[0068] Implement multi-level consistency assessment, compare the deviations between quality control samples and reference values ​​of children's endocrine indicators and analyze repeated measurement results to ensure that the test results are accurate and repeatable.

[0069] A unique identification code is assigned to each sample, and blockchain technology is used to record information of the entire process to ensure that the test results can be traced and verified.

[0070] Set quality control standards based on historical data of children's endocrine testing, automatically detect abnormalities and provide treatment plans to reduce human bias.

[0071] Regularly generate quality assessment reports, summarize quality control data, provide a basis for equipment maintenance, process optimization and laboratory certification, and continuously improve testing quality.

[0072] Compared with the prior art, this application has the following beneficial effects: This application integrates rapid sampling, automatic dilution, multi-channel microfluidic detection, real-time signal processing, high-sensitivity detection, deep learning analysis and quality control mechanisms to achieve full process automation from blood collection to test report generation, improve detection efficiency and accuracy, and provide personalized growth and development assessments and intervention recommendations, which facilitates early detection and intervention of endocrine disease risks in children. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 This is a schematic flow chart of the rapid detection method for screening dwarfism in children disclosed in an embodiment of the present application. DETAILED DESCRIPTION

[0074] In order to make the purpose, technical scheme and advantages of the implementation of the present invention clearer, the technical scheme in the embodiment of the present invention will be described in more detail below in conjunction with the drawings in the embodiment of the present invention. In the drawings, the same or similar reference numerals throughout represent the same or similar elements or elements with the same or similar functions. The described embodiments are part of the embodiments of the present invention, not all of the embodiments.

[0075] Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in the field without making any creative work shall fall within the scope of protection of the present invention.

[0076] The embodiments and directional terms described below with reference to the accompanying drawings are exemplary and intended to be used to explain the present invention, but should not be construed as limiting the present invention.

[0077] like Figure 1 As shown, the rapid detection method for screening dwarfism in children includes the following steps.

[0078] Step 1: Use a sampling device with integrated micro-blood collection and automatic dilution functions to accurately sample through the capillary effect and automatically dilute the sample to ensure accurate concentration.

[0079] Step 2: The collected samples are introduced into a multi-channel microfluidic detection chip, and parallel quantitative detection of endocrine hormones is achieved through a nano-biorecognition probe array.

[0080] Step 3: Based on high-sensitivity detectors and adaptive signal processing algorithms, weak biomolecule recognition signals are collected and amplified in real time, effectively improving the signal-to-noise ratio and ensuring the accuracy of the detection results.

[0081] Step 4: Based on the deep learning algorithm and combined with the dynamically updated reference database of children's growth and development and endocrine indicators, the test results are analyzed in multiple dimensions to achieve quantitative evaluation of growth and development, body mass index and sexual development indicators and prediction of endocrine disease risks.

[0082] Step 5: Automatically integrate the analysis results to generate a standardized test report, including quantitative test data, personalized growth curve, body mass index curve and sexual development assessment analysis, abnormal risk grading assessment and targeted intervention recommendations.

[0083] Step 6: Establish a quality control mechanism throughout the entire testing process, including real-time monitoring of sample quality, tracking of testing process parameters, result consistency assessment, and full traceability of testing data.

[0084] The core of step 1 is to accurately and efficiently obtain and process children's blood samples through integrated micro-blood collection and automatic dilution functions. Traditional blood sampling often relies on manual operation, which is not only easily affected by the technical level of the operator, but also has the risk of biased test results due to irregular sampling or inaccurate dilution. In order to solve this problem, the capillary effect is used to accurately control the drainage speed and sampling volume of the blood, so that highly consistent sample collection can be achieved with a small amount of blood. Through the microstructure of the capillary, blood can flow into the device naturally without external force, and is accurately collected through a well-designed channel, avoiding errors caused by human intervention. In addition, the automatic dilution function can dilute the blood to the appropriate concentration immediately after sampling to ensure the precise control of sample concentration during the subsequent multi-channel microfluidic detection chip analysis process. The advantage of this technology is that it provides a high-precision, automated and efficient sampling and dilution solution, which not only improves the convenience of operation, but also greatly reduces the incidence of human errors, thereby improving the accuracy and reliability of subsequent analysis.

[0085] Step 2 uses a combination of a multi-channel microfluidic detection chip and a nano-biorecognition probe array. This technical architecture realizes the parallel quantitative detection of endocrine hormones for children with dwarfism. Microfluidic technology is an advanced technology that combines microfluidic operations with integrated circuit design. It can integrate multiple detection processes into one chip, which not only greatly saves time and space, but also improves the efficiency and accuracy of detection. The multi-channel design enables the chip to handle the detection of multiple biomarkers at the same time, greatly improving the detection speed and avoiding the cumbersome one-by-one detection steps. The nano-biorecognition probe array can accurately identify and capture various endocrine hormones by binding to target molecules with high specificity. Compared with traditional biorecognition methods, nanoprobes have stronger binding ability and higher sensitivity due to their ultra-small size and high surface area, and can detect hormone molecules at extremely low concentrations, thus playing a key role in the screening of children with dwarfism. By detecting multiple hormones in parallel, quantitative data of multiple biomarkers can be obtained simultaneously in one test, providing a more comprehensive basis for subsequent comprehensive analysis.

[0086] The innovation of step 3 is mainly reflected in the combined application of high-sensitivity detectors and adaptive signal processing algorithms. Biomolecule recognition signals are usually very weak, especially in the detection of low-concentration endocrine hormones, where the signals are easily interfered by noise, resulting in inaccurate test results. Therefore, the use of high-sensitivity detectors is the key to technical implementation. High-sensitivity detectors can efficiently capture weak biorecognition signals through optimized sensor materials and precise detection circuit design, and transmit them to the processing unit in real time. In order to further improve the signal quality, adaptive signal processing algorithms play an important role here. The algorithm can automatically adjust the processing method according to the intensity and noise characteristics of the detection signal, and perform dynamic optimization, thereby realizing real-time amplification and denoising of the signal. The algorithm can adaptively identify and isolate background noise, thereby significantly improving the signal-to-noise ratio and ensuring the accuracy of the test results. This step ensures that even in a complex biological sample environment, weak endocrine hormone signals can be accurately captured and processed, providing a reliable technical guarantee for the final diagnostic results.

[0087] In step 4, multi-dimensional analysis based on deep learning algorithms achieves accurate assessment of children's growth and development and endocrine indicators and disease risk prediction. With the rapid development of big data technology, deep learning-based analysis methods can extract complex nonlinear features from a large amount of historical data and make more accurate trend predictions. Through the deep neural network model, a comprehensive analysis can be performed based on multiple parameters such as detected hormone levels, body mass index, sexual development indicators, etc., and the growth and development status of children can be automatically calculated. In combination with the dynamically updated children's growth and development database and the risk assessment of endocrine diseases, the children's development prediction results are given. Especially in the risk prediction of endocrine diseases, deep learning algorithms can identify subtle physiological changes, predict potential risks in advance, and help early intervention. Because the deep learning model has the ability to self-optimize, with the addition of more data, the prediction accuracy will continue to improve, thereby providing more accurate guidance for clinicians.

[0088] The core technical effect of step 5 is to realize the automatic integration of analysis results and generate standardized test reports. Through the aforementioned multi-dimensional analysis, data from various growth and development indicators can be automatically summarized to generate detailed reports including quantitative test data, personalized growth curves, body mass index curves, sexual development assessment analysis, etc. These reports can clearly show the health status of children and help doctors and parents better understand their children's growth and development. At the same time, it is also possible to conduct abnormal risk grading assessments based on the analysis results to provide a basis for whether children have problems such as dwarfism and endocrine diseases. Based on these assessments, targeted intervention recommendations will be automatically given to help parents take appropriate measures for early intervention and avoid further development of potential health problems. The automatically generated reports not only save doctors' time, but also improve screening efficiency, and are especially suitable for large-scale screening scenarios.

[0089] In step 6, the quality control mechanism throughout the entire testing process ensures the accuracy and reliability of each step. Through real-time monitoring of sample quality, tracking of test process parameters, and evaluation of result consistency, potential deviations can be discovered and corrected in a timely manner to ensure the stability and reliability of the test. The quality control mechanism not only relies on the self-calibration of the equipment, but also includes full data traceability to ensure the traceability and transparency of each step, which is crucial to ensuring the rigor and legality of the screening process. When abnormal situations occur, by being able to quickly track the source of the problem and propose reasonable improvement measures, high standards of quality assurance are provided for the screening of dwarfism in children.

[0090] In summary, by integrating advanced technologies such as micro-blood collection, automatic dilution, multi-channel microfluidic chips, high-sensitivity detection and deep learning analysis, this method can achieve efficient and accurate screening for dwarfism in children. The close integration of each step improves the accuracy, timeliness and operability of screening, while also ensuring high-quality control during the detection process. The integration of these steps not only provides strong support for the early identification of dwarfism in children, but also lays a solid foundation for personalized intervention and early diagnosis of endocrine diseases, and has broad clinical application prospects.

