A neural development disease detection system based on image recognition

Abstract

The application provides a neurodevelopmental disease detection system based on image recognition. The neurodevelopmental disease detection system based on image recognition comprises a data collection terminal, an image quality determination terminal, a parameter adjustment terminal and a deep learning detection terminal; wherein the data collection terminal and the image quality determination terminal are connected in data communication; the image quality determination terminal and the parameter adjustment terminal are connected in data communication; the parameter adjustment data information output end of the parameter adjustment terminal is connected in data communication with the data adjustment data information input end of the data collection terminal; and the medical image signal output end of the image quality determination terminal is connected in data communication with the image signal input end of the deep learning detection terminal.

Description

Technical Field

[0001] This invention proposes a neurodevelopmental disease detection system based on image recognition, belonging to the field of image recognition data processing technology. Background Technology

[0002] Neurodevelopmental disorders, including attention deficit hyperactivity disorder (ADHD) and learning disabilities, are a group of diseases that affect the normal development and function of children's nervous systems. These diseases typically exhibit specific behavioral, cognitive, and neuroimaging features in their early stages. However, due to their complexity and diversity, traditional clinical diagnosis often relies on the subjective judgment and experience of physicians, lacking objective and quantitative assessment methods. Therefore, developing an accurate, efficient, and objective detection system for neurodevelopmental disorders is crucial for improving diagnostic accuracy and enabling early intervention. In recent years, the rapid development of image processing and machine learning technologies, especially the widespread application of deep learning technology, has provided new ideas and methods for the detection of neurodevelopmental disorders. Deep learning models can automatically extract key disease-related information by learning features from large amounts of medical image data, thereby achieving automatic detection and assisted diagnosis of diseases.

[0003] However, in practical applications, image recognition-based neurodevelopmental disease detection systems still face many challenges. First, the quality of medical image acquisition directly affects the accuracy and reliability of subsequent analysis. Due to various factors during medical image acquisition (such as equipment performance, operator skill, and patient cooperation), image quality can vary significantly, thus impacting the detection performance of deep learning models. Second, the construction and training of deep learning models require a large amount of high-quality medical image data, but acquiring and labeling such data is often time-consuming and labor-intensive, and is subject to limitations such as privacy protection. Furthermore, differences in images acquired using different medical devices and with different acquisition parameters can also challenge the model's generalization ability. Summary of the Invention

[0004] This invention provides an image recognition-based detection system for neurodevelopmental disorders to address the aforementioned problems in the prior art. The technical solution adopted is as follows:

[0005] A neurodevelopmental disease detection system based on image recognition is disclosed. The system includes a data collection terminal, an image quality assessment terminal, a parameter adjustment terminal, and a deep learning detection terminal. The data collection terminal and the image quality assessment terminal are connected via data communication. The image quality assessment terminal and the parameter adjustment terminal are connected via data communication. The parameter adjustment data output terminal of the parameter adjustment terminal is connected via data communication with the parameter adjustment data input terminal of the data collection terminal. The medical image signal output terminal of the image quality assessment terminal is connected via data communication with the image signal input terminal of the deep learning detection terminal.

[0006] Furthermore, the data collection terminal is used to acquire medical images for the detection of neurodevelopmental diseases in real time, and to perform image preprocessing on the medical images to obtain preprocessed medical images;

[0007] The image quality assessment terminal is used to assess the image quality of the preprocessed medical image. If the preprocessed medical image does not meet the image quality requirements, the image is re-acquired. If the medical image meets the image quality requirements, the medical image that meets the image quality requirements is input to the deep learning detection terminal.

[0008] The parameter adjustment terminal is used to determine whether the image quality of the medical image after noise reduction meets the image quality requirements, and to adjust the unified image acquisition parameters of the medical device if the image quality after noise reduction still does not meet the image quality requirements.

[0009] The deep learning detection terminal is used to receive medical images that meet the image quality requirements in real time, and input the medical images that meet the image quality requirements into a deep learning model that has been constructed and trained for the detection of neurodevelopmental diseases, and obtain the detection results.

[0010] Furthermore, the data collection terminal includes:

[0011] The unified image acquisition parameter setting module is used to retrieve the original image acquisition parameters of medical images generated by medical devices for the detection of neurodevelopmental diseases, and set the unified image acquisition parameters according to the original image parameters.

[0012] The acquisition parameter adjustment terminal is used to adjust the original image acquisition parameters of the medical device according to the unified image acquisition parameters, and after the original image acquisition parameters are adjusted, control the medical device to acquire medical images for the detection of neurodevelopmental diseases.

[0013] Furthermore, the steps for setting the unified image acquisition parameters corresponding to the unified image acquisition parameter setting module include:

[0014] Extract the raw image acquisition parameters of the medical device, wherein the raw image acquisition parameters include resolution parameters, contrast parameters, and exposure time parameters;

[0015] Uniform image acquisition parameters are obtained using the raw image acquisition parameters of medical equipment, wherein the uniform image acquisition parameters include uniform resolution, uniform contrast, and uniform exposure time.

[0016] Furthermore, the unified resolution is obtained through the following steps:

[0017] Extract the types of medical devices collected by the data collection terminal;

[0018] Extract the device performance parameters of the medical devices corresponding to the above-mentioned device types. The device performance parameters include dynamic range parameters (i.e., the ratio between the maximum and minimum signals that the device can detect), sensitivity parameters, signal-to-noise ratio parameters, and uniformity parameters.

[0019] A device characteristic matrix is ​​generated using the device performance parameters, wherein the structure of the device characteristic matrix is ​​as follows:

[0020]

[0021] Where D represents the device characteristic matrix; F 01 F 02 ... F n L represents the dynamic range parameters of medical devices corresponding to n device types; 01 L 02 ... L n S represents the sensitivity parameters of medical devices corresponding to n device types; 01 S 02 ... S n J represents the signal-to-noise ratio parameters of medical devices corresponding to n device types; 01 J 02 ... J n These represent the uniformity parameters of medical equipment corresponding to n equipment types;

[0022] The nonlinear variation function f(D) is obtained using the device characteristic matrix D, where the function structure of f(D) is as follows:

[0023]

[0024] Where f(D) represents the nonlinear variation function; n represents the number of rows in the equipment characteristic matrix D, i.e., the number of equipment types in the medical equipment; m represents the number of columns in the equipment characteristic matrix D, i.e., the number of parameter types included in the equipment performance parameters; D ij Let represent the parameter value corresponding to the i-th row and j-th column of the device characteristic matrix D; ε represents the minimum constant used to ensure that the logarithmic function is non-negative within its domain; α, β, and γ represent the first adjustment parameter, the second adjustment parameter, and the third adjustment parameter, respectively, and these parameters are obtained using the following formula:

[0025] , ,

[0026] Where, α x+1 β x+1 and γ x+1 This represents the first, second, and third adjustment parameters corresponding to the (x+1)th iteration; α x β x and γ x This represents the first, second, and third adjustment parameters corresponding to the x-th iteration; This represents the loss function that includes the first adjustment parameter, the second adjustment parameter, and the third adjustment parameter; η represents the learning rate.

[0027] The uniform resolution is obtained using the nonlinear transformation function f(D), wherein the uniform resolution is obtained by the following formula:

[0028]

[0029] Among them, R t Represents uniform resolution; f(D) represents the nonlinear variation function; n represents the number of rows in the device characteristic matrix D, i.e., the number of types of medical equipment; R p R represents the average resolution corresponding to the original image acquisition parameters of n medical devices; i R represents the resolution corresponding to the original image acquisition parameters of the i-th type of medical device; z This represents the median resolution value corresponding to the original image acquisition parameters of n types of medical devices.

[0030] Furthermore, the uniform contrast is obtained through the following steps:

[0031] Extract the contrast parameters corresponding to the original image acquisition parameters of each medical device;

[0032] Retrieve the third adjustment parameter corresponding to the current nonlinear change function f(D);

[0033] A uniform contrast is obtained by combining the contrast parameter corresponding to the original image acquisition parameters of each medical device with the third adjustment parameter corresponding to the current nonlinear change function f(D), wherein the uniform contrast is obtained by the following formula:

[0034]

[0035] Among them, C t Indicates uniform contrast; n represents the number of rows in the device characteristic matrix D, i.e., the number of types of medical equipment; C p The value of C represents the average contrast ratio corresponding to the original image acquisition parameters of n medical devices; γ represents the third adjustment parameter; i C represents the contrast ratio corresponding to the original image acquisition parameters of the i-th type of medical device; z C represents the median contrast value corresponding to the original image acquisition parameters of n types of medical devices; max This represents the maximum contrast value corresponding to the original image acquisition parameters of n types of medical devices.

