A portable DR image lesion analysis method and system

Abstract

The application provides a portable DR image lesion analysis method and system, including: encapsulating performance state information of each imaging unit inside a detector and image data collected by the imaging unit to form an integrated data stream, and transmitting the integrated data stream; receiving the integrated data stream and parsing the performance state information and the image data from the integrated data stream; performing gain adjustment, offset correction and geometric correction processing on the image data according to the parsed performance state information, so as to eliminate local image defects in the image data caused by detector performance drift; and providing the processed image data to an image analysis module for lesion identification. The application can accurately correct the image data before image analysis, and effectively eliminate local image defects caused by detector performance drift.

Description

Technical Field

[0001] This application relates to the field of DR image processing technology, and more specifically, to a portable DR image lesion analysis method and system. Background Technology

[0002] Portable DR equipment, when frequently moved or used in harsh environments (such as dusty or humid conditions) for extended periods, experiences gradual and localized subtle shifts in its internal structure and performance. For example, the scintillator layer, photodiode array, and readout circuit board inside the detector may undergo extremely minor changes in their physical connections or material properties due to prolonged vibration and minute deformation. Furthermore, dust particles or moisture in the environment can cause slight corrosion of the internal circuit boards, affecting the performance of individual imaging units within the detector, such as increased dark current, decreased response sensitivity, or uneven pixel spacing. These changes are so subtle that they are difficult for non-experts to detect during routine image quality checks, such as observing uniform phantom images.

[0003] When DR detectors with subtle, localized performance drift are used to acquire patient images, the resulting digital images carry imperceptible defects such as increased local noise, decreased sensitivity, or non-uniform quality issues like geometric distortion. Existing image analysis systems and their core AI methods lack the ability to precisely perceive the real-time operating status of DR detectors, making it impossible to adaptively correct or compensate for these localized image quality degradations caused by detector performance drift. These systems typically only address obvious image problems, often failing to recognize subtle, localized image quality degradations caused by internal detector performance drift. This results in potentially defective images being deemed acceptable and continued into subsequent lesion analysis processes. Summary of the Invention

[0004] This application provides a portable DR image lesion analysis method and system to solve at least one of the above-mentioned technical problems.

[0005] To achieve the above objectives, this application adopts the following technical solution: In a first aspect, this application discloses a portable DR image lesion analysis method, specifically including the following steps: Acquire performance status information of each imaging unit inside the detector; The performance status information and the image data acquired by the imaging unit are encapsulated to form an integrated data stream, and the integrated data stream is transmitted. Receive the integrated data stream and parse the performance status information and image data from it; Based on the parsed performance status information, the image data is processed by gain adjustment, offset correction and geometric correction to eliminate local image defects caused by detector performance drift. The processed image data is provided to the image analysis module for lesion identification.

[0006] Secondly, this application also discloses a portable DR image lesion analysis system, which includes: The information acquisition module is used to acquire the performance status information of each imaging unit inside the detector. The performance status information includes the dark current deviation, response sensitivity deviation and position offset of the imaging unit. The data encapsulation and transmission module is used to encapsulate performance status information and image data acquired by the imaging unit to form an integrated data stream and transmit the integrated data stream. The data parsing module is used to receive the integrated data stream and parse performance status information and image data from it. The image correction module is used to perform gain adjustment, offset correction and geometric correction on the image data based on the parsed performance status information, so as to eliminate local image defects caused by detector performance drift in the image data. The image analysis module provides processed image data to the image analysis module for lesion identification.

[0007] Compared with the prior art, this application has at least the following beneficial effects: This application effectively solves the technical problem that when existing portable DR devices are used for a long time in complex environments, the image acquisition quality deteriorates due to detector performance drift, which in turn affects the accuracy of lesion analysis. Attached Figure Description

[0008] Figure 1 A flowchart illustrating a portable DR image lesion analysis method provided in this application; Figure 2 This is a schematic diagram of the structure of a portable DR image lesion analysis system provided in this application. Detailed Implementation

[0009] The technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0010] like Figure 1 As shown in the embodiment of this application, a portable DR image lesion analysis method is proposed, including: Acquire the performance status information of each imaging unit inside the detector, including the dark current deviation, response sensitivity deviation, and position offset of the imaging unit. The performance status information and the image data acquired by the imaging unit are encapsulated to form an integrated data stream, and the integrated data stream is transmitted. Receive the integrated data stream and parse the performance status information and image data from it; Based on the parsed performance status information, the image data is processed by gain adjustment, offset correction and geometric correction to eliminate local image defects caused by detector performance drift. The processed image data is provided to the image analysis module for lesion identification.

[0011] The detector usually refers to the digital X-ray detector, which is the core component of a portable DR system and is responsible for converting X-ray signals into digital image data.

[0012] An imaging unit refers to the basic pixels or pixel arrays that make up an image inside the detector, and each imaging unit has independent performance parameters.

[0013] Performance status information is a key parameter describing the operating status of the imaging unit, including dark current deviation, response sensitivity deviation, and position offset. Dark current deviation refers to the difference between the current generated by the imaging unit and the ideal value when there is no X-ray irradiation, which introduces image noise. Response sensitivity deviation refers to the difference between the imaging unit's response to X-ray signals and the ideal value, affecting the brightness uniformity of the image. Position offset refers to the deviation between the actual physical position of the imaging unit and its theoretically designed position, which may lead to geometric distortion of the image.

[0014] Image data refers to the raw digital image information acquired by the imaging unit.

[0015] An integrated data stream refers to a data packet or data stream that encapsulates performance status information and image data together, facilitating synchronous transmission and processing.

[0016] Gain adjustment refers to the correction of brightness or contrast in image data based on the deviation in response sensitivity.

[0017] Offset correction refers to baseline correction of image data based on dark current deviation to eliminate dark current noise.

[0018] Geometric correction refers to spatial transformation of image data based on positional offset to correct image distortion.