[0091] Furthermore, step 1 includes the following steps.

[0092] Use 75% medical alcohol to fully disinfect the sampling head and dilution chamber of the sampling device, and check the metering accuracy of the device with a standard solution to ensure that the device is in the best working condition. The core innovation of this step is reflected in the comprehensive disinfection measures and precise calibration. The accuracy and hygiene of the sampling device are the key to ensuring the reliability of the test results. Using 75% medical alcohol to disinfect the sampling head and dilution chamber of the sampling device can not only effectively kill potential pathogens, but also avoid cross contamination. Especially when handling children's samples, it is very important to ensure the cleanliness of the sampling tools, so as to minimize the impact of external contamination sources on the test results. In this step, alcohol disinfection can ensure that the sampling head and the inside of the dilution chamber are not interfered by environmental factors, thereby ensuring the accuracy and hygiene requirements in the subsequent blood sampling and dilution process. At the same time, the use of standard solutions for metering accuracy checks is a necessary means to verify whether the equipment is in the best working condition. In this way, it is possible to detect in real time whether the equipment has deviations or inaccuracies, thereby effectively avoiding measurement errors caused by equipment problems and ensuring that subsequent sampling and dilution operations are accurate and correct.

[0093] Use alcohol cotton pads to disinfect the fingertips of children. After they are dried naturally, use a special blood collection needle for children to puncture, and the depth is controlled at 1.8-2.2 mm. The key to this step is the strict operating standards and the use of special blood collection tools for children. Alcohol cotton pads are used to disinfect the fingertips of children. This operation not only ensures the cleanliness of the sampling site, but also reduces the risk of infection caused by uncleanness. After disinfection, the fingertips are dried naturally to ensure that alcohol does not interfere with the subsequent sampling process, especially when blood is collected to avoid alcohol residue affecting the quality of the blood sample. Using a special blood collection needle for children for puncture can effectively avoid the discomfort or injury caused by too deep puncture. Unlike adult blood collection tools, the design of children's blood collection needles is more in line with the physiological characteristics of children. The needle design is smaller and the puncture depth is strictly controlled at 1.8-2.2 mm to avoid tissue damage caused by too deep puncture. Through this meticulous depth control, the discomfort and pain of children can be minimized, while ensuring sufficient blood sample volume to provide high-quality raw materials for subsequent analysis. This design specifically for children reflects the humanization of the equipment and full protection of children's health.

[0094] Relying on the capillary effect, the preset volume of blood sample is automatically sucked in. The capillary effect is a phenomenon in which liquid flows along a small pipe under the action of surface tension. The fine capillary in the sampling device is designed as a precise micro-channel. Through the attraction of the capillary, the accurate volume of blood sample can be automatically sucked in without additional manual operation or mechanical pumping. The advantage of this technology is that it can maintain high accuracy and consistency during the process of micro-blood sampling, avoiding errors in manual operation. In addition, the application of the capillary effect means that the design of the device can be very compact and simple to operate, which not only reduces costs but also improves the convenience of operation. Through precise blood suction, the amount of blood sample collected can be guaranteed to be moderate, providing sufficient and non-wasted samples for subsequent analysis, while avoiding the situation of over-sampling or under-sampling that may occur in traditional methods. The application of the capillary effect reflects the innovation of this technology in microscopic physical principles and improves the efficiency and accuracy of sampling.

[0095] After the blood reaches the preset amount, the buffer solution pre-installed in the dilution chamber is mixed with the blood sample in a precise proportion to form a sample solution with uniform concentration. The key to this step is to accurately control the mixing ratio of blood and buffer solution to ensure the uniform concentration of the sample solution. The quality of the blood sample and the ratio of the diluent are crucial to the test results. By pre-installing the buffer solution in the dilution chamber, the proportion of mixing with the blood sample can be accurately controlled. This precise proportion control not only ensures the stability of the solution, but also avoids the impact of too high or too low concentration on the test results. The use of buffer solution can adjust the pH value of the sample, maintain the stability of the blood sample, and prevent the blood components from changing or precipitating. By achieving precise proportion control during the mixing process, the consistency and repeatability of each operation are ensured, thereby improving the reliability and accuracy of the test. The implementation of this technology further improves the quality of the sample, ensures that the biomolecules in the sample can be detected under optimal conditions, and effectively avoids detection errors caused by uneven dilution or concentration gradient phenomena.

[0096] After the dilution process is completed, the mixed liquid sample automatically flows into the collection chamber and is fully mixed to avoid the concentration gradient phenomenon. In order to avoid the concentration gradient phenomenon caused by uneven mixing, the design of the collection chamber ensures that the liquid can be evenly distributed and fully mixed in the collection chamber through an optimized fluid dynamic structure. The advantage of this technical design is that it can automatically perform the final mixing of the sample without manual intervention, thereby avoiding errors that may be caused during the operation. In addition, by effectively preventing the concentration gradient phenomenon, the uniformity of the sample is ensured, so that subsequent tests can obtain accurate and consistent data. This automated mixing process improves work efficiency, avoids the influence of human factors on sample quality, and ensures the stability of the sample throughout the detection process, providing an ideal basis for subsequent precise analysis.

[0097] In summary, through the careful design and precise control of the above steps, each link is closely coordinated to ensure the efficiency, accuracy and safety of blood sampling, dilution and mixing. The use of capillary effect, precise proportional dilution and automatic mixing technology makes the entire sampling process efficient and stable, avoiding errors and inconsistencies in manual operation. In particular, the blood collection equipment specially designed for children and the tailor-made technical design reflect the greatest concern and protection for children's health. Through the integration of these technologies, not only the convenience of operation and detection accuracy are improved, but also the potential biosafety risks are reduced.

[0098] Furthermore, step 2 includes the following steps.

[0099] The multi-channel microfluidic detection chip is preheated to 37°C and pretreated with a special buffer to ensure that the hydrophilicity of the inner wall of each microchannel reaches the optimal state to prevent the generation of bubbles and ensure the uniform flow of samples in the channel. Preheating to 37°C is to simulate the body temperature environment so that the biological samples in the chip and the probe reaction can be carried out under physiological conditions. The simulation of body temperature can not only ensure the authenticity and reliability of the reaction, but also optimize the reaction speed and efficiency of biomolecules. In addition, pretreatment of the chip with a special buffer can effectively improve the hydrophilicity of the inner wall of the microchannel and prevent unnecessary bubble formation. This design is very important because bubbles will affect the uniform flow of the liquid, resulting in uneven sample distribution, which in turn affects the accuracy of the detection. Microfluidic technology relies on the fine flow of liquid in tiny channels, so the hydrophilicity of the channel plays a vital role in the stable flow of the liquid. Through this treatment, the uniformity and controllability of the sample flow in the chip are ensured, which helps to improve the sensitivity and stability of the detection. At the same time, the optimized hydrophilic surface reduces the friction between the liquid and the material, improves the flow efficiency of the liquid, and ensures the smooth progress of subsequent reactions. Through the preheating and pretreatment process, it can be ensured that the chip's performance reaches the best state during formal testing, ensuring that it runs efficiently and stably during the working process.

[0100] The mixed solution is accurately introduced into the injection port of the chip using a microinjector, and the mixed solution is automatically distributed into multiple parallel detection channels under the action of microfluidics. The technical effect of this step reflects the advantages of microfluidics in sample guidance and distribution. The precise introduction of the mixed solution by the microinjector can ensure that the volume, concentration and uniformity of the sample remain consistent each time the sample is injected. This is crucial for subsequent detection, because the precise processing of trace samples directly affects the reliability of the final result. The principle of microfluidics realizes the automatic distribution of samples by precisely controlling the flow of liquids in tiny channels. The multi-channel microfluidic detection chip can effectively distribute samples evenly into multiple parallel detection channels through microfluidics, ensuring that there is enough sample volume in each channel for detection. Since the flow environment and transmission rate of each channel are accurately calculated during the design, this distribution method avoids the uneven distribution of samples in some channels due to differences in flow rate or uneven resistance. This automated fluid distribution method not only improves operational efficiency, but also ensures the uniformity and consistency of samples in the detection channel, which helps to improve the accuracy and repeatability of the test results.

[0101] Start the microfluidic control module on the chip to accurately adjust the flow rate and residence time of the sample in each detection channel to ensure that the sample fully contacts and reacts with the nanoprobes in each channel. By starting the microfluidic control module, the flow rate and residence time of the sample can be accurately adjusted according to the characteristics of each channel to ensure that the sample is evenly distributed and fully reacts in the chip. Flow rate and residence time are key parameters in microfluidic experiments. Too fast a flow rate may lead to insufficient contact between the sample and the probe and incomplete reaction, while too slow a flow rate may lead to time waste and affect the efficiency of the entire detection. Through microfluidic technology, the chip can accurately control the flow of liquid in each detection channel to ensure that the sample does not waste time in the process of contacting the probe and can ensure the best reaction conditions. The optimized flow rate and residence time settings enable each sample molecule to fully contact and react with the corresponding nanoprobe, thereby maximizing the detection sensitivity and accuracy. In addition, this technology can also be adaptively adjusted according to the type of sample and the needs of the reaction to ensure that different types of samples can obtain the best detection conditions.