[0036] Furthermore, the uniform exposure time is obtained through the following steps:

[0037] Extract the exposure time parameters corresponding to the original image acquisition parameters of each medical device;

[0038] Retrieve the first and second adjustment parameters corresponding to the current nonlinear change function f(D);

[0039] A uniform exposure time is obtained by using the exposure time parameter corresponding to the original image acquisition parameters of each medical device and the first adjustment parameter and the second adjustment parameter corresponding to the current nonlinear change function f(D). The uniform exposure time is obtained using the following formula:

[0040]

[0041] Among them, T t The uniform exposure time is represented by α and β, which represent the first and second adjustment parameters, respectively. n represents the number of rows in the device characteristic matrix D, i.e., the number of medical device types. T i The exposure time corresponds to the original image acquisition parameters of the i-th type of medical device; f(D) represents the nonlinear variation function; T z T represents the median exposure time corresponding to the raw image acquisition parameters of n medical devices; min This represents the minimum exposure time corresponding to the original image acquisition parameters of n types of medical devices.

[0042] Furthermore, the image quality assessment terminal includes:

[0043] The medical image retrieval module is used to retrieve denoised medical images for the detection of neurodevelopmental diseases.

[0044] The image quality assessment execution module is used to assess the image quality of the denoised medical image for detecting neurodevelopmental diseases using the image parameters of the denoised medical image for detecting neurodevelopmental diseases, and to obtain the image quality assessment result.

[0045] The secondary acquisition control module is used to perform secondary acquisition of medical images when the image quality judgment results indicate that the denoised medical images used for the detection of neurodevelopmental diseases do not meet the image quality requirements.

[0046] Furthermore, the image quality determination steps of the image quality determination execution module include:

[0047] Image parameters are extracted from denoised medical images used for the detection of neurodevelopmental diseases, wherein the image parameters include signal-to-noise ratio, image uniformity, and artifact area;

[0048] The signal-to-noise ratio, image uniformity, and artifact area are used to obtain quality assessment parameters for denoised medical images used in the detection of neurodevelopmental diseases; wherein, the quality assessment parameters are obtained by the following formula:

[0049]

[0050] Where Q represents the quality evaluation parameter; w 01 w 02 and w 03 These represent the weighting coefficients corresponding to signal-to-noise ratio, image uniformity, and artifact area, respectively; S represents the value corresponding to signal-to-noise ratio; U represents the value corresponding to image uniformity; A represents the artifact area; A y This represents the preset artifact area threshold;

[0051] The quality evaluation parameters are compared with preset quality evaluation parameter thresholds;

[0052] When the quality evaluation parameter is lower than the preset quality evaluation parameter threshold, the medical image for detecting neurodevelopmental diseases that has undergone noise reduction is determined to not meet the image quality requirements.

[0053] Furthermore, the parameter adjustment terminal includes:

[0054] The data quality assessment module is used to determine whether the image quality of the medical image after noise reduction meets the image quality requirements after the image is re-acquired.

[0055] The second quality evaluation parameter retrieval module is used to extract the quality evaluation parameters corresponding to the image quality of the re-acquired medical image after noise reduction when the image quality of the re-acquired medical image after noise reduction still does not meet the image quality requirements, and use them as the second quality evaluation parameters.

[0056] The first quality evaluation parameter retrieval module is used to retrieve the quality evaluation parameters corresponding to the first image acquisition for the medical image that has been re-acquired, and use them as the first quality evaluation parameters.

[0057] The parameter adjustment execution module is used to adjust the unified image acquisition parameters using the first quality evaluation parameter and the second quality evaluation parameter;

[0058] The steps for adjusting the unified image acquisition parameters in the parameter adjustment execution module include:

[0059] Retrieve the first quality evaluation parameter and the second quality evaluation parameter, and use the first quality evaluation parameter and the second quality evaluation parameter to obtain the difference between the first quality evaluation parameter and the second quality evaluation parameter, as the first difference data;

[0060] The difference between the first quality evaluation parameter and the preset quality evaluation parameter threshold is retrieved as the second difference data.

[0061] The difference between the second quality evaluation parameter and the preset quality evaluation parameter threshold is retrieved as the third difference data.

[0062] The compensation amounts corresponding to uniform resolution, uniform contrast, and uniform exposure time are obtained using the first difference data, the second difference data, and the third difference data; wherein, the compensation amounts corresponding to uniform resolution, uniform contrast, and uniform exposure time are obtained by the following formula:

[0063]

[0064] Among them, Y R Y C and Y T These represent the compensation amounts corresponding to uniform resolution, uniform contrast, and uniform exposure time, respectively; f(D) represents the nonlinear variation function; k 01 k 02 and k 03 These represent the compensation adjustment coefficients corresponding to uniform resolution, uniform contrast, and uniform exposure time, respectively; ΔC 01 ΔC 02 and ΔC 03 These represent the first difference data, the second difference data, and the third difference data, respectively.

[0065] The unified resolution, unified contrast, and unified exposure time are adjusted using the compensation amounts corresponding to the unified resolution, unified contrast, and unified exposure time, respectively. The adjusted unified resolution, unified contrast, and unified exposure time are obtained using the following formulas:

[0066]

[0067] Among them, R z Indicates the adjusted uniform resolution; C z Indicates the adjusted uniform contrast; T z Indicates the adjusted uniform exposure time; Y R Y C and Y T These represent the compensation amounts corresponding to uniform resolution, uniform contrast, and uniform exposure time, respectively; R t Indicates uniform resolution; C t Indicates uniform contrast; T t Indicates uniform exposure time; λ R , λ C and λ T These represent the adjustment ratios corresponding to uniform resolution, uniform contrast, and uniform exposure time, respectively.

[0068] Beneficial effects of this invention:

[0069] This invention provides an image recognition-based neurodevelopmental disease detection system that integrates multiple modules, including data collection, image quality assessment, parameter adjustment, and deep learning detection. It aims to achieve efficient acquisition, quality control, parameter optimization, and disease detection of medical images. By acquiring medical images in real time and performing preprocessing and quality assessment, the system ensures that the image quality input to the deep learning model meets requirements. Simultaneously, the system optimizes the unified image acquisition parameters of the medical equipment through a parameter adjustment terminal, further improving image quality and detection accuracy. Finally, the system uses the constructed and trained deep learning model to detect medical images that meet quality requirements, obtaining objective and accurate detection results, providing strong support for the early diagnosis and intervention of neurodevelopmental diseases. Through rigorous image quality control and precise analysis by the deep learning model, this system can significantly improve the detection accuracy of neurodevelopmental diseases and reduce misdiagnosis and missed diagnosis. Because the system can acquire and analyze medical images in real time, it can detect abnormal signals in the early stages of disease, providing doctors with timely intervention suggestions and improving patient prognosis. The automated and intelligent workflow reduces the need for manual intervention and improves overall work efficiency. Doctors can obtain detection results more quickly and formulate treatment plans accordingly. The system generates a large amount of high-quality medical image data and test results, providing valuable data support for scientific research on neurodevelopmental diseases and promoting scientific and technological progress and development in related fields.

[0070] In summary, the above-mentioned technical solution integrates multiple modules such as data collection, image quality assessment, parameter adjustment, and deep learning detection, achieving efficient and accurate detection of neurodevelopmental diseases, and has significant technical effects and clinical application value. Attached Figure Description

[0071] Figure 1 This is a system block diagram of the system described in this invention. Detailed Implementation

[0072] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0073] This invention proposes an image recognition-based neurodevelopmental disease detection system, such as... Figure 1 As shown, the image recognition-based neurodevelopmental disease detection system includes a data collection terminal, an image quality assessment terminal, a parameter adjustment terminal, and a deep learning detection terminal. The data collection terminal and the image quality assessment terminal are connected via data communication. The image quality assessment terminal and the parameter adjustment terminal are connected via data communication. The parameter adjustment data output terminal of the parameter adjustment terminal is connected via data communication with the parameter adjustment data input terminal of the data collection terminal. The medical image signal output terminal of the image quality assessment terminal is connected via data communication with the image signal input terminal of the deep learning detection terminal.