[0019] An image analysis module is a software or hardware unit responsible for identifying, classifying, and quantifying lesions in processed image data, and it typically includes artificial intelligence algorithms.

[0020] The implementation environment of this application typically includes a portable DR device (containing detectors and data acquisition units) and a portable computer or mobile workstation (containing data receiving, processing, and image analysis modules), which transmits data via wired or wireless means.

[0021] The core of the portable DR image lesion analysis method proposed in this application lies in the precise perception and adaptive correction of detector performance drift. The main technical features will be described in detail below.

[0022] First, regarding the acquisition of performance status information for each imaging unit within the detector, this step involves obtaining information such as dark current deviation, response sensitivity deviation, and positional offset. This step aims to acquire the health status of each imaging unit within the detector in real-time or near real-time. One implementation involves integrating multiple miniature sensors within the detector. These sensors periodically measure the dark current, response sensitivity, and physical position of each imaging unit. For example, dark current can be measured by placing a photodiode array within the detector, response sensitivity can be measured using a specific calibration light source and a known X-ray dose, and the relative position of the imaging unit can be monitored using microelectromechanical systems (MEMS) sensors or optical encoders. These raw measurements are then processed to calculate the deviations from ideal reference values, thus obtaining the dark current deviation, response sensitivity deviation, and positional offset. Another implementation involves the detector undergoing detailed performance calibration at the factory, generating a set of reference performance parameters. During actual use, the system periodically runs a self-test program, collecting the performance data of the current imaging unit and comparing it with the reference parameters to calculate various deviations. For example, images can be acquired periodically without X-ray irradiation and their noise levels analyzed to assess dark current deviation; or images can be acquired under uniform X-ray irradiation and their brightness uniformity analyzed to assess response sensitivity deviation.

[0023] Secondly, regarding the step of encapsulating performance status information with image data acquired by the imaging unit to form an integrated data stream and transmitting this integrated data stream, this step aims to ensure that performance status information and image data are transmitted synchronously, providing a foundation for subsequent accurate calibration. One implementation is to embed the performance status information as metadata into the header or footer of the image data frame in the detector's data acquisition unit. For example, data such as dark current deviation, response sensitivity deviation, and position offset can be packaged into a small binary data block and then appended to the beginning of each image data frame to form a complete data packet. These data packets are then transmitted to a portable computer via the detector's data interface (such as USB, Ethernet, or PCIe). Another implementation is that performance status information and image data can be generated as independent data packets, but associated with a common timestamp or sequence number. For example, the detector can send a performance status information data packet at regular intervals, containing a timestamp; simultaneously, each image data packet also contains the same timestamp. During transmission, these data packets are transmitted to the receiving end via wired connections (such as USB 3.0 or Gigabit Ethernet) or wireless connections (such as Wi-Fi 6 or 5G).

[0024] Secondly, regarding the step of receiving the integrated data stream and parsing the performance status information and image data from it, this step is the starting point for data processing, ensuring that subsequent corrections can obtain accurate input. One implementation is to design a dedicated data parser within the receiving module of the portable computer. This parser can recognize the encapsulation format of the integrated data stream. For example, if performance status information is embedded in the header of the image data frame, the parser will first read the header of the data frame, extract the performance status information, and then continue reading the remaining part as image data. For example, a data packet structure can be defined, containing a fixed-length header for storing performance status information and a variable-length payload for storing image data. After receiving the complete data packet, the receiving module will parse the header according to a preset protocol to obtain the performance status information and separate the image data.

[0025] Next, regarding the step of performing gain adjustment, offset correction, and geometric correction on the image data based on the resolved performance status information to eliminate local image defects caused by detector performance drift, this is the core correction step of this application. One implementation is that the system performs pixel-by-pixel offset correction on the image data based on the resolved dark current deviation, that is, subtracts the corresponding dark current deviation value from the gray value of each pixel to eliminate dark current noise. Next, based on the response sensitivity deviation, the image data is gain-adjusted, that is, the gray value of each pixel is multiplied by a correction coefficient, which is calculated from the deviation between the ideal response sensitivity and the actual response sensitivity, to compensate for sensitivity unevenness. Finally, based on the positional offset information, the image data is geometrically corrected, for example, by remapping the pixels to the correct spatial position using interpolation algorithms (such as bilinear interpolation or cubic spline interpolation) to correct image distortion. For example, if the dark current deviation of a certain imaging unit is +50, then all pixel values ​​acquired by that imaging unit will be subtracted by 50. If a deviation in response sensitivity causes a 10% reduction in response, the pixel value will be multiplied by 1.1 as compensation.

[0026] Finally, regarding the step of providing the processed image data to the image analysis module for lesion identification, this is the final stage where the corrected image data realizes its value. One implementation involves storing the image data, after gain adjustment, offset correction, and geometric correction, in a standard medical image format (such as DICOM) and loading it into the image analysis module. The image analysis module can be a standalone software application or a component integrated into a portable computer's operating system. Internally, this module runs a pre-trained deep learning model or other image processing algorithms to automatically identify abnormal regions in the image, such as lung nodules, fracture lines, or inflammatory lesions. For example, the processed image data is input into a convolutional neural network (CNN), which, trained on a large number of labeled images, can identify features related to lesions in the image and output the location, size, and probability score of the lesion.

[0027] The portable DR image lesion analysis method proposed in this application achieves adaptive correction of local image defects caused by detector performance drift by tightly integrating the performance status information of each imaging unit within the detector with the image data. Specifically, the method first acquires performance status information such as dark current deviation, response sensitivity deviation, and positional offset of the imaging units. This information reflects subtle performance degradation that may occur in the detector under long-term use or complex environments. Subsequently, this performance status information and the image data acquired by the imaging units are encapsulated into an integrated data stream and transmitted. At the receiving end, the integrated data stream is parsed, separating the performance status information and the original image data. Crucially, the system performs refined gain adjustment, offset correction, and geometric correction processing on the image data based on the parsed performance status information. For example, dark current deviation is used for offset correction to eliminate noise, response sensitivity deviation is used for gain adjustment to compensate for brightness unevenness, and positional offset is used for geometric correction to correct image distortion. Through these correction steps, local image defects caused by detector performance drift in the image data are effectively eliminated, thereby obtaining higher quality and more accurate diagnostic images. Finally, this processed image data is provided to the image analysis module for lesion identification.