[0102] During the reaction, various endocrine hormones bind specifically to the corresponding nanoprobes, causing changes in the physical and chemical properties of the probes and capturing and recording them in real time. In this step, the specific binding of the nanoprobes to the endocrine hormones triggers changes in the physical and chemical properties of the probes, thereby generating detectable signals. High-sensitivity, real-time monitoring in microfluidic chips is achieved through specific molecular recognition reactions. The nanoprobes are designed to be highly selective and can specifically bind to target molecules (such as hormones) without interfering with other substances, thereby ensuring the specificity and sensitivity of the detection. After the hormone molecules bind to the nanoprobes, they usually cause changes in the physical and chemical properties of the probe surface, such as charge, fluorescence intensity or optical absorption, which can be captured and recorded in real time. This technology enables the chip to not only have efficient detection capabilities, but also monitor the reaction process in real time, avoiding errors caused by time delays or sampling problems in traditional methods. Real-time capture and recording of signals makes the detection process more efficient, and accurate detection data can be obtained in the shortest time, thereby speeding up the entire detection process and improving the reliability and accuracy of the results.

[0103] After the reaction is completed, the detection channel is gently rinsed with a special cleaning buffer to remove unbound interfering substances. This step ensures the purity and accuracy of the test results. After the reaction is completed, the cleaning process in the chip is very important, because if the unbound interfering substances are not removed in time, they will interfere with the subsequent signal reading. Using a special cleaning buffer, these unbound substances can be effectively removed without damaging the nanoprobes in the chip or the bound target molecules. The cleaning process needs to be very delicate to ensure that effective binding substances are not taken away or the reactivity of the probe surface is destroyed. Through delicate cleaning, the signal strength and accuracy of each detection channel can be guaranteed, ensuring that the final read signal only reflects the concentration of the target hormone, excluding the influence of nonspecific binding or background signals.

[0104] Through the signal acquisition module integrated in the chip, the signal intensity of each detection channel is read synchronously, and converted into the specific concentration value of the corresponding hormone level according to the pre-established standard curve. The signal acquisition module integrated in the chip can read the signal intensity in each detection channel in real time and synchronously, ensuring that the signal of each channel can be accurately captured and processed when multiple channels are running simultaneously. By comparing with the pre-established standard curve, the collected signal intensity can be converted into the corresponding hormone concentration value. This process ensures the accuracy and repeatability of the data through the application of precise algorithms and standard curves. The standard curve is established through a large amount of data, which can convert the measured physical signal into specific biological meaning, that is, the concentration of each hormone. In this way, the chip can not only provide quantitative hormone level data, but also compare with the preset clinical reference standard, thereby providing a more accurate basis for diagnosis. This process ensures the high reliability of the test results and can meet the needs of clinical screening and analysis.

[0105] In summary, through the precise design and operation of the microfluidic chip, the above steps have achieved full automation of the entire process from sample input, precise distribution, reaction control, signal capture to result analysis. This highly integrated and automated design significantly improves the efficiency and accuracy of detection and reduces the errors that may be caused by manual operation. At the same time, the chip ensures the optimization of each reaction step through meticulous control, providing reliable and accurate data for the final hormone concentration detection.

[0106] Furthermore, the signal intensity of each detection channel is synchronously read through the signal acquisition module integrated in the chip, and converted into a specific concentration value of the corresponding hormone level according to a pre-established standard curve, including the following steps.

[0107] The signal acquisition of all detection channels is synchronously controlled by a multiplexer, and the analog-to-digital converter is used to convert the analog signal into a digital signal, and the interference of random noise is eliminated. The multiplexer (MUX) enables multiple signal channels to be read synchronously through a shared acquisition system, which not only reduces the complexity of the hardware, but also effectively improves the operating efficiency of the system. Through multiplexing, the signals in multiple detection channels can be monitored in real time without adding a lot of hardware costs, and their synchronization can be guaranteed. During the processing, the analog signal is converted into a digital signal by an analog-to-digital converter (ADC) for further analysis and processing. The conversion of analog signals is usually accompanied by a certain amount of noise, and if the random noise is not eliminated in time, it will affect the authenticity and accuracy of the signal. In order to effectively suppress noise, the analog-to-digital converter usually adopts noise filtering technology, such as low-pass filter and average filter algorithm, which can reduce the noise interference generated by the external environment or the hardware itself, so as to obtain a clearer and more reliable signal. This link ensures that the signals read from each channel of the chip are accurate and interference-free, which is helpful for the subsequent signal processing and the accuracy of concentration calculation.

[0108] The relative signal intensity value is calculated by comparing the signal amplitude of each channel with the signal difference of the reference channel. Different detection channels will produce differences in signal intensity due to different reactions of the probes, the types of samples or slight errors in the experimental process. Therefore, comparing the signal amplitude of each channel with the signal difference of a reference channel (generally a standard sample with a known hormone concentration) can effectively correct these errors and calculate the signal intensity of the sample in each detection channel relative to the reference channel. This relative intensity calculation can eliminate systematic errors caused by fluctuations in experimental conditions or equipment differences. By comparing with the reference channel, the hormone concentration of the sample can be determined more accurately. This relative value calculation not only makes the signal quantification more accurate, but also provides strong support for subsequent concentration calculations. Through standardized comparison, individual differences that may occur in a single channel during signal acquisition can be eliminated, thereby ensuring the comparability of results between different samples and further improving the stability and reliability of detection.

[0109] According to the standard curve equation established by samples of known concentration, the relative signal intensity value is substituted into the calculation to obtain the actual concentration of the target hormone in the sample to be tested. The establishment of the standard curve is usually achieved by measuring the signal intensity in samples of different known concentrations to obtain the relationship between concentration and signal intensity. This relationship generally presents linear or certain specific nonlinear characteristics, which can reflect the quantitative relationship between hormone concentration and measured signal under specific experimental conditions. Substituting the relative signal intensity value into this standard curve equation, the hormone concentration in the sample to be tested can be calculated. The technical advantage of this method is that it can convert the original signal (usually amplitude) obtained during the experiment into specific biological data, that is, the concentration value of the hormone. The standard curve is the basis for quantitative analysis and can effectively compensate for various changing factors in the experimental process, thereby improving the accuracy of the data. Through this calculation step, the complex relationship between signal intensity and target hormone concentration can be simplified into a mathematical model, making the sample analysis process automated and efficient.

[0110] The actual concentration value calculated is validated to ensure the accuracy and reliability of the measurement results. The hormone concentration value obtained by the standard curve is only an extrapolated result based on the experimental data, and still needs to be validated to ensure its accuracy. Validation generally includes multiple aspects of inspection, such as comparative experiments with samples of known concentrations, comparison of sample results under different batches or experimental conditions, etc. By comparing with these verification methods, it is possible to check whether there are systematic deviations or errors in the measurement results to ensure that the data obtained is credible. In addition, through validation, outliers or extreme values ​​can be identified and excluded, further improving the reliability of the test results. For example, the signal of a certain channel may be abnormal due to factors such as chip aging and sample contamination. At this time, through validation, these errors can be discovered and corrected in time to avoid adverse effects on the final results. Validation can also evaluate the repeatability and consistency of the detection process based on the stability of multiple measurement results to ensure that highly similar results can be obtained each time.

[0111] In summary, by integrating the signal acquisition module with the standard curve analysis, the whole process from signal acquisition, signal processing, concentration calculation to result verification can be completed efficiently. The design of each link plays a vital role in improving the accuracy, sensitivity and reliability of the detection. The use of multiplexers and analog-to-digital converters ensures the synchronous acquisition and noise suppression of multi-channel signals, avoiding signal distortion and errors. By comparing with the reference channel, the difference in signal intensity can be effectively corrected and the accuracy of the calculation can be improved. The establishment of the standard curve converts the original signal into a specific concentration value, thereby realizing the quantitative analysis of hormone concentration. Finally, through the validity verification step, the test results are verified at multiple levels, further enhancing their accuracy and reliability. The automation and precision of the entire process not only reduces the interference of human operation, but also makes the test results more stable and reliable, providing an efficient and accurate technical solution for the screening of dwarfism in children.

[0112] Furthermore, step 3 includes the following steps.