[0074] The data collection terminal is used to acquire medical images for the detection of neurodevelopmental diseases in real time, and to perform image preprocessing on the medical images to obtain preprocessed medical images.

[0075] The image quality assessment terminal is used to assess the image quality of the preprocessed medical image. If the preprocessed medical image does not meet the image quality requirements, the image is re-acquired. If the medical image meets the image quality requirements, the medical image that meets the image quality requirements is input to the deep learning detection terminal.

[0076] The parameter adjustment terminal is used to determine whether the image quality of the medical image after noise reduction meets the image quality requirements, and to adjust the unified image acquisition parameters of the medical device if the image quality after noise reduction still does not meet the image quality requirements.

[0077] The deep learning detection terminal is used to receive medical images that meet the image quality requirements in real time, and input the medical images that meet the image quality requirements into a deep learning model that has been constructed and trained for the detection of neurodevelopmental diseases, and obtain the detection results.

[0078] The working principle of the above technical solution is as follows: real-time acquisition of medical images for the detection of neurodevelopmental disorders. These images may come from different medical devices, such as MRI and CT.

[0079] Inside the data collection terminal, the acquired raw medical images are preprocessed to remove noise, enhance contrast, and improve image quality, laying the foundation for subsequent analysis.

[0080] The image quality assessment terminal receives the preprocessed medical images and assesses their quality. This step uses a series of algorithms to evaluate whether the image meets preset quality requirements, such as sharpness and signal-to-noise ratio. If the image does not meet the quality requirements, the system will send feedback to the data collection terminal for image re-acquisition.

[0081] If the image quality after noise reduction still does not meet the requirements after re-acquiring, the parameter adjustment terminal will intervene. This terminal is responsible for analyzing the reasons for the insufficient image quality and adjusting the unified image acquisition parameters of the medical equipment, such as exposure time and resolution, in order to obtain higher quality images in the next acquisition. Once the image quality meets the requirements, the image quality assessment terminal will transmit the qualified medical image to the deep learning detection terminal.

[0082] The deep learning detection terminal has a pre-built and trained deep learning model specifically designed for the detection of neurodevelopmental disorders. After receiving a medical image, the model automatically extracts key features and compares them with a pre-stored disease feature database to provide the final detection result.

[0083] The effects of the above technical solution are as follows: This embodiment provides an image recognition-based neurodevelopmental disease detection system that integrates multiple modules such as data collection, image quality assessment, parameter adjustment, and deep learning detection. It aims to achieve efficient acquisition, quality control, parameter optimization, and disease detection of medical images. By acquiring medical images in real time and performing preprocessing and quality assessment, the system ensures that the image quality input to the deep learning model meets the requirements. Simultaneously, the unified image acquisition parameters of the medical equipment are optimized through a parameter adjustment terminal, further improving image quality and detection accuracy. Finally, the system uses the constructed and trained deep learning model to detect medical images that meet the quality requirements, obtaining objective and accurate detection results, providing strong support for the early diagnosis and intervention of neurodevelopmental diseases. Through strict image quality control and precise analysis by the deep learning model, the system can significantly improve the detection accuracy of neurodevelopmental diseases and reduce the occurrence of misdiagnosis and missed diagnosis. Because the system can acquire and analyze medical images in real time, it can detect abnormal signals in the early stages of the disease, providing doctors with timely intervention suggestions and improving patient prognosis. The automated and intelligent workflow reduces the need for manual intervention and improves overall work efficiency. Doctors can obtain detection results more quickly and formulate treatment plans accordingly. The system generates a large amount of high-quality medical image data and test results, providing valuable data support for scientific research on neurodevelopmental diseases and promoting scientific and technological progress and development in related fields.

[0084] In summary, the above-mentioned technical solution integrates multiple modules such as data collection, image quality assessment, parameter adjustment, and deep learning detection, achieving efficient and accurate detection of neurodevelopmental diseases, and has significant technical effects and clinical application value.

[0085] In one embodiment of the present invention, the data collection terminal includes:

[0086] The unified image acquisition parameter setting module is used to retrieve the original image acquisition parameters of medical images generated by medical devices for the detection of neurodevelopmental diseases, and set the unified image acquisition parameters according to the original image parameters.

[0087] The acquisition parameter adjustment terminal is used to adjust the original image acquisition parameters of the medical device according to the unified image acquisition parameters, and after the original image acquisition parameters are adjusted, control the medical device to acquire medical images for the detection of neurodevelopmental diseases.

[0088] The steps for setting the unified image acquisition parameters corresponding to the unified image acquisition parameter setting module include:

[0089] Extract the raw image acquisition parameters of the medical device, wherein the raw image acquisition parameters include resolution parameters, contrast parameters, and exposure time parameters;

[0090] Uniform image acquisition parameters are obtained using the raw image acquisition parameters of medical equipment, wherein the uniform image acquisition parameters include uniform resolution, uniform contrast, and uniform exposure time.

[0091] The working principle of the above technical solution is as follows: First, the module retrieves the original image acquisition parameters of medical images used for the detection of neurodevelopmental diseases from the medical device. These parameters typically include resolution, contrast, and exposure time parameters, which directly affect the image clarity and quality. Next, using the extracted original image acquisition parameters and a preset standard or optimization algorithm, unified image acquisition parameters are calculated. These unified parameters (such as unified resolution, unified contrast, and unified exposure time) aim to ensure that images acquired by different medical devices or by the same device under different conditions have consistent quality standards. Based on the unified parameters provided by the unified image acquisition parameter setting module, the terminal adjusts the original image acquisition parameters of the medical device. This step is automated, requiring no manual intervention, ensuring the accuracy and efficiency of the adjustment. After the original image acquisition parameters are adjusted, the acquisition parameter adjustment terminal controls the medical device to acquire medical images according to the new unified parameters. The acquired images thus have higher quality consistency, which is beneficial for subsequent image processing and disease detection.

[0092] The above technical solution achieves the following results: By unifying the setting and adjustment of image acquisition parameters, it ensures that medical images acquired by different medical devices or the same device under different conditions maintain consistent quality standards. This solution reduces detection errors caused by differences in image quality. Automated parameter adjustment and control processes reduce the need for manual intervention and improve detection efficiency. Doctors can obtain high-quality medical images more quickly and use them for disease diagnosis and analysis. High-quality medical images are the foundation of precision medicine. By ensuring consistent image quality, this system provides strong support for the accurate detection of neurodevelopmental diseases, enabling doctors to develop more personalized treatment plans. Consistent and high-quality medical image datasets are of great significance to scientific research. The datasets generated by this system can be used for in-depth research on neurodevelopmental diseases, promoting technological progress and development in related fields.

[0093] In summary, the above technical solutions, through the unified setting and adjustment of image acquisition parameters and automated acquisition control processes, achieve consistent medical image quality and high-efficiency acquisition, providing strong support for the accurate detection of neurodevelopmental diseases.

[0094] In one embodiment of the present invention, the uniform resolution is obtained through the following steps:

[0095] Extract the types of medical devices collected by the data collection terminal;

[0096] Extract the device performance parameters of the medical devices corresponding to the above-mentioned device types. The device performance parameters include dynamic range parameters (i.e., the ratio between the maximum and minimum signals that the device can detect), sensitivity parameters, signal-to-noise ratio parameters, and uniformity parameters.

[0097] A device characteristic matrix is ​​generated using the device performance parameters, wherein the structure of the device characteristic matrix is ​​as follows:

[0098]

[0099] Where D represents the device characteristic matrix; F 01 F 02 ... F n L represents the dynamic range parameters of medical devices corresponding to n device types; 01 L 02 ... L n S represents the sensitivity parameters of medical devices corresponding to n device types; 01 S 02 ... S n J represents the signal-to-noise ratio parameters of medical devices corresponding to n device types; 01 J 02 ... J n These represent the uniformity parameters of medical equipment corresponding to n equipment types;

[0100] The nonlinear variation function f(D) is obtained using the device characteristic matrix D, where the function structure of f(D) is as follows:

[0101]

[0102] Where f(D) represents the nonlinear variation function; n represents the number of rows in the equipment characteristic matrix D, i.e., the number of equipment types in the medical equipment; m represents the number of columns in the equipment characteristic matrix D, i.e., the number of parameter types included in the equipment performance parameters; D ij Let represent the parameter value corresponding to the i-th row and j-th column of the device characteristic matrix D; ε represents the minimum constant used to ensure that the logarithmic function is non-negative within its domain; α, β, and γ represent the first adjustment parameter, the second adjustment parameter, and the third adjustment parameter, respectively, and these parameters are obtained using the following formula:

[0103] , ,

[0104] Where, α x+1 β x+1 and γx+1 This represents the first, second, and third adjustment parameters corresponding to the (x+1)th iteration; α x β x and γ x This represents the first, second, and third adjustment parameters corresponding to the x-th iteration; This represents the loss function that includes the first adjustment parameter, the second adjustment parameter, and the third adjustment parameter; η represents the learning rate.