[0028] This application uses the performance status information of each imaging unit within the detector as a key input, and integrates it with image data for unified encapsulation and transmission. This design enables the system to acquire the most detailed and nuanced operating status information of the detector in real-time or near real-time. By analyzing and utilizing this performance status information (including dark current deviation, response sensitivity deviation, and positional offset), this application can perform highly customized and adaptive gain adjustment, offset correction, and geometric correction processing on the image data. This correction mechanism based on the detector's own state can accurately eliminate local image defects caused by detector performance drift, which is not available in existing technologies. For example, when the dark current of a certain imaging unit increases due to aging, this application can perform precise offset correction based on its specific dark current deviation, rather than simply applying a global noise suppression algorithm. This refined correction significantly improves image quality and uniformity, enabling the image analysis module to receive cleaner and more reliable image data for lesion identification. Therefore, this application not only improves the accuracy and reliability of portable DR image lesion analysis but also enhances the system's robustness in complex and non-ideal environments, providing primary healthcare personnel with a more reliable diagnostic aid.

[0029] In some embodiments, the step of providing the processed image data to the image analysis module for lesion identification includes: Acquire sensor readings from the imaging unit; Analyze the dynamic patterns of sensor readings to identify abnormal fluctuations that are inconsistent with the drift of the detector's own performance; Electromagnetic interference feature maps are generated by quantifying the intensity, duration, and affected pixel regions of abnormal fluctuations. Electromagnetic interference feature maps, image data, and performance status information are encapsulated to form an integrated data stream; Transmitting integrated data streams; It receives the integrated data stream and extracts electromagnetic interference feature maps, image data, and performance status information from the integrated data stream; Based on the parsed performance status information, the image data is initially corrected; The preliminarily corrected image data and electromagnetic interference feature maps are provided to the image analysis module for lesion identification.

[0030] Specifically, acquiring sensor readings from the imaging unit refers to the real-time or periodic collection of non-image data reflecting the detector's operating environment and its own state through various environmental sensors integrated inside or near the DR detector, such as electromagnetic field sensors, temperature sensors, or humidity sensors. These sensor readings can provide clues about potential interference sources. Analyzing the dynamic patterns of sensor readings and identifying anomalous fluctuations inconsistent with the detector's own performance drift can be understood as performing time-series analysis and pattern recognition on the acquired sensor readings. By establishing baseline patterns of sensor readings under normal operating conditions and a detector performance drift model, any sharp or persistent fluctuations that deviate from these preset patterns and cannot be explained by performance drift can be identified as anomalous fluctuations, which are usually related to external electromagnetic interference events. For example, statistical methods (such as moving averages or standard deviation analysis) or machine learning algorithms (such as anomaly detection models) can be used to identify these anomalies.

[0031] Generating an electromagnetic interference (EMI) feature map by quantifying the intensity, duration, and affected pixel regions of abnormal fluctuations involves transforming identified abnormal fluctuations into a structured data representation. Intensity can refer to the peak or average value of the electromagnetic field strength, duration refers to the length of time the interference event lasts, and affected pixel regions can be determined by analyzing specific noise patterns in the image data that occur synchronously with the abnormal fluctuations. The EMI feature map can be a multi-dimensional vector or matrix containing key information such as the type, location, intensity, and frequency of the interference, aiming to provide detailed contextual information about the interference source for subsequent image analysis. Furthermore, encapsulating the EMI feature map with image data and performance status information to form an integrated data stream means integrating all relevant information into a unified data packet. This encapsulation can employ specific data structures or protocols to ensure the integrity and relevance of the data during transmission. Transmitting the integrated data stream typically involves wired or wireless communication interfaces, such as USB, Ethernet, or Wi-Fi, sending data from the detector to the processing end.

[0032] At the receiving end, after receiving the integrated data stream, it is necessary to extract the electromagnetic interference (EMI) feature map, image data, and performance status information from it. This typically involves decapsulating and parsing the data stream to recover the original information. Based on the extracted performance status information, preliminary correction is performed on the image data. This means that before providing the image data to the image analysis module, basic gain adjustment, offset correction, and geometric correction are applied using the detector performance status information to eliminate local image defects caused by detector performance drift. This preliminary correction aims to address inherent, predictable performance issues of the detector, laying the foundation for subsequent, more refined analysis. Finally, the pre-corrected image data and EMI feature map are provided to the image analysis module for lesion identification. This means that when performing lesion identification, the image analysis module not only receives the pre-corrected image data but also obtains a detailed feature map of potential EMI. The image analysis module can utilize this additional interference information to weight, filter, or compensate specific regions or patterns in the image when identifying lesions, thereby improving the accuracy and robustness of lesion identification.

[0033] This application effectively addresses the problem of electromagnetic interference (EMI) potentially misleading lesion identification by introducing a detection, quantification, and characterization process for EMI before providing image data to the image analysis module. Specifically, by acquiring sensor readings from the imaging unit and analyzing their dynamic patterns, abnormal fluctuations unrelated to detector performance drift can be identified, thereby accurately locating and identifying EMI events. Furthermore, by quantifying the intensity, duration, and affected pixel regions of these abnormal fluctuations, a detailed EMI feature map is generated. This map, as important auxiliary information, is provided to the image analysis module along with the pre-corrected image data. When identifying lesions, the image analysis module can combine the EMI feature map to perform special processing on potentially affected areas of the image, such as reducing their weight, performing targeted denoising, or correction, thereby avoiding misidentifying artifacts caused by EMI as lesions or preventing EMI from obscuring true lesion information. Therefore, this solution enables the image analysis module to identify lesions more intelligently and accurately, improving diagnostic reliability.