[0113] Under completely light-shielded conditions, record the baseline noise level of the detector and collect the background signal of the blank sample for subsequent signal processing and correction. In this link, it is crucial to ensure the baseline noise level of the detector and the collection of background signals. The baseline noise level refers to the signal detected by the detector in the absence of sample or hormone signals. This signal usually comes from electronic components, environmental interference or system noise, so it must be recorded under completely light-shielded conditions to minimize the impact of light sources and other external environments on the detector. By collecting the background signal of the blank sample, the original noise level can be accurately understood, which plays a key role in subsequent signal processing and correction. Collecting background signals can not only provide noise baseline data, but also help determine the subsequent signal processing strategy, such as the degree and method of noise filtering. Accurate baseline noise recording is the basis for signal correction, ensuring that the noise removed in the signal processing stage is truly random noise, and will not accidentally damage the valid signal. In addition, understanding the characteristics of the background signal can also help further optimize the sensitivity and accuracy of the detection process and ensure the reliability of subsequent data analysis.

[0114] The high-sensitivity photodetector is turned on, and a high-performance photomultiplier tube is used to collect the weak hormone-antibody binding signals of each detection channel of the chip in real time. The photomultiplier tube (PMT) is a very efficient light detector that can significantly increase the intensity of weak light signals through multiple electron multiplication processes. Especially in the weak signals generated by the hormone-antibody binding reaction, the use of photomultiplier tubes can ensure that even the detection of low-concentration hormones can be accurately measured. The use of high-sensitivity photodetectors can improve the sensitivity of the signal without significantly increasing the background noise, so that it can respond to extremely small signals. This is crucial for the hormone-antibody reaction signals in each detection channel in the microfluidic chip, especially when detecting low-concentration hormones in the human body, which can effectively avoid signal loss. In addition, real-time acquisition can quickly track the progress of sample reactions, monitor changes in signal intensity in a timely manner, and ensure the timeliness and accuracy of data, providing a reliable input signal for subsequent analysis.

[0115] The method of combining wavelet transform and adaptive Wiener filtering is adopted to automatically adjust the filtering parameters according to the signal characteristics, effectively remove high-frequency random noise and low-frequency drift, and retain the effective information of the target signal to the greatest extent. The method of combining wavelet transform and adaptive Wiener filtering is an advanced signal processing technology with powerful noise removal and signal enhancement functions. Signals are usually interfered by various noise sources, and the noise includes not only high-frequency random noise but also low-frequency drift. As a time-frequency analysis method, wavelet transform can effectively separate the different frequency components of the signal and extract the changing characteristics of the signal in time and frequency. Through wavelet transform, the noise and effective signal in different frequency components can be identified, and the noise in different frequency ranges can be filtered out. Adaptive Wiener filtering further enhances this effect. It can automatically adjust the filtering parameters according to the characteristics of the signal, making the removal of high-frequency noise and low-frequency drift more accurate during the signal processing process, while retaining the effective information of the target signal to the greatest extent. Through this signal processing technology, the signal-to-noise ratio can be greatly improved to ensure the high accuracy and reliability of the detection results. The advantage of this step is its automated parameter adjustment capability, which makes the processing more intelligent, adapts to signal changes under different experimental conditions, and improves the robustness of the overall detection system.

[0116] The amplification factor is automatically adjusted according to the change of signal strength to ensure that weak signals are amplified while avoiding saturation distortion of strong signals, so that the signal is always kept within the optimal detection range, which can be expressed as: G = G min +(G max -G min )·[A threshold / (A in +A threshold )], where G is the gain factor, i.e. the magnification; Gmin is the minimum gain, i.e. the lower limit of the gain coefficient; G max is the maximum gain, that is, the upper limit of the gain coefficient; A in is the input signal strength; A threshold It is a preset signal strength threshold, which is used to determine whether the change in signal strength requires adjustment of the gain. In this step, the signal amplification process is an important link to ensure that weak hormone signals are accurately collected. In order to avoid signal distortion, especially the saturation phenomenon that may occur in the case of strong signals, it is necessary to intelligently adjust the amplification factor according to the change in signal intensity. In practical applications, hormone detection signals are usually distributed in different intensity ranges. Some signals may be very weak and require a higher amplification factor to be effectively detected, while other signals may be relatively strong. If too high a gain is used, the signal will be oversaturated and distorted. By dynamically adjusting the amplification factor (gain coefficient), it is ensured that both weak and strong signals can be processed under the optimal amplification state.

[0117] Taking into account the signal strength, time characteristics and spatial distribution characteristics, the effective signal is accurately extracted through pattern recognition methods to further improve the signal-to-noise ratio, which is expressed as: SNR optimized =[I signal (T) F signal (ω)·S~ effective (k x ,k y ,k z )·M(S effective ,N)] / P noise , where SNR optimized is the optimized signal-to-noise ratio, that is, the signal-to-noise ratio that is finally improved after pattern recognition and multi-dimensional feature analysis; P noise is the noise power, which means the noise signal in the entire time window The average power within signal (T) is the signal strength, that is, the signal in a specific time window The average intensity within signal (ω) is the spectrum of the signal, that is, the frequency distribution of the signal obtained by Fourier transform, reflecting the distribution of the signal in different frequency components; S~ effective (k x ,k y ,k z） is the spatial spectrum of the effective signal, that is, the frequency component of the effective signal in the spatial domain; M(S effective ,N) is the pattern recognition function, which represents the relationship between the effective signal and the noise signal processed by the pattern recognition method; N is the noise signal; ω is the representation of the signal in the frequency domain; k x ,k y ,k zRepresent the frequencies in the X, Y, and Z directions, respectively. In this step, the effective signal is accurately extracted by comprehensively analyzing the signal's strength, time characteristics, and spatial distribution characteristics. The signal's time characteristics (such as the time when the peak occurs, the duration of the signal) and spatial distribution characteristics (such as the distribution of the signal at different locations) can provide important information about the nature of the signal. Through pattern recognition methods, these features can be analyzed to identify effective signals that are different from noise signals. Pattern recognition can not only help identify and remove noise, but also optimize the signal extraction process through the spatial and temporal characteristics of the effective signal, thereby further improving the signal-to-noise ratio. Specifically, pattern recognition technology analyzes the signal's frequency spectrum F signal (ω), spatial spectrum S~ effective (k x ,k y ,k z） And the relationship between signal and noise M(S effective ,N), and finally improve the quality of effective signals and reduce noise interference through optimization algorithms. The key advantage of this method is that it can accurately identify and extract truly valuable information in a complex signal environment, ensuring the reliability and accuracy of the final detection results.

[0118] In summary, the technical implementation of the entire step 3 improves the acquisition accuracy and reliability of weak hormone signals and ensures the accuracy of each link from signal acquisition to signal processing. The baseline noise record provides a solid foundation for subsequent signal correction, and the combination of high-sensitivity photodetectors and photomultiplier tubes ensures the effective capture of weak signals. Through the combined application of wavelet transform and adaptive Wiener filtering, the system noise is effectively removed and the effective components of the target signal are retained. The automatic gain adjustment function ensures the optimal amplification effect of the signal at different intensities, while the pattern recognition technology further improves the signal-to-noise ratio and ensures the quality and accuracy of the signal. The combination of these technologies enables the overall detection system to always maintain an efficient, stable and accurate working state when facing complex biological samples, thus providing strong technical support for the screening of dwarfism in children.

[0119] Furthermore, step 4 includes the following steps.

[0120] Establish digital archives for children, collect growth and development indicators, body mass index, development stage and bone age assessment information, and perform standardization to ensure that the data is suitable for deep learning model analysis. In this step, establishing digital archives for children is the core link. By collecting children's growth and development indicators, body mass index, development stage and bone age assessment information, we can fully grasp the growth and development of each child. These data provide rich input features for subsequent deep learning models and can provide a basis for accurately analyzing children's health status and development. Growth and development indicators usually include basic information such as height and weight. Body mass index (BMI) reflects the child's body fat level. Sexual development stage assessment helps to determine the child's sexual development process. Bone age assessment is related to the level of bone development and can indirectly reflect the child's growth potential. After data collection, standardization is a very critical step. Since there are large individual differences in children's growth and development, and different physical examination hospitals or equipment may use different measurement standards and methods, directly inputting these data into the deep learning model will lead to data deviation. Through standardization, data from different sources can be adjusted to a unified scale range, eliminating the impact of unit differences and outliers, and ensuring data consistency and comparability. Standardized data can provide clear and unbiased data input for deep learning models, which helps improve the accuracy and robustness of analysis results. In addition, standardized processing steps can enhance data adaptability, so that data of children of different ages and genders can be effectively processed, and developmental conditions in different contexts can be adapted, so as to make predictions and analyses more accurately.