[0105] The uniform resolution is obtained using the nonlinear transformation function f(D), wherein the uniform resolution is obtained by the following formula:

[0106]

[0107] Among them, R t Represents uniform resolution; f(D) represents the nonlinear variation function; n represents the number of rows in the device characteristic matrix D, i.e., the number of types of medical equipment; R p R represents the average resolution corresponding to the original image acquisition parameters of n medical devices; i R represents the resolution corresponding to the original image acquisition parameters of the i-th type of medical device; z This represents the median resolution value corresponding to the original image acquisition parameters of n types of medical devices.

[0108] The working principle of the above technical solution is as follows: First, the system extracts information about the types of medical devices collected by the data collection terminal. Next, for each type of device, its corresponding performance parameters are extracted, including dynamic range, sensitivity, signal-to-noise ratio, and uniformity. These parameters directly reflect the imaging capability and image quality of the device. Using the extracted device performance parameters, a device characteristic matrix D is constructed. Each row of this matrix represents a type of device, and each column represents a performance parameter. A nonlinear variation function f(D) is generated from the device characteristic matrix D. This function considers the diversity of device types and performance parameters, balancing the differences between different devices through complex mathematical operations (such as logarithmic functions and iterative parameter adjustment). During the calculation of the iteratively adjusted parameters (α, β, γ), a loss function is used to evaluate the function performance under the current parameter values, and the parameter values ​​are gradually optimized using a learning rate η until convergence is achieved.

[0109] Finally, using the nonlinear transformation function f(D) and the original resolution information of the devices (including the average resolution Rp, the resolution Ri for each device, and the median resolution Rz), the uniform resolution Rt is calculated. This step takes into account the overall performance of the devices, individual differences, and the distribution of resolutions, aiming to find a uniform resolution value that is neither too high nor too low and applicable to most devices.

[0110] The effects of the above technical solution are as follows: By setting a uniform resolution, images acquired by different medical devices tend to have consistent resolution, reducing image quality fluctuations caused by device differences. Automated computation reduces the need for manual intervention, improving detection efficiency and accuracy. Doctors can obtain high-quality medical images more quickly and use them for disease diagnosis and analysis. The above technical solution can flexibly adapt to different types and performance levels of medical devices, dynamically adjusting the uniform resolution value through a nonlinear variation function, enhancing the system's adaptability and robustness. Consistent and high-quality medical image datasets are of great significance for scientific research. The datasets generated by the above technical solution can be used for in-depth research on neurodevelopmental diseases, promoting the translation of research results into clinical applications.

[0111] On the other hand, the aforementioned technical solution first extracts the types of medical devices and their corresponding performance parameters, including dynamic range, sensitivity, signal-to-noise ratio, and uniformity. These parameters comprehensively reflect the imaging quality and performance of the medical devices, providing a solid foundation for subsequent resolution unification. Simultaneously, by constructing a device characteristic matrix, the technical solution integrates and quantifies the performance parameters of different types of medical devices. This matrix form facilitates subsequent calculations and analysis, and also provides convenience for the introduction of nonlinear variation functions. The introduction of the nonlinear variation function f(D) is one of the core innovations of the aforementioned technical solution. This function can dynamically adjust the resolution of different medical devices based on the parameter values ​​in the device characteristic matrix, thereby achieving resolution unification. The aforementioned nonlinear processing method fully considers the differences and complexity of medical device performance parameters, making resolution adjustment more precise and flexible.

[0112] The above technical solution determines the adjustment parameters (α, β, and γ) in the nonlinear variation function through iterative optimization. This iterative process gradually approximates the optimal solution, allowing the nonlinear variation function to more accurately reflect the performance differences of medical devices and achieve effective resolution unification. After determining the nonlinear variation function and adjustment parameters, the above technical solution calculates the unified resolution using a formula. This unified resolution considers both the performance differences of different medical devices and ensures the clarity and accuracy of image acquisition. This precise calculation method improves the quality of medical images and the accuracy of diagnosis.

[0113] Meanwhile, by achieving a unified resolution across medical devices, the aforementioned technical solution enhances the comparability and interoperability of images acquired by different devices. This solution enables doctors to compare and analyze images across different devices, thereby improving diagnostic accuracy and efficiency. Achieving unified resolution also optimizes the utilization of medical resources. Doctors can more flexibly select and use different medical devices without worrying about differences in image quality and resolution. This technical solution can reduce medical costs and improve the efficiency and quality of medical services. Furthermore, it provides strong support for the development of medical informatization. By achieving a unified resolution across medical devices, image data acquired by different devices can be integrated and analyzed more conveniently, thus accelerating the development of medical informatization.

[0114] In summary, the above technical solution, by comprehensively considering the performance parameters of medical equipment and the original resolution information, uses a nonlinear variation function to calculate a uniform resolution value, aiming to improve the consistency of medical image quality and the efficiency and accuracy of the detection process.

[0115] In one embodiment of the present invention, the uniform contrast is obtained through the following steps:

[0116] Extract the contrast parameters corresponding to the original image acquisition parameters of each medical device;

[0117] Retrieve the third adjustment parameter corresponding to the current nonlinear change function f(D);

[0118] A uniform contrast is obtained by combining the contrast parameter corresponding to the original image acquisition parameters of each medical device with the third adjustment parameter corresponding to the current nonlinear change function f(D), wherein the uniform contrast is obtained by the following formula:

[0119]

[0120] Among them, C t Indicates uniform contrast; n represents the number of rows in the device characteristic matrix D, i.e., the number of types of medical equipment; C p The value of C represents the average contrast ratio corresponding to the original image acquisition parameters of n medical devices; γ represents the third adjustment parameter; i C represents the contrast ratio corresponding to the original image acquisition parameters of the i-th type of medical device; z C represents the median contrast value corresponding to the original image acquisition parameters of n types of medical devices; max This represents the maximum contrast value corresponding to the original image acquisition parameters of n types of medical devices.

[0121] The working principle of the above technical solution is as follows: The system first extracts the contrast parameter corresponding to the original image acquisition parameters of each medical device. This contrast parameter reflects the device's sensitivity to brightness differences during image acquisition. Next, the system retrieves the third adjustment parameter γ corresponding to the current nonlinear change function f(D). This parameter, obtained through iterative optimization in previous steps, is used to adjust the calculation of the uniform contrast. Using the extracted contrast parameter and the third adjustment parameter γ, the system calculates the uniform contrast using a specific formula. This formula comprehensively considers the average contrast of n medical devices, the contrast of each device, the median contrast value, and the maximum contrast value. The formula is designed to balance the contrast differences between different devices, finding a uniform contrast that reflects the characteristics of most devices without deviating too much from extreme values.

[0122] The effects of the above technical solution are as follows: By setting a unified contrast ratio, images acquired by different medical devices tend to have consistent contrast, reducing differences in image brightness and loss of detail caused by device variations. A unified contrast ratio makes images clearer and more layered when displayed, allowing doctors to better observe and analyze detailed information within the images. This technical solution can flexibly adapt to different types and performance levels of medical devices, dynamically adjusting the unified contrast ratio value through the third adjustment parameter of a nonlinear variation function, enhancing the system's adaptability and robustness. Setting a unified contrast ratio promotes the standardization of medical image acquisition and processing, improves interoperability between different devices, and facilitates data sharing and exchange.

[0123] The above technical solution first extracts the contrast parameter from the original image acquisition parameters of each medical device. This is a crucial step in obtaining a uniform contrast ratio, as contrast is one of the important indicators of image quality, directly affecting the visual effect and diagnostic accuracy of the image. Accurate extraction of the contrast parameter provides a reliable data foundation for subsequent uniform contrast ratio calculations.