[0034] Suppose that during a portable DR examination, a working medical device is present near the detector, generating periodic electromagnetic radiation. At this time, the imaging unit's electromagnetic field sensor acquires periodic high-intensity readings that deviate from the normal baseline pattern. The system analyzes the dynamic patterns of these sensor readings, identifying these anomalous fluctuations unrelated to detector performance drift and classifying them as electromagnetic interference (EMI). The system then quantifies the intensity of these anomalous fluctuations (e.g., electromagnetic field intensity peaks reaching a certain threshold), duration (e.g., each interference lasting 100 milliseconds), and the affected pixel regions determined through image analysis (e.g., striped noise appearing in the upper left corner of the image). Based on this information, the system generates an EMI feature map, which includes the periodicity, intensity, frequency, approximate location, and affected area of ​​the interference within the image. After image data acquisition, the image data is first preliminarily corrected based on detector performance information to eliminate inherent defects such as dark current deviation, response sensitivity deviation, and positional offset. Subsequently, the preliminarily corrected image data and the generated EMI feature map are packaged together and transmitted to the image analysis module. After receiving this data, the image analysis module uses electromagnetic interference feature maps to guide its lesion identification algorithm. For example, if the feature map indicates that a certain area of ​​the image has been subjected to a specific type of electromagnetic interference, the image analysis algorithm can perform special processing on the pixel values ​​of that area, such as applying a specific denoising filter, or reducing the confidence level of that area when identifying lesions, to avoid misclassifying interference-induced stripes or spots as tiny lesions. Conversely, if a certain area is not affected by interference, lesion identification can be performed with greater confidence. In this way, even in environments with electromagnetic interference, the method of this application can ensure the accuracy of lesion identification and avoid misdiagnosis or missed diagnosis caused by external interference.

[0035] In some embodiments, the step of obtaining the performance status information of each imaging unit inside the detector, including the dark current deviation, response sensitivity deviation, and position offset of the imaging unit, includes: The steps for acquiring performance status information of each imaging unit inside the detector, including dark current deviation, response sensitivity deviation, and position offset of the imaging unit, include: Acquire sensor readings of the imaging unit, including raw measurements of the imaging unit's dark current, response sensitivity, and position offset; Track and analyze historical readings of the sensor to establish long-term performance drift trends; Identify systematic biases in sensor readings that do not conform to long-term performance drift trends, and these systematic biases are related to environmental changes or usage duration. Based on systematic bias, it is determined that the sensor itself has measurement errors due to performance drift or degradation; Based on the preset sensor calibration parameters, the sensor readings are compensated to correct the sensor's own measurement error; The compensated sensor readings are used as performance status information, which includes dark current deviation, response sensitivity deviation, and position offset of the imaging unit.

[0036] Specifically, when acquiring performance status information of each imaging unit within the detector, it is first necessary to obtain sensor readings from the imaging units. Sensor readings refer to raw data collected in real time or periodically by the sensors integrated within the imaging unit. These readings include raw measurements of the imaging unit's dark current, response sensitivity, and positional offset under its current operating condition. These raw measurements are the fundamental data for evaluating the imaging unit's performance. Based on this, by continuously tracking and deeply analyzing the historical sensor readings, a long-term performance drift trend of the sensors can be established. This trend reflects the performance changes of the imaging unit under long-term operation or different environmental conditions, providing a benchmark for subsequent deviation identification.

[0037] The current sensor readings are compared with established long-term performance drift trends to identify systematic deviations that do not conform to these trends. Systematic deviations typically manifest as persistent or periodic variations and may be closely related to changes in the external environment (such as temperature, humidity, or electromagnetic interference) or the cumulative usage time of the detector. Once such systematic deviations are identified, it can be determined that the sensor itself may have measurement errors due to performance drift or degradation. This error is not random noise but a systematic deviation caused by sensor hardware aging, calibration parameter failure, or environmental influences. To eliminate or reduce this measurement error, the sensor readings need to be compensated according to preset sensor calibration parameters. These calibration parameters can be set at the factory or updated based on periodic calibration results; their purpose is to adjust the sensor's raw readings to a more accurate level to correct the sensor's own measurement errors. Finally, the compensated sensor readings are used as performance status information. This performance status information accurately reflects the dark current deviation, response sensitivity deviation, and positional offset of the imaging unit, providing a reliable basis for subsequent image correction.

[0038] This application ensures the accuracy and reliability of acquired performance status information by refining the sensor readings of the imaging unit. Specifically, by tracking and analyzing historical sensor readings, a long-term performance drift trend can be established, thereby distinguishing between systematic deviations caused by sensor performance drift or degradation and normal performance fluctuations. Based on the identification and determination of systematic deviations, the original sensor readings can be compensated in a targeted manner using preset sensor calibration parameters, effectively correcting the sensor's own measurement errors. As a result, the obtained performance status information, such as dark current deviation, response sensitivity deviation, and position offset, more realistically reflects the actual working state of the imaging unit, providing more accurate correction parameters for subsequent gain adjustment, offset correction, and geometric correction of image data, thus more effectively eliminating local image defects caused by detector performance drift.

[0039] In some embodiments, the steps of performing gain adjustment, offset correction, and geometric correction on the image data based on the parsed performance status information to eliminate local image defects caused by detector performance drift in the image data further include: Content encoding is performed on the performance status information and the image data acquired by the imaging unit to obtain the encoded performance status information and the encoded image data. The encoded performance status information and the encoded image data are encapsulated to form an integrated encoded data stream; Generate unique identification information related to the integrated coding data stream. The unique identification information can reflect the integrity and origin of the integrated coding data stream. The unique identification information and the encoded integrated data stream are encapsulated to form a secure integrated data stream; Transmit secure integrated data stream; Upon receiving the secure integrated data stream, the integrity of the secure integrated data stream is verified to confirm that the secure integrated data stream has not been tampered with during transmission. After the integrity verification is passed, the secure integrated data stream is decoded to recover performance status information and image data; Based on the recovered performance status information, the image data is processed by gain adjustment, offset correction and geometric correction to eliminate local image defects caused by detector performance drift.