[0121] The standardized data is input into the deep learning model, and the key features are extracted through the multi-layer convolution structure and attention mechanism, focusing on analyzing the dynamic changes of endocrine hormone levels and growth and development indicators. In this step, the standardized data will be input into the deep learning model for analysis, and the deep learning model can effectively handle complex data relationships and nonlinear features. Specifically, the model extracts key features from the data through a multi-layer convolutional neural network (CNN) structure. The convolution layer can locally perceive the input data and automatically capture local patterns and high-dimensional features in the data through the sliding of the convolution kernel. In the analysis of children's growth and development, the multi-layer structure of the convolutional neural network can effectively learn the deep relationship between indicators such as height, weight, body mass index (BMI), and sexual development, and identify potential factors that affect children's growth and development. At the same time, the introduction of the attention mechanism further improves the feature extraction ability of the model. The attention mechanism can automatically adjust the weight of each feature according to the importance of the data, give higher attention to important features, and thus avoid interference with irrelevant features. In the dynamic analysis of growth and development, endocrine hormone levels are one of the key factors, which are closely related to children's growth rate, bone age, etc. Through the convolutional structure and attention mechanism in the deep learning model, it is possible to automatically identify features closely related to endocrine hormone changes from a large amount of input data, ensuring that not only static growth data is captured during the analysis process, but also the dynamic changes between endocrine hormone levels and children's growth and development can be tracked. In this way, the model can more accurately predict the development trend of children and identify any possible growth abnormalities or risks of developmental delay.

[0122] Using the dynamically updated reference database of children's growth and endocrine diseases, the transfer learning technology is used to compare children's data with similar cases in multiple dimensions to generate quantitative evaluation indicators. The application of transfer learning technology in this step is key, which enables the model to analyze and reason about new children's data based on existing knowledge and experience. Through the dynamically updated reference database of children's growth and endocrine diseases, it is possible to quickly generate comparative analysis with the help of health data and cases from a large number of children. The advantage of this method is that the model not only relies on the historical data of a single child, but also improves the accuracy of prediction by comparing similar cases. Through transfer learning, the model will learn more patterns and rules from known similar cases, and then generate personalized quantitative evaluation indicators for each child. This process is particularly important for multi-dimensional growth and development assessment. Children's growth and development is not just a static process, it also involves dynamic changes in endocrine hormone levels, body mass index, bone age and other aspects. Therefore, through transfer learning, children's historical data can be comprehensively compared in multiple dimensions, and a comprehensive evaluation indicator can be generated by combining these comparison results. These evaluation indicators can reflect the health status of children at different developmental stages, help doctors and parents understand the growth and development trends of children in real time, and provide a basis for the next step of intervention measures.

[0123] Ensemble learning methods are used to predict the risk of dwarfism, obesity and precocious puberty, and to calculate the risk probability of children developing related diseases in the future. In this step, ensemble learning methods are used to predict the risk of children developing dwarfism, obesity and precocious puberty in the future. Ensemble learning is a technology that combines the prediction results of multiple learning models to improve the overall prediction accuracy. Common ensemble learning methods include random forest, boosting and bagging. These methods can effectively reduce the bias and variance of the model when dealing with complex problems by combining the prediction results of multiple basic learners, and improve the robustness and accuracy of the prediction. For children's health screening, dwarfism, obesity and precocious puberty are common endocrine-related diseases. Their occurrence is usually closely related to multiple factors, such as genes, endocrine hormone levels, body mass index, sexual development process, etc. Through ensemble learning methods, the model can comprehensively analyze these complex factors and accurately predict the disease risk of each child. In this process, the model first learns the prediction rules of each disease through training data, and then generates a more accurate risk probability by combining the prediction results of different models. The advantage of this approach is that it can handle complex nonlinear relationships and high-dimensional data, ensuring that the comprehensive assessment of the risks of multiple diseases is more accurate and comprehensive.

[0124] In summary, through the above steps, the entire process forms a complete child health screening system, which comprehensively analyzes multi-dimensional data such as children's growth and development, body mass index, sexual development and endocrine hormones through deep learning and ensemble learning methods. On the basis of establishing digital archives for children and standardized data processing, the deep learning model effectively extracts key features through convolutional structure and attention mechanism to capture dynamic changes in growth and development. Transfer learning technology generates personalized evaluation indicators through comparative analysis with similar cases, providing strong data support for predicting the risks of dwarfism, obesity and precocious puberty. The ensemble learning method improves the accuracy and robustness of disease prediction by integrating the results of multiple prediction models, and ultimately provides accurate risk assessment for children's health management.

[0125] Furthermore, the standardized data are input into the deep learning model, and the key features are extracted through the multi-layer convolution structure and attention mechanism, focusing on analyzing the dynamic changes of endocrine hormone levels and growth and development indicators, including the following steps.

[0126] A multi-layer convolutional neural network is designed as the backbone network, which includes multiple layers of convolutional layers, pooling layers and fully connected layers. The convolutional layers are responsible for extracting local features, the pooling layers are responsible for feature dimensionality reduction and main information extraction, and the fully connected layers are responsible for integrating high-dimensional features.

[0127] Attention layers are embedded in key positions of multi-layer convolutional neural networks to automatically learn the importance weights between different features, so as to better focus on the key correspondence between endocrine hormones and various growth and development indicators.

[0128] Recurrent neural networks are used to process time series data to capture the patterns of changes in various indicators over time, with special attention paid to the lagged effects of changes in hormone levels on growth and development.

[0129] The extracted static features and temporal features are multimodally fused through a multi-layer perceptron to deeply analyze the nonlinear relationship between endocrine hormones and growth and development indicators, and a prediction model is established.

[0130] The performance of the prediction model was evaluated through cross-validation, the prediction results were displayed using visualization technology, and the impact of different hormone indicators on growth and development was analyzed.

[0131] In summary, through the comprehensive application of multi-layer convolutional neural networks, attention mechanisms, recurrent neural networks, and multimodal fusion technologies, deep learning models can effectively capture the dynamic characteristics of children's growth and development and the complex influence of endocrine hormones. Cross-validation and visualization techniques further ensure the accuracy and interpretability of the model and provide a reliable prediction tool. The combination of these technologies not only helps to deeply understand the relationship between endocrine hormones and children's growth and development, but also provides a scientific basis for future personalized health management and early intervention.

[0132] Furthermore, an integrated learning method is used to predict the risks of dwarfism, obesity and precocious puberty, and calculate the risk probability of children developing related diseases in the future, including the following steps.

[0133] Multiple types of machine learning algorithms were selected as base classifiers, and each base classifier was modeled for dwarfism, obesity, and precocious puberty respectively.

[0134] The cross-validation method is used to generate the prediction results of the base classifiers, and logistic regression is used as a meta-learner to optimize the combination of the prediction outputs of each base classifier.

[0135] According to the individual characteristics of children, the incidence probabilities of dwarfism, obesity and precocious puberty were calculated, and a comprehensive evaluation was conducted based on the disease correlation, which was expressed as: R total =∑ j=1 3 α j ·P(d j |X)+λ·∑ i≠j P(d i ,d j |X), where R total It is a comprehensive risk assessment result, which indicates the overall risk of children developing dwarfism, obesity and precocious puberty, taking into account the correlation between individual characteristics and diseases; α j is the weight factor of the jth disease, which is used to measure the relative importance of the disease in the comprehensive risk assessment; P(d j |X) is the disease d under the given child individual characteristics X j The probability of disease occurrence; λ is the adjustment factor, which indicates the strength of the correlation between diseases; P(d i ,d j |X) is the disease d under the given child individual characteristics X i and d j The joint probability of simultaneous occurrence.

[0136] In summary, the ensemble learning method can effectively improve the risk prediction accuracy of dwarfism, obesity and precocious puberty in children. Selecting multiple base classifiers and optimizing their prediction results through cross-validation can capture the diversity characteristics of different diseases; using logistic regression as a meta-learner to further optimize the prediction results improves the stability and accuracy of the model. In the comprehensive risk assessment process, by combining the individual characteristics of the disease, the probability of occurrence and the correlation between diseases, a personalized and accurate health risk prediction can be provided for each child. This method not only improves the accuracy of disease prediction, but also provides strong support for early intervention and health management.

[0137] Furthermore, step 5 includes the following steps.

[0138] According to the clinical laboratory report specifications, the quantitative data of endocrine hormones, the results of growth and development and sexual development assessment, and the risk prediction analysis information are classified and formatted, and abnormal indicators are automatically marked and displayed in the report in a striking manner. In the analysis of children's growth and development and endocrine hormone levels, the accurate presentation and standardization of clinical data are crucial for doctors and parents to understand and interpret. By classifying and formatting various data according to the clinical laboratory report specifications, the consistency and reliability of information transmission can be ensured, and misunderstandings or misdiagnoses caused by chaotic or non-standard data can be avoided. The standardized data format includes the timestamp, unit label, reference interval, and related annotations of the data, so that each data has a clear context. Automatically marking abnormal indicators is an important part of the data formatting process. Through the preset standard range and clinical specifications, abnormal values ​​can be automatically detected and marked prominently, helping doctors to pay attention to possible health risks in a short time. This not only improves the efficiency of data processing, but also reduces the workload of doctors. Especially when facing a large amount of children's data, automatic marking of abnormalities can significantly reduce the probability of manual missed diagnosis or misdiagnosis. The abnormal indicators are prominently displayed in the report, not only limited to color identification or symbol prompts, but also can highlight the degree of deviation of the data through charts. For example, using color gradients or bold fonts to highlight outliers can help doctors see the problem at a glance and focus on these abnormal indicators in subsequent analysis.