[0124] The above technical solution retrieves the third adjustment parameter γ corresponding to the current nonlinear transformation function f(D). This parameter plays a role in adjusting the contrast within the nonlinear transformation function. By introducing γ, the contrast of different medical devices can be adjusted and optimized under a unified standard. The method described above, utilizing the nonlinear transformation function and adjustment parameters, fully considers the performance differences of medical devices and the diversity of image acquisition parameters, providing an effective means to achieve uniform contrast.

[0125] After obtaining the contrast and adjustment parameters, the technical solution calculates the uniform contrast ratio Ct using a formula. This formula comprehensively considers multiple factors, including the type and number of medical devices, the average contrast ratio, the median contrast ratio, the maximum contrast ratio, and a third adjustment parameter, ensuring the accuracy and reliability of the calculation results. By accurately calculating the uniform contrast ratio, images acquired by different medical devices can maintain consistent contrast, improving image comparability and diagnostic accuracy. Achieving uniform contrast ratio optimizes the quality of medical images. By adjusting the contrast of different medical devices, images become visually clearer and brighter, allowing doctors to better observe and analyze image details, thus improving diagnostic accuracy and efficiency.

[0126] Meanwhile, the aforementioned technical solutions provide strong support for medical informatization and standardization. By achieving uniform contrast across medical devices, image data collected from different devices can be integrated and analyzed more conveniently, promoting the development of medical informatization. Furthermore, achieving uniform contrast also facilitates the formulation and promotion of medical standards, providing a strong guarantee for improving medical quality and optimizing medical services. Uniform contrast makes it more convenient and efficient for doctors to view and analyze images, eliminating the need to spend extra time and effort adjusting images due to contrast differences between different medical devices. This not only improves the efficiency of medical services but also enhances the satisfaction of both doctors and patients.

[0127] In summary, the above technical solution achieves uniform contrast in medical devices through steps such as accurately extracting contrast parameters, utilizing nonlinear variation functions and adjusting parameters, and precisely calculating unified contrast. This not only optimizes the quality of medical images and improves the accuracy and efficiency of diagnosis, but also promotes the development of medical informatization and standardization, demonstrating significant technical effectiveness and application value. Furthermore, by comprehensively considering the original contrast parameters of the medical device and the third adjustment parameter of the nonlinear variation function, the above technical solution aims to improve the consistency and visual effect of medical image quality, while enhancing the system's adaptability and promoting standardization and interoperability.

[0128] In one embodiment of the present invention, the uniform exposure time is obtained through the following steps:

[0129] Extract the exposure time parameters corresponding to the original image acquisition parameters of each medical device;

[0130] Retrieve the first and second adjustment parameters corresponding to the current nonlinear change function f(D);

[0131] A uniform exposure time is obtained by using the exposure time parameter corresponding to the original image acquisition parameters of each medical device and the first adjustment parameter and the second adjustment parameter corresponding to the current nonlinear change function f(D). The uniform exposure time is obtained using the following formula:

[0132]

[0133] Among them, T t The uniform exposure time is represented by α and β, which represent the first and second adjustment parameters, respectively. n represents the number of rows in the device characteristic matrix D, i.e., the number of medical device types. T i The exposure time corresponds to the original image acquisition parameters of the i-th type of medical device; f(D) represents the nonlinear variation function; T z T represents the median exposure time corresponding to the raw image acquisition parameters of n medical devices; min This represents the minimum exposure time corresponding to the original image acquisition parameters of n types of medical devices.

[0134] The working principle of the above technical solution is as follows: First, the system extracts the exposure time parameter (T) corresponding to the original image acquisition parameters from various medical devices. i The above parameters directly reflect the exposure time required by the device when acquiring images. Next, the system retrieves the first adjustment parameter α and the second adjustment parameter β corresponding to the current nonlinear change function f(D). These two parameters were obtained through complex mathematical calculations and iterative optimization in previous steps, aiming to adjust the exposure time according to the device's performance characteristics and image acquisition requirements. Using the extracted exposure time parameter (T... i ) and the retrieved first and second adjustment parameters (α, β), combined with the nonlinear change function f(D) and the median exposure time (T) z ) and minimum exposure time (T) min The uniform exposure time (T) is calculated using a specific formula. t The above formula was designed considering various factors, including the diversity of equipment, the needs of image acquisition, and the characteristics of nonlinear changes, with the aim of finding a uniform exposure time value that both conforms to the characteristics of most equipment and meets the requirements for image acquisition quality.

[0135] The effects of the above technical solution are as follows: By setting a unified exposure time, different medical devices can use more similar exposure durations when acquiring images, thereby improving the overall efficiency of image acquisition. Doctors or operators no longer need to set different exposure times for different devices, simplifying the operation process. A unified exposure time can reduce problems such as uneven image brightness and excessive noise caused by device differences, improving the overall image quality. This is particularly important for medical fields such as neurodevelopmental diseases that require high-precision image analysis. The above technical solution can flexibly cope with different types and performance of medical devices, ensuring the accuracy and applicability of the unified exposure time through nonlinear variation functions and dynamic adjustment of adjustment parameters. This adaptability enhances the robustness and stability of the system. Setting a unified exposure time can promote the standardization and normalization of medical image acquisition. It makes the exposure time of images acquired by different devices more consistent, facilitating subsequent image processing and analysis. At the same time, this also provides convenient conditions for data sharing and exchange between different medical institutions.

[0136] The above technical solution first extracts the exposure time parameter from the original image acquisition parameters of each medical device. Exposure time is one of the key factors affecting image quality and clarity, especially in the field of medical imaging, where accurate exposure time is crucial for accurate diagnosis. By precisely extracting the exposure time parameter, a reliable data foundation is provided for subsequent unified exposure time calculation. The above technical solution retrieves the first adjustment parameter α and the second adjustment parameter β corresponding to the current nonlinear change function f(D). These two parameters play a role in adjusting the exposure time in the nonlinear change function. Their introduction allows the exposure time of different medical devices to be adjusted and optimized under a unified standard. The above method, utilizing nonlinear change functions and adjustment parameters, fully considers the performance differences of medical devices and the diversity of image acquisition parameters, providing an effective means to achieve unified exposure time.

[0137] Meanwhile, after obtaining the exposure time parameters and adjustment parameters, the above technical solution calculates the uniform exposure time T using a formula. t The above formula comprehensively considers multiple factors, including the type and number of medical devices, the initial exposure time for each device, the nonlinear variation function, the median exposure time, and the minimum exposure time, ensuring the accuracy and reliability of the calculation results. By accurately calculating a uniform exposure time, images acquired by different medical devices can be kept consistent in exposure, improving image comparability and diagnostic accuracy. Achieving a uniform exposure time can improve the quality of medical images. By adjusting the exposure time of different medical devices, images are exposed more evenly and clearly, allowing doctors to better observe and analyze image details, improving diagnostic accuracy and efficiency. Especially in low-light environments, correct exposure time is crucial to avoiding overexposure or underexposure of images.

[0138] The implementation of standardized exposure times enhances the compatibility and interoperability between different medical devices. Doctors can more flexibly select and use different medical devices without worrying about differences in image exposure. This can reduce medical costs and improve the efficiency and quality of medical services. The aforementioned technical solutions provide strong support for medical standardization and normalization. By achieving standardized exposure times for medical devices, the standardization process in the field of medical imaging can be promoted, providing a strong guarantee for improving medical quality and optimizing medical services. Standardized exposure times reduce human intervention and errors when doctors adjust image exposure. Doctors can focus more on image diagnosis and analysis, rather than spending time and energy adjusting image exposure. This not only improves the efficiency of medical services but also enhances diagnostic accuracy.

[0139] In summary, the above technical solution, by comprehensively considering multiple factors and employing nonlinear variation functions and parameter adjustment methods to calculate the uniform exposure time, aims to improve image acquisition efficiency, optimize image quality, enhance system adaptability, and promote standardization and normalization. The aforementioned technical effects are significant for improving the accuracy and reliability of medical image diagnosis. Furthermore, this technical solution achieves uniform exposure time for medical devices through precise extraction of exposure time parameters, utilization of nonlinear variation functions and parameter adjustment, and accurate calculation of the uniform exposure time. This not only improves the quality of medical images and enhances the compatibility and interoperability of medical devices but also promotes the development of medical standardization and normalization, demonstrating significant technical effects and application value.