[0040] Specifically, content encoding refers to the preprocessing of raw performance status information and image data to enhance data security and transmission efficiency. This can include operations such as data compression, encryption, or adding error detection codes. For example, Advanced Encryption Standard (AES) can be used to encrypt the data to prevent unauthorized access, while lossless compression algorithms such as Huffman coding can be used to reduce the data volume. The encoded performance status information and encoded image data are then encapsulated to form a coherent data stream, which is a collection of the original data after security and efficiency processing. Unique identification information can be understood as a fingerprint or signature of the data, and its purpose is to provide a mechanism to verify whether the data has been tampered with during transmission and the origin of the data. This is usually achieved by calculating a hash value (e.g., SHA-256) or generating a digital signature on the coherent data stream. This unique identification information is encapsulated together with the coherent data stream to form a secure coherent data stream, which is a complete packet containing the data itself and its security verification information. At the receiving end, integrity verification refers to recalculating the hash value of the received coherent data stream and comparing it with the received unique identification information to confirm that the data has not been altered during transmission. Only when the verification is successful will the secure integrated data stream be decoded to restore the encoded data to the original performance status information and image data, thereby ensuring the accuracy of subsequent image correction.

[0041] This application encodes performance status information and image data before data transmission, generating unique identifiers reflecting data integrity and origin. This information is then encapsulated into a secure, integrated data stream for transmission. At the receiving end, the received secure, integrated data stream undergoes rigorous integrity verification. This verification process effectively detects whether the data has been tampered with, damaged, or lost during transmission. Only when the data is confirmed to be complete and untampered is content decoding performed to recover the original performance status information and image data. This ensures that the data used for subsequent gain adjustment, offset correction, and geometric correction processing is authentic, reliable, and uncontaminated, fundamentally eliminating local image defects and potential diagnostic errors caused by insecure data transmission.

[0042] The following is a specific example to illustrate this.

[0043] Suppose that in a portable DR device, the imaging unit acquires image data and generates corresponding performance status information. To ensure the security of this data when transmitted to a portable computer for analysis, the image data and performance status information are first content-encoded. For example, the data can be encrypted using the AES-256 encryption algorithm and combined with the LZW algorithm for lossless compression to form encoded performance status information and encoded image data. Then, this encoded data is encapsulated into a unified encoded data stream. Next, a SHA-256 hash value is calculated for this unified encoded data stream as its unique identifier, which uniquely represents the content of the data stream. Then, this SHA-256 hash value is encapsulated together with the unified encoded data stream to form a secure unified data stream. This secure unified data stream is transmitted to the portable computer via a wireless network. After receiving the secure unified data stream, the portable computer first extracts the unified encoded data stream and independently calculates its SHA-256 hash value. Then, the calculated hash value is compared with the received unique identifier (i.e., the original SHA-256 hash value). If both are completely consistent, it indicates that the data has not been tampered with or damaged during transmission, and the integrity verification is successful. At this point, the portable computer will decode the encoded integrated data stream to recover the original performance status information and image data. Finally, based on this recovered and verified performance status information, precise gain adjustment, offset correction, and geometric correction processing are performed on the image data to eliminate local image defects caused by detector performance drift, and the processed image is provided to the image analysis module for lesion identification.

[0044] In some embodiments, the steps of encapsulating performance status information with image data acquired by the imaging unit to form an integrated data stream and transmitting the integrated data stream include: The image data acquired by the imaging unit is divided into multiple image data blocks; The performance status information is divided into multiple performance status information blocks based on the image data blocks; A data packet is generated for each image data block and its corresponding performance status information block. The data packet contains a data packet header, which includes the data packet sequence number, the data block type identifier, and verification information. Data packets are prioritized, with performance status information packets marked as high priority and image data packets marked as normal priority. Data packets are transmitted to the portable computer via the detector's data interface in order of priority.

[0045] Specifically, the image data acquired by the imaging unit is usually quite large. It is divided into multiple image data blocks, for example, by rows, columns, or specific regions of the image. This facilitates batch processing and transmission of data, improving transmission flexibility and efficiency. The size of the image data blocks can be dynamically adjusted based on actual transmission bandwidth, processing power, and real-time requirements.

[0046] Performance status information, including dark current deviation, response sensitivity deviation, and positional offset of the imaging unit, is crucial data for image correction. This performance status information is segmented according to image data blocks, forming multiple performance status information blocks. This ensures that each image data block can be associated with its corresponding real-time performance status information, thus providing a basis for subsequent accurate correction. For example, if the image data is divided into N blocks, the performance status information is also correspondingly segmented into N blocks or performance status information blocks corresponding to the N image data blocks.

[0047] A data packet is generated for each image data block and its corresponding performance status information block. Each data packet contains a header, which includes the packet's sequence number, the data block's type identifier, and checksum information. The sequence number is used for data reassembly and order recovery at the receiving end; the data block's type identifier distinguishes whether the data packet is an image data packet or a performance status information data packet; and the checksum information is used to detect errors or corruption during transmission, ensuring data integrity and reliability. For example, the checksum information can take the form of a Cyclic Redundancy Check (CRC) code or a hash value.

[0048] Data packets are prioritized. Specifically, performance status information packets are marked as high priority, while image packets are marked as regular priority. This is to ensure that critical performance status information is transmitted and processed first, as this information is essential for subsequent image correction, and its timeliness directly affects the accuracy of the correction. For example, in situations of network congestion or bandwidth limitations, high-priority packets will receive priority transmission.

[0049] Data packets are transmitted to a portable computer via the detector's data interface, following a priority order. The detector's data interface can be USB, Ethernet, PCIe, or a wireless communication interface, the choice depending on the specific hardware configuration and transmission requirements. Priority transmission means that high-priority data packets (i.e., performance status information packets) are sent before regular-priority packets (i.e., image packets), thus ensuring the real-time performance and effectiveness of critical calibration parameters.