[0139] According to the children's previous test data, combined with the standard growth curve, body mass index curve and development assessment standards, a personalized development trajectory map is generated. The personalized development trajectory map is a dynamic visualization of the children's growth and development process, which can clearly show the changing trends of various indicators such as children's height, weight, bone age, and endocrine hormone levels over time. By combining the children's previous test data with the standard growth curve (such as the WHO growth curve, CDC standard, etc.), a personalized development trajectory map can be generated for each child, and compared with the standard growth curve to intuitively show the children's growth and development. The technical key to generating a personalized development trajectory map lies in how to process multi-source data and perform effective curve fitting. First, the children's height, weight, bone age and other data are standardized and compared with the standard growth curve to generate a personalized curve compared with the standard growth trajectory. This curve shows the actual development of children at different growth stages, helping doctors identify potential developmental abnormalities. For example, if a child's height development deviates greatly from the standard curve, it may indicate problems such as malnutrition, abnormal hormone levels or growth retardation. In addition, the body mass index (BMI) curve can be combined to further assess the risk of obesity or thinness in children, especially in the early screening of childhood obesity, malnutrition and other related endocrine diseases. The BMI curve, as one of the important indicators, can provide doctors with a more comprehensive risk assessment. The sexual development assessment standards help identify problems such as precocious puberty or delayed sexual development, and form a targeted developmental assessment. This developmental trajectory chart not only provides doctors with an intuitive growth assessment tool, but also makes it easier for parents to understand and track their children's growth process. Through clear graphical displays, parents can understand their children's developmental progress and whether there are abnormal development trends that require special attention.

[0140] Comprehensively analyze various test indicators, combine with the diagnostic standards for children's endocrine diseases, carry out multi-dimensional risk grading assessment, and graphically display the assessment results and explanations. In the process of children's growth and development and the diagnosis of endocrine diseases, multi-dimensional risk grading assessment is a key task. A single indicator cannot fully reflect the health status of children, so it is necessary to comprehensively analyze multiple test indicators, including growth hormone, thyroid hormone, sex hormone levels, etc., as well as developmental indicators such as height, weight, and bone age, and conduct multi-dimensional risk assessment through a comprehensive evaluation system. Technically, the comprehensive analysis combines various test indicators with clinical diagnostic standards, and conducts model training and evaluation based on relevant data of known endocrine disease standards (such as dwarfism, obesity, precocious puberty, etc.). First, by multi-dimensional modeling of children's endocrine hormone levels, growth rates, and sexual development, machine learning algorithms (such as random forests, support vector machines, etc.) are used to grade different health risks and give probability predictions for each risk level. These analysis results need to be presented in a graphical way so that doctors and parents can understand them more clearly. For example, using visualization techniques such as risk distribution maps and heat maps, the distribution of each risk level can be intuitively displayed, helping doctors quickly identify high-risk groups of children. Through graphical presentation, the results of risk assessment and the weight relationship of various indicators become more intuitive, ensuring that the assessment results are more scientific and easy to understand. Graphically displayed risk assessment results help doctors make more targeted decisions, such as whether further examinations are needed, whether intervention is recommended, etc. At the same time, parents can also understand their children's health risks through these clear diagrams, facilitating early prevention and intervention.

[0141] Personalized intervention recommendations are automatically generated based on the risk assessment results, and are prioritized according to the specific circumstances of the child, with scientific evidence and expected results. The formulation of intervention recommendations requires comprehensive consideration of the child's specific health status, endocrine hormone levels, developmental status, family history, and other relevant factors. Based on this comprehensive information, specific intervention measures can be automatically recommended, such as diet adjustment, exercise program, drug therapy, hormone therapy, etc., and ranked according to the priority of intervention. In order to ensure the scientific nature of the intervention recommendations, corresponding scientific evidence and expected results should also be provided. For example, if a child is assessed as being at high risk for obesity, a personalized diet and exercise intervention plan will be recommended, and relevant medical literature and clinical research data will be cited to illustrate the effectiveness and expected effects of this intervention. This not only enhances the credibility of the intervention recommendations, but also helps parents and doctors better understand the necessity of intervention and the feasibility of implementation. In addition, dynamic adjustment recommendations can be provided based on the implementation cycle and goals of the intervention to ensure the continuous optimization and updating of intervention measures. This personalized, dynamically adjusted intervention plan can be continuously optimized according to the growth changes of the child, thereby achieving the best health management effect.

[0142] The information is integrated to generate a hierarchical electronic report, in which the doctor version provides professional data, and the parent version uses plain language and illustrations to ensure that the information is prominent and easy to understand. The hierarchical structure of the electronic report includes a doctor version and a parent version, which are designed with different information presentation methods for different user groups. In the doctor version, the report will be presented with professional data and medical terms, covering all test indicators, analysis results and intervention recommendations to provide decision support for doctors. In the parent version, the report is presented in plain language and vivid illustrations to ensure that parents can understand the health status of their children and related intervention measures, helping parents make more scientific health decisions. This hierarchical electronic report not only increases the readability of the information, but also ensures that the needs of different users are met. Through clear charts, concise language and highlighting key content, parents can quickly grasp their children's health information, and doctors can obtain detailed professional data for clinical decision-making.

[0143] In summary, through multi-dimensional test index analysis, risk assessment, intervention recommendation generation and personalized report design, it can provide doctors and parents with a scientific, comprehensive and easy-to-understand child health management tool. This comprehensive solution not only improves the efficiency and accuracy of diagnosis and treatment, but also provides strong data support for children's early intervention and personalized health management.

[0144] Furthermore, based on the children's previous test data, combined with the standard growth curve, body mass index curve and sexual development assessment standards, a personalized development trajectory chart is generated, including the following steps.

[0145] Collect and organize all the children's test data, arrange all the test data in chronological order, and record the specific age at the time of the test.

[0146] Select the latest version of the standard reference curve, including the height standard curve, weight standard curve and BMI standard curve for the same gender and age group.

[0147] The children’s actual test data points are plotted onto the standard curve and connected to form the child’s individual growth and development trajectory, while also marking important developmental events and turning points.

[0148] The development of secondary sexual characteristics was assessed according to the Tanner staging criteria, and the progression of sexual development was correlated with height and weight development to fully understand the overall development status of the child.

[0149] Calculate development trends based on historical data, and calculate changing trends in height growth rate and weight growth rate.

[0150] Calculate the prediction interval of the child in the future and mark the upper and lower limits of the prediction interval on the trajectory graph. The prediction interval is expressed as: Y^(t f)±t α / 2 ·√[MSE·(1+1 / n+(t f -tˉ) 2 / ∑ r=1 n (t r -tˉ) 2 )], where Y^(t f） is predicted by the regression model at a certain time in the future f Estimated value of developmental index; t α / 2 is the critical value determined according to the selected confidence level; MSE represents the average of the squares of the differences between the model prediction value and the actual observed value; n is the number of sample data points, that is, the number of known historical test data; tˉ is the average value of the time data, which represents the average value of the age values ​​corresponding to all historical test data; t r Represents the age value corresponding to the rth data point.

[0151] In summary, through these steps, the generated personalized developmental trajectory map can accurately and comprehensively display the growth and development trends of children and provide a scientific basis for future development predictions. The technical solutions in each step are closely combined to form a complete child development assessment system based on data analysis. Through data collection and organization, standard curve selection, developmental trajectory drawing, developmental assessment, trend calculation and future prediction, it helps parents and doctors better understand the development status of children, detect developmental abnormalities in time and intervene.

[0152] Furthermore, step 6 includes the following steps.

[0153] The volume, concentration and uniformity parameters of blood samples are monitored in real time. When abnormalities are found, an alarm is automatically sounded and treatment suggestions are provided to ensure that the sample quality meets the standards. The quality of blood samples directly affects the accuracy of the test results. Therefore, real-time monitoring of the volume, concentration and uniformity of blood samples is a key link in ensuring the quality of the test. Monitoring of volume and concentration can help testers determine whether the sample is appropriate and whether it meets the laboratory test requirements. Blood samples often have uneven concentrations during processing, especially in microfluidic chips. Due to the limitations of blood viscosity and tiny channels, uneven sample concentrations will cause distortion of the results of some test points. Therefore, monitoring the uniformity of samples is equally important. Real-time monitoring of these parameters and automatic alarms when abnormalities are found can immediately take corrective measures to avoid unqualified samples from continuing to be used for testing and ensure the reliability of the final results. Technically, this monitoring mechanism uses a variety of sensors and image recognition technologies to collect relevant parameters of samples in real time. The sensor can accurately measure the volume, concentration and distribution of the sample, and automatically detect whether it meets the specified test conditions by comparing it with the preset standards. Once an abnormality is found, an alarm will automatically sound and provide treatment suggestions, such as resampling, adjusting the concentration or changing the sample processing method. This automated monitoring can greatly reduce human omissions and interventions, and improve laboratory work efficiency and accuracy.