[0140] In one embodiment of the present invention, the image quality determination terminal includes:

[0141] The medical image retrieval module is used to retrieve denoised medical images for the detection of neurodevelopmental diseases.

[0142] The image quality assessment execution module is used to assess the image quality of the denoised medical image for detecting neurodevelopmental diseases using the image parameters of the denoised medical image for detecting neurodevelopmental diseases, and to obtain the image quality assessment result.

[0143] The secondary acquisition control module is used to perform secondary acquisition of medical images when the image quality judgment results indicate that the denoised medical images used for the detection of neurodevelopmental diseases do not meet the image quality requirements.

[0144] The image quality determination steps of the image quality determination execution module include:

[0145] Image parameters are extracted from denoised medical images used for the detection of neurodevelopmental diseases, wherein the image parameters include signal-to-noise ratio, image uniformity, and artifact area;

[0146] The signal-to-noise ratio, image uniformity, and artifact area are used to obtain quality assessment parameters for denoised medical images used in the detection of neurodevelopmental diseases; wherein, the quality assessment parameters are obtained by the following formula:

[0147]

[0148] Where Q represents the quality evaluation parameter; w 01 w 02 and w 03 These represent the weighting coefficients corresponding to signal-to-noise ratio, image uniformity, and artifact area, respectively; S represents the value corresponding to signal-to-noise ratio; U represents the value corresponding to image uniformity; A represents the artifact area; A y This represents the preset artifact area threshold;

[0149] The quality evaluation parameters are compared with preset quality evaluation parameter thresholds;

[0150] When the quality evaluation parameter is lower than the preset quality evaluation parameter threshold, the medical image for detecting neurodevelopmental diseases that has undergone noise reduction is determined to not meet the image quality requirements.

[0151] The working principle of the above technical solution is as follows: The medical image retrieval module is responsible for retrieving denoised medical images used for the detection of neurodevelopmental diseases from the storage system. These images form the basis for subsequent quality assessment. The image quality assessment execution module first extracts key image parameters from the above medical images, including signal-to-noise ratio (S), image uniformity (U), and artifact area (A).

[0152] Next, using the aforementioned image parameters, the quality assessment parameter (Q) is calculated using a specific formula. This formula comprehensively considers the importance of signal-to-noise ratio, image uniformity, and artifact area, and uses weighting coefficients (w... 01 w 02 w 03 To achieve balance, an artifact area threshold (A) is also introduced into the formula. y To further evaluate the impact of artifacts on image quality, the calculated quality assessment parameter (Q) is compared with a preset quality assessment parameter threshold. This threshold, set based on actual needs and experience, is used to determine whether the image meets quality requirements. If the quality assessment parameter (Q) is lower than the preset threshold, the medical image is deemed not to meet image quality requirements. In this case, the secondary acquisition control module will be activated to guide or automatically perform a secondary acquisition of the medical image.

[0153] The above technical solution achieves the following results: Through a rigorous image quality assessment process, it ensures that medical images used for the detection of neurodevelopmental diseases are of sufficiently high quality. This solution enables doctors to diagnose diseases more accurately, improving treatment outcomes. Low-quality medical images may lead to misdiagnosis or missed diagnosis, posing unnecessary risks to patients. The above technical solution reduces these risks by promptly identifying and correcting image quality issues. Automated image quality assessment and secondary acquisition control processes reduce the need for manual intervention, improving work efficiency. Simultaneously, it reduces the possibility of human error. High-quality medical images and accurate diagnostic results enhance patient satisfaction and trust in medical services. This is of great significance to the reputation and long-term development of medical institutions.

[0154] The above technical solution first extracts image parameters from the denoised medical image, including signal-to-noise ratio (S), image uniformity (U), and artifact area (A). These parameters are key indicators for evaluating image quality, comprehensively reflecting the image's sharpness, consistency, and defects. The quality evaluation parameter (Q) is calculated using the formula described above, which considers signal-to-noise ratio, image uniformity, and artifact area, assigning each a different weighting coefficient (w). 01 w 02 w 03 The above comprehensive evaluation method can more fully reflect the overall image quality and avoid the one-sidedness of evaluating a single parameter. At the same time, the artifact area part of the formula also introduces a preset artifact area threshold (A). y This method is used to relatively evaluate the area of ​​artifacts, further enhancing the accuracy and practicality of the evaluation. Furthermore, the calculated quality evaluation parameters are compared with preset quality evaluation parameter thresholds to determine whether the image meets quality requirements. This determination method is objective, reducing the subjectivity and uncertainty of human judgment. When the quality evaluation parameters are lower than the preset quality evaluation parameter thresholds, the image is determined to be unqualified. This helps to promptly identify and process substandard medical images, ensuring that images used for the detection of neurodevelopmental diseases have sufficient accuracy and reliability. The above technical solution performs quality judgment on images that have already undergone denoising, thus indirectly verifying the effectiveness of the denoising process. If the denoised image still does not meet quality requirements, further optimization of the denoising algorithm or parameters may be necessary.

[0155] In summary, this technical solution effectively controls and evaluates the quality of medical images used in the detection of neurodevelopmental diseases by extracting key image parameters, comprehensively calculating quality evaluation parameters, and objectively determining image quality. This not only helps improve the accuracy and reliability of medical diagnosis but also provides strong support for the development of medical image processing technology. Furthermore, by comprehensively utilizing techniques such as medical image retrieval, image quality assessment, and secondary acquisition control, the above-mentioned technical solution effectively improves the quality of medical images used for the detection of neurodevelopmental diseases, providing doctors with more reliable diagnostic evidence, while also optimizing workflows and improving patient satisfaction.

[0156] In one embodiment of the present invention, the parameter adjustment terminal includes:

[0157] The data quality assessment module is used to determine whether the image quality of the medical image after noise reduction meets the image quality requirements after the image is re-acquired.

[0158] The second quality evaluation parameter retrieval module is used to extract the quality evaluation parameters corresponding to the image quality of the re-acquired medical image after noise reduction when the image quality of the re-acquired medical image after noise reduction still does not meet the image quality requirements, and use them as the second quality evaluation parameters.

[0159] The first quality evaluation parameter retrieval module is used to retrieve the quality evaluation parameters corresponding to the first image acquisition for the medical image that has been re-acquired, and use them as the first quality evaluation parameters.

[0160] The parameter adjustment execution module is used to adjust the unified image acquisition parameters using the first quality evaluation parameter and the second quality evaluation parameter;

[0161] The steps for adjusting the unified image acquisition parameters in the parameter adjustment execution module include:

[0162] Retrieve the first quality evaluation parameter and the second quality evaluation parameter, and use the first quality evaluation parameter and the second quality evaluation parameter to obtain the difference between the first quality evaluation parameter and the second quality evaluation parameter, as the first difference data;

[0163] The difference between the first quality evaluation parameter and the preset quality evaluation parameter threshold is retrieved as the second difference data.

[0164] The difference between the second quality evaluation parameter and the preset quality evaluation parameter threshold is retrieved as the third difference data.

[0165] The compensation amounts corresponding to uniform resolution, uniform contrast, and uniform exposure time are obtained using the first difference data, the second difference data, and the third difference data; wherein, the compensation amounts corresponding to uniform resolution, uniform contrast, and uniform exposure time are obtained by the following formula:

[0166]

[0167] Among them, Y R Y C and Y T These represent the compensation amounts corresponding to uniform resolution, uniform contrast, and uniform exposure time, respectively; f(D) represents the nonlinear variation function; k 01 k 02 and k 03 These represent the compensation adjustment coefficients corresponding to uniform resolution, uniform contrast, and uniform exposure time, respectively; ΔC 01 ΔC 02 and ΔC 03 These represent the first difference data, the second difference data, and the third difference data, respectively.

[0168] The unified resolution, unified contrast, and unified exposure time are adjusted using the compensation amounts corresponding to the unified resolution, unified contrast, and unified exposure time, respectively. The adjusted unified resolution, unified contrast, and unified exposure time are obtained using the following formulas:

[0169]

[0170] Among them, R z Indicates the adjusted uniform resolution; C z Indicates the adjusted uniform contrast; T z Indicates the adjusted uniform exposure time; Y R Y C and Y T These represent the compensation amounts corresponding to uniform resolution, uniform contrast, and uniform exposure time, respectively; R t Indicates uniform resolution; C t Indicates uniform contrast; T t Indicates uniform exposure time; λ R , λ C and λ T These represent the adjustment ratios corresponding to uniform resolution, uniform contrast, and uniform exposure time, respectively.