[0050] This application effectively solves the efficiency and reliability problems that may be faced by single data stream transmission by refining the integrated data stream into independent data packets with specific identifiers and priorities. Specifically, image data is divided into multiple image data blocks, and performance status information blocks are correspondingly segmented, allowing data transmission to be performed in parallel or in batches, improving the flexibility and efficiency of transmission. Data packets with sequence numbers, type identifiers, and checksums are generated for each data block, enabling the receiving end to accurately identify data types, restore data order, and verify data integrity, thereby avoiding the impact of data corruption or misinterpretation on subsequent processing. Crucially, by assigning high priority to performance status information data packets, it ensures that performance status information critical to image correction is transmitted and processed first, even with limited transmission resources, avoiding correction errors caused by delays or loss of critical information. This mechanism of segmentation, encapsulation, and prioritized transmission makes the data transmission process more robust and efficient, laying the foundation for subsequent accurate image correction.

[0051] Suppose a portable DR detector acquires a chest X-ray image with a data size of 20MB. Simultaneously, sensors inside the detector acquire performance status information such as dark current deviation, response sensitivity deviation, and positional offset of the current imaging unit; this information is approximately 10KB.

[0052] First, the image data was divided into 200 image data blocks, each approximately 100KB in size. Simultaneously, performance status information was also divided into 200 performance status information blocks corresponding to these image data blocks, each containing correction parameters related to the corresponding image region.

[0053] Next, the system generates a separate data packet for each image data block and each performance status information block. Each data packet has a header that includes the packet's sequence number (e.g., from 1 to 400), the data block type identifier (e.g., 0 for image data, 1 for performance status information), and a CRC checksum.

[0054] The system then prioritizes these data packets. All packets containing performance status information are marked as high priority, while all packets containing image data are marked as normal priority.

[0055] Finally, the detector transmits these data packets to the connected portable computer via its USB data interface, prioritizing them. For example, if a brief congestion occurs on the USB bus during transmission, the system will prioritize sending high-priority performance status information data packets to ensure the portable computer can obtain the latest correction parameters in a timely manner. Even if there is a slight delay in the transmission of image data packets, it will not affect the timely start and accuracy of the correction algorithm. After receiving these data packets, the portable computer will reassemble, verify, and decode them based on the information in the packet header, ultimately obtaining complete and corrected image data, which will then be provided to the image analysis module for lesion identification.

[0056] In some embodiments, the steps of encapsulating performance status information with image data acquired by the imaging unit to form an integrated data stream and transmitting the integrated data stream include: The performance status information and image data acquired by the imaging unit are processed in blocks, and each data block is redundantly encoded. The encoded performance status information data block and the encoded image data block are encapsulated to form multiple data packets of an integrated data stream; Obtain a health status map of the storage medium; Based on the health status map of the storage medium, select a healthy storage area and write multiple data packets of the integrated data stream into the healthy storage area; During the writing of multiple data packets into the integrated data stream, monitor the success rate of the write operation and data integrity; When a write error is detected, the damaged data packet is rewritten to the backup storage area, and the health status map of the storage medium is updated. After multiple data packets of the integrated data stream are successfully written, the integrated data stream in the storage medium is verified to confirm the integrity of the data. Transmitted integrated data stream that has passed verification.

[0057] Specifically, before encapsulating the performance status information and image data acquired by the imaging unit, this data is first processed by segmenting it into several smaller data blocks. This helps improve the flexibility and efficiency of data processing. Subsequently, each data block is redundantly encoded, for example using error correction coding techniques such as Cyclic Redundancy Check (CRC), Hamming codes, or Reed-Solomon codes, to detect and recover from errors that occur during data transmission or storage, thereby enhancing the data's fault tolerance.

[0058] A storage media health map can be understood as a mapping table that real-time or periodically assesses and records the performance, reliability, remaining lifespan, and potential failure risks of various storage areas of a storage device (such as a solid-state drive or flash memory card). This map dynamically reflects the availability and stability of the storage media, aiming to provide a reliable basis for selecting storage areas for data writing.

[0059] Based on the health status map of the storage media, the system intelligently selects storage areas that are currently in good condition, have stable performance, and show no obvious signs of failure for writing multiple data packets to the integrated data stream. During the data packet writing process, the system continuously monitors the success rate and data integrity of the write operation, for example, by comparing the consistency between the written data and the source data and checking the status codes returned by the write operation.

[0060] When a write error is detected, such as a data packet failing to be written successfully or a data verification failure after writing, the system will rewrite the damaged data packet to a pre-defined backup storage area. Simultaneously, the health status map of the storage medium will be immediately updated to mark the area where the write error occurred, preventing the area from being used again, and recording the usage of the backup area.

[0061] After multiple data packets of the integrated data stream are successfully written, a comprehensive verification of the integrated data stream in the storage medium is performed, such as by calculating the hash value of the data or performing decoding verification of redundant encoding, to ultimately confirm the integrity and accuracy of all data packets.

[0062] This application enhances the fault tolerance of data by introducing block processing and redundant coding before data encapsulation and transmission. Due to the redundant coding of data blocks, even if partial data corruption occurs during transmission or storage, errors can be detected and even recovered through the encoded information. Furthermore, by acquiring and utilizing the health status map of the storage medium, the system can intelligently select healthy storage areas for data writing, thereby reducing the risk of data storage failure from the source. During the writing process, the success rate and data integrity of the write operation are monitored in real time. Once a write error is detected, the damaged data packet is immediately rewritten to a backup storage area, and the health status map is updated promptly, forming a closed-loop error handling mechanism. Finally, by verifying the successfully written data stream, the final integrity of the data is further ensured, effectively solving the integrity problems that may occur when data is stored and transmitted in unstable environments in traditional solutions.