[0154] The key parameters in the detection process are tracked and recorded throughout the process, and the detection channels of the microfluidic chip are monitored through multi-point sensors to ensure the optimal state of the detection process. In order to ensure that the detection process of blood samples such as endocrine hormones is as close to the ideal state as possible, the key parameters in the detection process must be tracked and recorded throughout the process. As a core tool in modern biomedical experiments, microfluidic chips can efficiently process small volumes of blood samples and perform accurate analysis. However, in actual operation, parameters such as temperature, pressure, and flow rate will have a significant impact on the test results. Therefore, it is crucial to establish a multi-point sensor system to monitor these key parameters in real time to ensure the optimal state of the detection process. This embodiment arranges sensors in each detection channel of the microfluidic chip to collect information such as the flow rate, temperature change, and internal pressure of the chip in real time to ensure that all detection parameters are in an optimized state. For example, too fast or too slow a flow rate may cause the sample to fail to fully react, and temperature changes may affect the stability of hormones and the sensitivity of detection. Therefore, these sensors will continue to detect and record any abnormalities in the detection process and provide real-time feedback to the experimenter through data transmission. If there are deviations that do not meet the standards during the test, the system will automatically prompt the operator to make adjustments to reduce the occurrence of errors. In terms of data recording, the collection and processing of key parameters are all recorded and automatically archived. In this way, complete test process data can be provided for later analysis and auditing. In addition, combined with advanced cloud computing technology, these data can also be remotely monitored and diagnosed online, ensuring that the quality of the test process can be guaranteed even in a multi-location and multi-laboratory environment.

[0155] Implement multi-level consistency assessment, compare the deviation between quality control samples and reference values ​​of children's endocrine indicators and analyze the results of repeated measurements to ensure that the test results are accurate and repeatable. Consistency assessment is the basis for ensuring the accuracy and reliability of test results. In the detection of endocrine hormone levels, it is necessary not only to ensure that the quality of the sample itself meets the standard, but also to conduct consistency inspections on the test equipment and results. Through multi-level consistency assessment, the deviation between the test results of quality control samples and the reference values ​​of children's endocrine indicators can be regularly compared, and combined with the results of repeated measurements to analyze whether there are systematic errors. This evaluation method can ensure that the test equipment will not have equipment aging or other problems during long-term use, resulting in reduced accuracy of the results. Technically, by setting standard quality control samples and comparing them with the reference values ​​of children's endocrine indicators within the normal range. Through this comparison, it is possible to judge in real time whether the test results of the current experimental process are normal. In addition, the results of repeated measurements also need to be analyzed. If the deviation between a test value and the reference value is large, or an indicator fluctuates abnormally in multiple repeated tests, the operator will be automatically reminded to check the equipment or replace the test sample. Through this multi-level consistency assessment, not only can the reliability of each test be ensured, but also the stability of the test process and the operating status of the equipment can be regularly verified. In addition, combined with data analysis tools, a detailed evaluation report can be generated for each test result, including deviation analysis, error source location, and recommended treatment measures. This function helps laboratory personnel discover potential quality problems at an early stage, reduce errors caused by improper equipment or operation, and ensure the reliability of test results.

[0156] Assign a unique identification code to each sample, use blockchain technology to record the entire process information, and ensure that the test results are traceable and verifiable. The uniqueness of the sample and the traceability of the test process are important links in laboratory management, especially in the field of medical testing. Assigning a unique identification code to each blood sample and using blockchain technology to record the entire process information is an effective means to ensure that the sample can be traced and verified at every stage from collection to testing. Through blockchain technology, the system can ensure the security and immutability of all data in a decentralized manner, and ensure the transparency and reliability of the test process and results.

[0157] Setting quality control standards based on historical data of children's endocrine tests, automatically detecting abnormalities and providing treatment plans to reduce human bias. In order to ensure the efficiency and accuracy of children's endocrine hormone level testing, setting quality control standards based on historical data and automatically detecting abnormalities are effective means to reduce human bias. By analyzing previous test data, reasonable quality control standards can be set and new test results can be monitored in real time to see if they meet these standards. Once an abnormality occurs, it will automatically prompt and provide possible solutions, reducing human errors and improving laboratory work efficiency.

[0158] Regularly generate quality assessment reports, summarize quality control data, provide a basis for equipment maintenance, process optimization and laboratory certification, and continuously improve the quality of testing. Regularly generating quality assessment reports is an important step to ensure the continuous improvement of testing quality. By summarizing quality control data and analyzing various parameters and equipment operating conditions during the testing process, the laboratory can identify potential problems and optimize them in periodic assessments. The quality assessment report can provide a basis for equipment maintenance, laboratory process optimization and laboratory certification, ensuring that the laboratory always maintains a high standard of testing.

[0159] In summary, through the technical implementation of the above steps, the entire blood sample testing process has been effectively quality controlled, ensuring that each link is in the best state. From sample quality monitoring to full-process data traceability, to multi-level consistency evaluation and automated anomaly detection, it can comprehensively improve the accuracy of detection and reduce human interference and errors. In modern medical testing, quality control is not only the test itself, but also includes data security, equipment maintenance and process optimization. Through the comprehensive application of these technical means, the laboratory can provide reliable test results while ensuring efficient operation, and continuously optimize the work process to improve the quality of medical services.

[0160] Finally, it should be pointed out that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them. Although the present invention has been described in detail with reference to the above embodiments, a person skilled in the art should understand that the technical solutions described in the above embodiments can still be modified, or some of the technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A rapid detection method for screening dwarfism in children, characterized in that: The following steps are involved: Step 1: Use a sampling device with integrated micro-blood collection and automatic dilution functions to accurately sample through the capillary effect and automatically dilute the sample to ensure accurate concentration; Step 2, introducing the collected samples into a multi-channel microfluidic detection chip, and realizing parallel quantitative detection of endocrine hormones through a nano-biorecognition probe array; Step 3: Based on high-sensitivity detectors and adaptive signal processing algorithms, weak biomolecule recognition signals are collected and amplified in real time, effectively improving the signal-to-noise ratio and ensuring the accuracy of the detection results; Step 4: Based on the deep learning algorithm and combined with the dynamically updated reference database of children's growth and endocrine indicators, the test results are analyzed in multiple dimensions to achieve quantitative evaluation of growth and development, body mass index and sexual development indicators and prediction of endocrine disease risks; Step 5: Automatically integrate the analysis results and generate a standardized test report, including quantitative test data, personalized growth curve, body mass index curve and sexual development assessment analysis, abnormal risk grading assessment and targeted intervention recommendations; Step 6: Establish a quality control mechanism throughout the entire testing process, including real-time monitoring of sample quality, tracking of testing process parameters, result consistency assessment, and full traceability of testing data.

2. The rapid detection method for screening dwarfism in children according to claim 1, characterized in that: Step 1 includes the following steps: Use 75% medical alcohol to fully disinfect the sampling head and dilution chamber of the sampling device, and check the metering accuracy of the device with a standard solution to ensure that the device is in the best working condition; Use alcohol cotton pads to disinfect the fingertips of children. After they are dried naturally, use a blood collection needle specially designed for children to perform puncture, and control the depth to 1.8-2.2 mm. Relying on the capillary effect, the preset volume of blood sample is automatically sucked in; After the blood reaches a preset volume, the buffer solution pre-installed in the dilution chamber is mixed with the blood sample in a precise proportion to form a sample solution with uniform concentration; After the dilution process is completed, the mixed liquid sample automatically flows into the collection chamber and is fully mixed to avoid concentration gradient phenomenon.

3. The rapid detection method for screening dwarfism in children according to claim 1, characterized in that: Step 2 includes the following steps: Preheat the multi-channel microfluidic detection chip to 37°C and pre-treat it with a dedicated buffer to ensure that the hydrophilicity of the inner wall of each microchannel reaches the optimal state to prevent the generation of bubbles and ensure the uniform flow of the sample in the channel; Use a microinjector to accurately introduce the mixed solution into the injection port of the chip. Under the action of microfluidics, the mixed solution is automatically distributed into multiple parallel detection channels. Start the microfluidic control module on the chip to accurately adjust the flow rate and residence time of the sample in each detection channel to ensure that the sample fully contacts and reacts with the nanoprobes in each channel; During the reaction, various endocrine hormones bind specifically to the corresponding nanoprobes, causing changes in the physical and chemical properties of the probes and real-time capture and recording; After the reaction is completed, the detection channel is gently rinsed with a dedicated cleaning buffer to remove unbound interfering substances; The signal acquisition module integrated in the chip can synchronously read the signal intensity of each detection channel and convert it into the specific concentration value of the corresponding hormone level according to the pre-established standard curve.