[0171] The working principle of the above technical solution is as follows: The data quality judgment module first evaluates whether the quality of the medical image after re-acquisition and noise reduction meets the standards. This is achieved by comparing the image's quality evaluation parameters with preset quality evaluation parameter thresholds. If the image quality still does not meet the requirements, the second quality evaluation parameter retrieval module will extract the quality evaluation parameters of the current image as the second quality evaluation parameters.

[0172] Simultaneously, the first quality evaluation parameter retrieval module retrieves the quality evaluation parameters corresponding to the image before re-acquisition (i.e., during the first acquisition) as the first quality evaluation parameter. The parameter adjustment execution module then calculates the difference between the first and second quality evaluation parameters (first difference data), and the differences between each of them and the preset quality evaluation parameter thresholds (second and third difference data). Using the aforementioned difference data and the nonlinear change function f(D), combined with the compensation adjustment coefficient (k... 01 k 02 k 03 ), calculate the compensation amount (Y) corresponding to uniform resolution, uniform contrast, and uniform exposure time. R Y C Y T ).

[0173] Finally, based on the calculated compensation amount and the original unified image acquisition parameters (R... t C t T t By adjusting the proportional coefficient (λ) R , λ C , λ T Adjustments were made to the uniform resolution, uniform contrast, and uniform exposure time to obtain the adjusted parameter values ​​(R). z C z T z ).

[0174] The effects of the above technical solution are as follows: By dynamically adjusting unified image acquisition parameters, this solution can specifically optimize images that do not meet quality standards, improving the overall quality of medical images and providing doctors with more accurate diagnostic information. High-quality medical images can reduce misdiagnosis and missed diagnosis, improve the diagnostic accuracy of diseases such as neurodevelopmental disorders, and thus improve patient treatment outcomes. This technical solution is adaptable to different types and performance levels of medical equipment, ensuring the stability and consistency of image quality through flexible parameter adjustment strategies. Automated parameter adjustment processes reduce the need for manual intervention, improve work efficiency, and reduce the possibility of human error. Through unified image acquisition parameter standards and quality evaluation processes, this technical solution can promote the standardization and normalization of medical image acquisition and processing.

[0175] On the other hand, the aforementioned technical solution first retrieves the first and second quality evaluation parameters and calculates the difference between them (first difference data), as well as the difference between each parameter and a preset quality evaluation parameter threshold (second and third difference data), providing a precise data foundation for subsequent parameter adjustments. Simultaneously, using the first, second, and third difference data, combined with the nonlinear change function f(D) and the compensation adjustment coefficient (k... 01 k 02 k 03 ), calculate the compensation amount (Y) corresponding to uniform resolution, uniform contrast, and uniform exposure time. R Y C Y T This calculation method considers the relative differences between parameters and achieves more flexible adjustment of the compensation amount through nonlinear functions, which helps to more accurately adapt to the needs of different image quality. Simultaneously, the compensation amount is applied to the original uniform resolution, uniform contrast, and uniform exposure time through a formula to calculate the adjusted parameter value (R). z C z T z This automated and intelligent adjustment method reduces human intervention and improves adjustment efficiency and accuracy. The adjusted uniform resolution, contrast, and exposure time can more effectively adapt to the needs of different image acquisition scenarios, thereby improving the quality and consistency of image acquisition. This is of great significance for improving the accuracy of medical diagnosis and reducing misdiagnosis and missed diagnosis.

[0176] The aforementioned technical solution enhances the robustness of the image acquisition system through precise parameter adjustment and optimization. It ensures image acquisition quality and stability even under varying lighting conditions, equipment status, or patient conditions. This modular approach facilitates integration and expansion with other image acquisition and processing systems. This contributes to the further development of medical imaging technology and improves the quality and efficiency of medical services. By dynamically adjusting unified image acquisition parameters to optimize medical image quality, this solution demonstrates significant technical effectiveness and broad application prospects.

[0177] In summary, this technical solution achieves precise adjustment and optimization of image acquisition parameters through accurate calculation of compensation amounts, automated parameter adjustment, and optimization of image acquisition quality. This not only improves the accuracy and reliability of medical imaging technology but also provides strong support for the further development and application of medical imaging.

[0178] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if the above modifications and variations of this invention fall within the scope of the claims of this invention and their equivalents, this invention also intends to include the above modifications and variations.

Claims

1. A system for detecting a neurodevelopmental disorder based on image recognition, the system comprising: The image recognition-based neurodevelopmental disease detection system includes a data collection terminal, an image quality assessment terminal, a parameter adjustment terminal, and a deep learning detection terminal. The data collection terminal and the image quality assessment terminal are connected via data communication. The image quality assessment terminal and the parameter adjustment terminal are connected via data communication. The parameter adjustment data output terminal of the parameter adjustment terminal is connected via data communication with the parameter adjustment data input terminal of the data collection terminal. The medical image signal output terminal of the image quality assessment terminal is connected via data communication with the image signal input terminal of the deep learning detection terminal. The deep learning detection terminal is used to receive medical images that meet image quality requirements in real time and input these images into a pre-constructed and trained deep learning model for neurodevelopmental disease detection to obtain detection results. The parameter adjustment terminal is used to determine whether the image quality of the medical image after noise reduction meets the image quality requirements, and to adjust the unified image acquisition parameters of the medical device if the image quality after noise reduction still does not meet the image quality requirements. The parameter adjustment terminal includes: The data quality assessment module is used to determine whether the image quality of the medical image after noise reduction meets the image quality requirements after the image is re-acquired. The second quality evaluation parameter retrieval module is used to extract the quality evaluation parameters corresponding to the image quality of the re-acquired medical image after noise reduction when the image quality of the re-acquired medical image after noise reduction still does not meet the image quality requirements, and use them as the second quality evaluation parameters. The first quality evaluation parameter retrieval module is used to retrieve the quality evaluation parameters corresponding to the first image acquisition for the medical image that has been re-acquired, and use them as the first quality evaluation parameters. The parameter adjustment execution module is used to adjust the unified image acquisition parameters using the first quality evaluation parameter and the second quality evaluation parameter; The steps for adjusting the unified image acquisition parameters in the parameter adjustment execution module include: Retrieve the first quality evaluation parameter and the second quality evaluation parameter, and use the first quality evaluation parameter and the second quality evaluation parameter to obtain the difference between the first quality evaluation parameter and the second quality evaluation parameter, as the first difference data; The difference between the first quality evaluation parameter and the preset quality evaluation parameter threshold is retrieved as the second difference data; The difference between the second quality evaluation parameter and the preset quality evaluation parameter threshold is retrieved as the third difference data; The compensation amounts corresponding to uniform resolution, uniform contrast, and uniform exposure time are obtained using the first difference data, the second difference data, and the third difference data. The uniform resolution, uniform contrast, and uniform exposure time are adjusted using the compensation amounts corresponding to the uniform resolution, uniform contrast, and uniform exposure time, respectively.