[0063] In some embodiments, the steps of performing gain adjustment, offset correction, and geometric correction on the image data based on the parsed performance status information to eliminate local image defects caused by detector performance drift include: Obtain the instantaneous interference signal strength of the detector's operating environment; Compare the instantaneous interference signal strength with the preset interference threshold; When the instantaneous interference signal strength exceeds the interference threshold, the parsed performance status information is corrected in real time. The correction includes weighted adjustment or range limitation of dark current deviation, response sensitivity deviation and position offset. Based on the corrected performance status information, the image data is subjected to gain adjustment, offset correction, and geometric correction.

[0064] Specifically, before performing gain adjustment, offset correction, and geometric correction on the image data, the instantaneous interference signal strength of the detector's operating environment is first acquired. Instantaneous interference signal strength can be understood as the electromagnetic interference (EMI) or noise level experienced by the detector in the current operating environment. It can be monitored in real time by a dedicated environmental sensor integrated inside the detector, or estimated by analyzing noise data from non-imaging areas. The purpose is to quantify the potential impact of the current environment on detector performance.

[0065] The acquired instantaneous interference signal strength is compared with a preset interference threshold. The preset interference threshold is empirically determined based on the detector's design specifications, calibration data, and acceptable noise levels in the actual application scenario. This threshold is used to define the level of interference signal that requires correction of performance status information.

[0066] When the instantaneous interference signal strength exceeds a preset interference threshold, the resolved performance status information needs to be corrected in real time. Correction includes weighted adjustment or range limitation of dark current deviation, response sensitivity deviation, and position offset. For example, weighted adjustment refers to applying a weighting factor to the original dark current deviation, response sensitivity deviation, and position offset values ​​based on the strength of the interference signal to compensate for additional errors caused by the interference. Range limitation refers to restricting the correction range of these deviation values ​​to a more conservative range under strong interference environments to avoid over-correction or the introduction of new errors. The purpose is to ensure that, in the presence of external interference, the performance status information used for image correction can more accurately reflect the actual operating state of the detector.

[0067] Based on the corrected performance status information, gain adjustment, offset correction, and geometric correction are performed on the image data. This allows for more effective elimination of local image defects caused by detector performance drift.

[0068] This application addresses the issue of insufficient accuracy in image correction based solely on preset or static performance status information under complex or variable environments by introducing a real-time monitoring and feedback mechanism for the instantaneous interference signal intensity in the detector's operating environment. When the instantaneous interference signal intensity in the detector's environment is high, traditional correction methods may not be able to completely eliminate image artifacts caused by interference, thus affecting the accuracy of lesion identification. This application dynamically corrects the performance status information used for correction, enabling gain adjustment, offset correction, and geometric correction to better adapt to current environmental conditions, thereby more accurately compensating for the detector's own performance drift and the effects of environmental interference.

[0069] Suppose a portable DR system is operating in the field or a temporary medical facility. This area may experience unstable power supplies, radio signals, or electromagnetic radiation from other medical equipment, causing significant fluctuations in the instantaneous interference signal intensity of the detector's operating environment. During image acquisition, the system first uses a built-in electromagnetic interference sensor to acquire the instantaneous interference signal intensity of the current environment in real time. For example, if the detected instantaneous interference signal intensity reaches 50mV / m, while the preset interference threshold is 30mV / m, it indicates significant environmental interference. At this point, the system will correct previously analyzed performance status information such as dark current deviation, response sensitivity deviation, and positional offset in real time. Specifically, the dark current deviation can be weighted and adjusted, for example, by increasing its compensation value by 10% to offset additional noise; simultaneously, the correction range for the response sensitivity deviation is limited to ensure that it does not fluctuate excessively due to interference. Subsequently, the image correction module uses this corrected performance status information to perform gain adjustment, offset correction, and geometric correction on the acquired image data. In this way, even in environments with strong interference, clearer and more accurate DR images can be generated, effectively reducing image artifacts caused by environmental interference, thereby improving doctors' confidence and accuracy in lesion identification.

[0070] In some embodiments, the steps of encapsulating performance status information with image data acquired by the imaging unit to form an integrated data stream and transmitting the integrated data stream include: The performance status information and image data acquired by the imaging unit are processed in blocks. Block processing refers to dividing the original performance status information and image data into several smaller data units to facilitate management, transmission and error recovery.

[0071] Each data block is appended with a sequence number and a timestamp, generating a data packet. The sequence number identifies the order of the data blocks, ensuring correct reordering at the receiving end; the timestamp can be used to determine data freshness or assist in out-of-order detection. A data packet is a transmission unit containing data blocks, sequence numbers, timestamps, and other control information (such as checksums).

[0072] Redundancy coding is used to enhance data fault tolerance. Redundancy coding refers to adding extra redundant information to the original data so that even if some data is lost or damaged during transmission, the receiving end can still recover the original data through the redundant information. For example, forward error correction (FEC) coding technology can be used.

[0073] The encoded data packets are transmitted through a wireless communication module, which can be Bluetooth, Wi-Fi, or a cellular network, depending on the specific application scenario and transmission distance requirements.

[0074] At the receiving end, out-of-order detection and duplicate packet removal are performed on the received data packets. Out-of-order detection determines whether the data packets arrived in the expected order based on their sequence numbers. Duplicate packet removal identifies and discards duplicate data packets that have already been received to avoid data redundancy and processing errors. Based on the sequence numbers and timestamps of the data packets, the data packets are reordered and their integrity is verified. Reordering restores out-of-order data packets to the correct order based on their sequence numbers. Integrity verification uses the verification information (such as CRC codes) contained in the data packets to check whether the data packets have been corrupted during transmission.

[0075] When a data packet loss is detected, a retransmission request is triggered. If the receiving end finds that a data packet with a certain sequence number has not been received for a long time, or that the integrity check has failed, it will send a retransmission request to the sending end, requesting that the lost or corrupt data packet be retransmitted.