4. The rapid detection method for screening dwarfism in children according to claim 1, characterized in that: Step 3 includes the following steps: Under completely light-shielded conditions, record the baseline noise level of the detector and collect the background signal of the blank sample for subsequent signal processing and correction; Turn on the high-sensitivity photodetector and use a high-performance photomultiplier tube to collect the weak hormone-antibody binding signals of each detection channel of the chip in real time; The method combining wavelet transform and adaptive Wiener filtering is adopted to automatically adjust the filtering parameters according to the signal characteristics, effectively remove high-frequency random noise and low-frequency drift, and at the same time retain the effective information of the target signal to the greatest extent; The amplification factor is automatically adjusted according to the change of signal strength to ensure that weak signals are amplified while avoiding saturation distortion of strong signals, so that the signal is always kept within the optimal detection range, which can be expressed as: G = G min +(G max -G min )·[A threshold / (A in +A threshold )], where G is the gain factor, i.e. the magnification; G min is the minimum gain, i.e. the lower limit of the gain coefficient; G max is the maximum gain, that is, the upper limit of the gain coefficient; A in is the input signal strength; A threshold is a preset signal strength threshold used to determine whether a change in signal strength requires gain adjustment; Taking into account the signal strength, time characteristics and spatial distribution characteristics, the effective signal is accurately extracted through pattern recognition methods to further improve the signal-to-noise ratio, which is expressed as: SNR optimized =[I signal (T) F signal (ω)·S~ effective (k x ,k y ,k z )·M(S effective ,N)] / P noise , where SNR optimized is the optimized signal-to-noise ratio, that is, the signal-to-noise ratio that is finally improved after pattern recognition and multi-dimensional feature analysis; P noise is the noise power, which represents the average power of the noise signal in the entire time window; I signal (T) is the signal strength, that is, the average strength of the signal in a specific time window; F signal (ω) is the spectrum of the signal, that is, the frequency distribution of the signal obtained by Fourier transform, reflecting the distribution of the signal in different frequency components; S~ effective (k x ,k y ,k z） is the spatial spectrum of the effective signal, that is, the frequency component of the effective signal in the spatial domain; M(S effective ,N) is the pattern recognition function, which represents the relationship between the effective signal and the noise signal processed by the pattern recognition method; N is the noise signal; ω is the representation of the signal in the frequency domain; k x ,k y ,k z Represent the frequencies in the X, Y and Z directions respectively.

5. The rapid detection method for screening dwarfism in children according to claim 1, characterized in that: Step 4 includes the following steps: Establish digital archives for children, collect growth and development indicators, body mass index, sexual development stages, and bone age assessment information, and perform standardized processing to ensure that the data is suitable for deep learning model analysis; The standardized data is input into the deep learning model, and key features are extracted through multi-layer convolutional structure and attention mechanism, focusing on analyzing the dynamic changes of endocrine hormone levels and growth and development indicators; Using a dynamically updated reference database of children's growth and endocrine diseases, we use transfer learning technology to compare children's data with similar cases in multiple dimensions and generate quantitative evaluation indicators; Use ensemble learning methods to predict the risks of dwarfism, obesity, and precocious puberty, and calculate the risk probability of children developing related diseases in the future.

6. The rapid detection method for screening dwarfism in children according to claim 5, characterized in that: The standardized data is input into the deep learning model, and key features are extracted through a multi-layer convolutional structure and attention mechanism, focusing on analyzing the dynamic changes of endocrine hormone levels and growth and development indicators, including the following steps: Design a multi-layer convolutional neural network as the backbone network, which includes multiple convolutional layers, pooling layers and fully connected layers. The convolutional layers are responsible for extracting local features, the pooling layers are responsible for feature dimensionality reduction and main information extraction, and the fully connected layers are responsible for integrating high-dimensional features. Attention layers are embedded in key positions of multi-layer convolutional neural networks to automatically learn the importance weights between different features, so as to better focus on the key correspondence between endocrine hormones and various growth and development indicators; Use recurrent neural networks to process time series data to capture the patterns of changes in various indicators over time, with a special focus on the lagged effects of changes in hormone levels on growth and development; The extracted static features and temporal features are multimodally fused through a multi-layer perceptron to deeply analyze the nonlinear relationship between endocrine hormones and growth and development indicators, and a prediction model is established; The performance of the prediction model was evaluated through cross-validation, the prediction results were displayed using visualization technology, and the impact of different hormone indicators on growth and development was analyzed.

7. The rapid detection method for screening dwarfism in children according to claim 5, characterized in that: The ensemble learning method is used to predict the risk of dwarfism, obesity and precocious puberty, and calculate the risk probability of children developing related diseases in the future, including the following steps: Select multiple types of machine learning algorithms as base classifiers, each of which is modeled for dwarfism, obesity, and precocious puberty; The prediction results of the base classifiers are generated by cross-validation method, and logistic regression is used as a meta-learner to optimize the combination of the prediction outputs of the base classifiers. According to the individual characteristics of children, the incidence probabilities of dwarfism, obesity and precocious puberty were calculated, and a comprehensive evaluation was conducted based on the disease correlation, which was expressed as: R total =∑ j=1 3 α j ·P(d j |X)+λ·∑ i≠j P(d i ,d j |X), where R total It is a comprehensive risk assessment result, which indicates the overall risk of children developing dwarfism, obesity and precocious puberty, taking into account the correlation between individual characteristics and diseases; α j is the weight factor of the jth disease, which is used to measure the relative importance of the disease in the comprehensive risk assessment; P(d j |X) is the disease d under the given child individual characteristics X j The probability of disease occurrence; λ is the adjustment factor, which indicates the strength of the correlation between diseases; P(d i ,d j |X) is the disease d under the given child individual characteristics X i and d j The joint probability of simultaneous occurrence.

8. The rapid detection method for screening dwarfism in children according to claim 1, characterized in that: Step 5 includes the following steps: According to the clinical test report specifications, the quantitative data of endocrine hormones, growth and sexual development assessment results, and risk prediction analysis information are classified and formatted, and abnormal indicators are automatically marked and displayed in the report in a striking manner; Generate a personalized developmental trajectory chart based on the child's previous test data, combined with standard growth curves, body mass index curves and developmental assessment standards; Comprehensively analyze various test indicators, combine with the diagnostic standards for pediatric endocrine diseases, conduct multi-dimensional risk grading assessment, and graphically display the assessment results and explanations; Automatically generate personalized intervention recommendations based on risk assessment results, prioritize them according to the child's specific situation, and provide scientific evidence and expected results; The information is integrated into a hierarchical electronic report, in which the doctor version provides professional data and the parent version uses plain language and illustrations to ensure that the information is prominent and easy to understand.

9. The rapid detection method for screening dwarfism in children according to claim 8, characterized in that: Based on the child's previous test data, combined with the standard growth curve, body mass index curve and sexual development assessment standards, a personalized development trajectory chart is generated, including the following steps: Collect and organize the children's previous test data, arrange all test data in chronological order, and record the specific age at the time of the test; Select the latest version of the standard reference curve, including the height standard curve, weight standard curve and BMI standard curve for the same gender and age group; Plot the actual test data points of the child onto the standard curve and connect them to form the child's personal growth and development trajectory, while marking important developmental events and turning points; The development of secondary sexual characteristics was assessed according to the Tanner staging standard, and the correlation analysis between the sexual development process and height and weight development was performed to fully understand the overall development status of the child; Calculate the development trend based on historical data, and calculate the changing trend of height growth rate and weight growth rate; Calculate the prediction interval of the child in the future and mark the upper and lower limits of the prediction interval on the trajectory graph. The prediction interval is expressed as: Y^(t f )±t α / 2 ·√[MSE·(1+1 / n+(t f -tˉ) 2 / ∑ r=1 n (t r -tˉ) 2 )], where Y^(t f） is predicted by the regression model at a certain time in the future f Estimated value of developmental index; t α / 2 is the critical value determined according to the selected confidence level; MSE represents the average of the squares of the differences between the model prediction value and the actual observed value; n is the number of sample data points, that is, the number of known historical test data; tˉ is the average value of the time data, which represents the average value of the age values ​​corresponding to all historical test data; t r Represents the age value corresponding to the rth data point.

10. The rapid detection method for screening dwarfism in children according to claim 1, characterized in that: Step 6 includes the following steps: The volume, concentration and uniformity parameters of blood samples are monitored in real time. When abnormalities are found, an alarm is automatically issued and treatment suggestions are provided to ensure that the sample quality meets the standards. The key parameters of the detection process are tracked and recorded throughout the process, and the detection channel of the microfluidic chip is monitored through multi-point sensors to ensure the optimal state of the detection process; Implement multi-level consistency assessment, compare the deviations between quality control samples and reference values ​​of children's endocrine indicators and analyze repeated measurement results to ensure that the test results are accurate and repeatable; Assign a unique identification code to each sample and use blockchain technology to record the entire process information to ensure that the test results can be traced and verified; Set quality control standards based on historical data of children's endocrine testing, automatically detect abnormalities and provide treatment plans to reduce human bias; Regularly generate quality assessment reports, summarize quality control data, provide a basis for equipment maintenance, process optimization and laboratory certification, and continuously improve testing quality.

Images