2. The image recognition-based neurodevelopmental disorder detection system of claim 1, wherein, The data collection terminal is used to acquire medical images for the detection of neurodevelopmental diseases in real time, and to perform image preprocessing on the medical images to obtain preprocessed medical images. The image quality assessment terminal is used to assess the image quality of the preprocessed medical image. If the preprocessed medical image does not meet the image quality requirements, the image is re-acquired. If the medical image meets the image quality requirements, the medical image that meets the image quality requirements is input to the deep learning detection terminal. The parameter adjustment terminal is used to determine whether the image quality of the medical image after noise reduction meets the image quality requirements, and to adjust the unified image acquisition parameters of the medical device if the image quality after noise reduction still does not meet the image quality requirements. 3.The image recognition-based neurodevelopmental disorder detection system of claim 1, wherein The data collection terminal includes: The unified image acquisition parameter setting module is used to retrieve the original image acquisition parameters of medical images generated by medical devices for the detection of neurodevelopmental diseases, and set the unified image acquisition parameters according to the original image acquisition parameters; The acquisition parameter adjustment terminal is used to adjust the original image acquisition parameters of the medical device according to the unified image acquisition parameters, and after the original image acquisition parameters are adjusted, control the medical device to acquire medical images for the detection of neurodevelopmental diseases. 4.The image recognition-based neurodevelopmental disorder detection system of claim 3, wherein The steps for setting the unified image acquisition parameters corresponding to the unified image acquisition parameter setting module include: Extract the raw image acquisition parameters of the medical device, wherein the raw image acquisition parameters include resolution parameters, contrast parameters, and exposure time parameters; Uniform image acquisition parameters are obtained using the raw image acquisition parameters of medical equipment, wherein the uniform image acquisition parameters include uniform resolution, uniform contrast, and uniform exposure time. 5.The image recognition-based neurodevelopmental disorder detection system of claim 4, wherein, The unified resolution is obtained through the following steps: Extract the types of medical devices collected by the data collection terminal; Extract the equipment performance parameters of the medical equipment corresponding to the above-mentioned equipment types, wherein the equipment performance parameters include dynamic range parameters, sensitivity parameters, signal-to-noise ratio parameters, and uniformity parameters; A device characteristic matrix is ​​generated using the device performance parameters, wherein the structure of the device characteristic matrix is ​​as follows: wherein, D represents a device characteristic matrix; F 01 , F 02 , …, F n respectively represent n dynamic range parameters of the medical devices corresponding to the L 01 , L 02 , …, L n respectively represent n sensitivity parameters of the medical devices corresponding to the S 01 , S 02 , …, S n respectively represent n signal-to-noise ratio parameters of the medical devices corresponding to the J 01 , J 02 , …, J n respectively represent n uniformity parameters of the medical devices corresponding to the Utilizing the device characteristic matrix D Obtaining a non-linear change function f D , wherein the f D function structure is as follows:​​ wherein, f D represents a nonlinear change function; n represents the number of rows of the equipment characteristic matrix D , i.e., the number of equipment types of the medical equipment; m represents the number of columns of the equipment characteristic matrix D , i.e., the number of parameter types included in the equipment performance parameters; D ij represents the parameter value corresponding to the i-th row and the j-th column of the equipment characteristic matrix D i j ε represents a minimum constant to ensure that the logarithmic function is non-negative within the domain of definition; α β γ represent a first adjustment parameter, a second adjustment parameter and a third adjustment parameter, and the first adjustment parameter, the second adjustment parameter and the third adjustment parameter are obtained by the following formula:​​​​​​ ， ， wherein, α x+1 , β x+1 and γ x+1 denotes the x+ 1iteration corresponding first adjustment parameter, second adjustment parameter and third adjustment parameter; α x , β x and γ x denotes the x iteration corresponding first adjustment parameter, second adjustment parameter and third adjustment parameter; denotes a loss function comprising the first adjustment parameter, the second adjustment parameter and the third adjustment parameter; η denotes a learning rate; using the non-linear change function f ( D ) obtaining a uniform resolution, wherein the uniform resolution is obtained by the following equation: wherein, R t denotes a uniform resolution; f D denotes a non-linear change function; n denotes a number of rows of the device characteristic matrix D , i.e. a number of device types of the medical devices; R p denotes a resolution average value corresponding to the original image acquisition parameters of the n th medical device; R i denotes a resolution corresponding to the original image acquisition parameters of the i th medical device; R z denotes a resolution median value corresponding to the original image acquisition parameters of the n th medical device.​ 6.The image recognition-based neurodevelopmental disorder detection system of claim 4, wherein The uniform contrast ratio is obtained through the following steps: Extract the contrast parameters corresponding to the original image acquisition parameters of each medical device; Retrieve the current nonlinear change function f ( D The third adjustment parameter corresponding to ) The contrast parameter corresponding to the original image acquisition parameters of each medical device is combined with the current nonlinear change function. f ( D The third adjustment parameter corresponding to the given parameter is used to obtain the uniform contrast ratio, which is obtained by the following formula: in, C t Indicates uniform contrast; n Represents the device characteristic matrix D The number of rows, i.e., the number of different types of medical equipment; C p express n The average contrast value corresponding to the original image acquisition parameters of a medical device; γ This indicates the third adjustment parameter; C i Indicates the first i Contrast ratio corresponding to the original image acquisition parameters of a medical device; C z express n The median contrast value corresponding to the original image acquisition parameters of this type of medical device; C max express n The maximum contrast value corresponding to the original image acquisition parameters of a medical device.

7. The image recognition-based neurodevelopmental disease detection system according to claim 4, characterized in that, The uniform exposure time is obtained through the following steps: Extract the exposure time parameters corresponding to the original image acquisition parameters of each medical device; Retrieve the current nonlinear change function f ( D The corresponding first adjustment parameter and second adjustment parameter; Using the exposure time parameter and current nonlinear change function corresponding to the original image acquisition parameters of each medical device f ( D The uniform exposure time is obtained by using the first and second adjustment parameters corresponding to the given parameters, wherein the uniform exposure time is obtained by the following formula: in, T t Indicates a uniform exposure time; α and β Indicates the first adjustment parameter and the second adjustment parameter; n Represents the device characteristic matrix D The number of rows, i.e., the number of different types of medical equipment; T i Indicates the first i The exposure time corresponding to the original image acquisition parameters of the medical device; f ( D ) represents a nonlinear variation function; T z express n The median exposure time corresponding to the original image acquisition parameters of the medical device; T min express n The minimum exposure time corresponding to the original image acquisition parameters of a certain medical device.

8. The image recognition-based neurodevelopmental disease detection system according to claim 2, characterized in that, The image quality assessment terminal includes: The medical image retrieval module is used to retrieve denoised medical images for the detection of neurodevelopmental diseases. The image quality assessment execution module is used to assess the image quality of the denoised medical image for detecting neurodevelopmental diseases using the image parameters of the denoised medical image for detecting neurodevelopmental diseases, and to obtain the image quality assessment result. The secondary acquisition control module is used to perform secondary acquisition of medical images when the image quality judgment results indicate that the denoised medical images used for the detection of neurodevelopmental diseases do not meet the image quality requirements.

9. The image recognition-based neurodevelopmental disease detection system according to claim 8, characterized in that, The image quality determination steps of the image quality determination execution module include: Image parameters are extracted from denoised medical images used for the detection of neurodevelopmental diseases, wherein the image parameters include signal-to-noise ratio, image uniformity, and artifact area; The signal-to-noise ratio, image uniformity, and artifact area are used to obtain quality assessment parameters for denoised medical images used in the detection of neurodevelopmental diseases; wherein, the quality assessment parameters are obtained by the following formula: in, Q Indicates quality evaluation parameters; w 01 , w 02 and w 03 These represent the weighting coefficients corresponding to signal-to-noise ratio, image uniformity, and artifact area, respectively. S This represents the numerical value corresponding to the signal-to-noise ratio; U The numerical value representing the uniformity of the image; A Indicates the area of ​​the artifact; A y This represents the preset artifact area threshold; The quality evaluation parameters are compared with preset quality evaluation parameter thresholds; When the quality evaluation parameter is lower than the preset quality evaluation parameter threshold, the medical image for detecting neurodevelopmental diseases that has undergone noise reduction is determined to not meet the image quality requirements.

10. The image recognition-based neurodevelopmental disease detection system according to claim 1, characterized in that, in, The compensation amounts corresponding to uniform resolution, uniform contrast, and uniform exposure time are obtained using the following formula: in, Y R , Y C and Y T These represent the compensation amounts corresponding to uniform resolution, uniform contrast, and uniform exposure time, respectively. f ( D ) represents a nonlinear variation function; k 01 , k 02 and k 03 These represent the compensation adjustment coefficients corresponding to uniform resolution, uniform contrast, and uniform exposure time, respectively; Δ C 01 Δ C 02 and Δ C 03 These represent the first difference data, the second difference data, and the third difference data, respectively. The adjusted uniform resolution, uniform contrast, and uniform exposure time are obtained using the following formula: in, R z This indicates the adjusted unified resolution; C z This indicates the adjusted uniform contrast. T z This indicates the adjusted uniform exposure time; Y R , Y C and Y T These represent the compensation amounts corresponding to uniform resolution, uniform contrast, and uniform exposure time, respectively. R t Indicates a uniform resolution; C t Indicates uniform contrast; T t Indicates a uniform exposure time; λ R , λ C and λ T These represent the adjustment ratios corresponding to uniform resolution, uniform contrast, and uniform exposure time, respectively.

Images