[0076] Decoding successfully retransmitted data packets is performed to recover the data blocks. Decoding is the reverse process of redundancy coding and is used to extract the original data blocks from the received (possibly corrected) data packets.

[0077] The recovered data blocks are reassembled to obtain complete performance status information and image data. Reassembly involves reassembling all correctly received and decoded data blocks according to their sequence numbers, thereby recovering the complete original performance status information and image data.

[0078] Finally, the complete performance status information and complete image data are encapsulated to form an integrated data stream, which is then transmitted. Here, encapsulation and transmission refer to the final encapsulation of the performance status information and image data, whose integrity and accuracy are ensured by the aforementioned robust transmission mechanism, according to the integrated data stream format defined in this application, and then transmitting it to the subsequent image correction module or image analysis module for processing.

[0079] This application effectively solves the problems of data loss, out-of-order delivery, and corruption that may occur in wireless transmission by introducing block processing, sequence numbers and timestamps, redundancy coding, and out-of-order detection, reordering, integrity verification, and retransmission mechanisms at the receiving end. Specifically, block processing makes data transmission more flexible, facilitating independent processing and recovery of small data blocks. Sequence numbers and timestamps provide unique identity and order information for data packets, enabling the receiving end to accurately identify the order of data packets and detect lost or duplicate data packets. Redundancy coding adds fault tolerance before data transmission, eliminating the need for immediate retransmission even if poor channel quality leads to partial data corruption, thereby improving transmission efficiency. When data packets are out of order or lost during transmission, the out-of-order detection and reordering functions at the receiving end can restore the correct order of data, while the retransmission request mechanism ensures that all lost data packets are eventually successfully received. Through these collaborative efforts, it is ensured that the original performance status information and image data can be transmitted and reassembled with high integrity and high reliability in complex wireless environments, laying a solid foundation for subsequent accurate image correction and lesion identification.

[0080] The following is a specific example to illustrate this.

[0081] Imagine a wilderness emergency scenario where medical personnel use a portable DR (Digital Radiography) device to perform an X-ray examination on a patient. After the DR detector acquires image data, its internal sensors simultaneously obtain performance status information such as dark current deviation, response sensitivity deviation, and positional offset of the imaging unit. To transmit this data to a portable computer for analysis, the system first divides the image data and performance status information into blocks, for example, dividing the image data into 1MB blocks, and correspondingly dividing the performance status information. Next, each data block is assigned a unique sequence number (e.g., incrementing from 1) and a timestamp, and redundantly encoded (e.g., using Reed-Solomon encoding) to form multiple data packets. These encoded data packets are then wirelessly transmitted to a nearby portable computer via the DR detector's built-in Wi-Fi module.

[0082] During transmission, some data packets may become out of order or be lost due to environmental interference. For example, data packet 5 may arrive after data packet 6, or data packet 3 may be completely lost. The portable computer's receiving module continuously receives these data packets and uses their sequence numbers for out-of-order detection. When data packet 5 is detected to have arrived after data packet 6, the system reorders it according to its sequence number. Simultaneously, the system checks the continuity of the sequence numbers; if data packet 3 is found to be lost, the portable computer automatically sends a retransmission request to the DR detector, requesting retransmission of data packet 3. Once data packet 3 is successfully retransmitted and received, all data packets undergo integrity verification and decoding to recover the original data blocks. Finally, these recovered data blocks are reassembled to form complete, lossless performance status information and image data. This complete and reliable data is then encapsulated into an integrated data stream and transmitted to the portable computer's internal image correction and image analysis modules for subsequent precise correction and lesion identification, ensuring high-quality diagnostic images even in harsh field environments.

[0083] Portable DR image lesion analysis systems are widely used in remote areas with relatively scarce medical resources or in scenarios requiring large-scale health screenings. These systems aim to provide initial image diagnostic assistance to primary healthcare personnel, thereby improving the efficiency of screening efforts. However, long-term use of these portable devices in complex real-world environments presents unique challenges that can affect the quality of image acquisition and the accuracy of subsequent lesion analysis. Specifically, prolonged use of portable DR devices in frequent handling or harsh environments (such as dusty or humid conditions) can lead to gradual and localized subtle drifts in their internal structure and performance. These subtle performance drifts result in non-uniform quality issues in the generated digital images, such as increased local noise, decreased sensitivity, or geometric distortion. Existing image analysis systems, lacking the ability to precisely perceive the real-time operating status of the DR detector, cannot adaptively correct or compensate for these localized image quality degradations caused by detector performance drifts, thus affecting the accuracy of lesion analysis.

[0084] like Figure 2 As shown in the embodiments, this application also discloses a portable DR image lesion analysis system, including: Information acquisition module 1 is used to acquire the performance status information of each imaging unit inside the detector. The performance status information includes the dark current deviation, response sensitivity deviation and position offset of the imaging unit. The data encapsulation and transmission module 2 is used to encapsulate the performance status information and the image data acquired by the imaging unit to form an integrated data stream and transmit the integrated data stream. Data parsing module 3 is used to receive the integrated data stream and parse performance status information and image data from the integrated data stream; Image correction module 4 is used to perform gain adjustment, offset correction and geometric correction on image data based on the parsed performance status information, so as to eliminate local image defects in image data caused by detector performance drift. Image analysis module 5 is used to provide processed image data to the image analysis module for lesion identification.

[0085] This application significantly improves the accuracy and reliability of lesion analysis in portable DR images, providing more reliable diagnostic assistance for primary healthcare personnel.

[0086] The above are merely embodiments of this application and are not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A portable DR image lesion analysis method, characterized in that, Includes the following steps: Acquire performance status information of each imaging unit inside the detector; The performance status information and the image data acquired by the imaging unit are encapsulated to form an integrated data stream, and the integrated data stream is transmitted. Receive the integrated data stream and parse the performance status information and image data from it; Based on the parsed performance status information, the image data is processed by gain adjustment, offset correction and geometric correction to eliminate local image defects caused by detector performance drift. The processed image data is provided to the image analysis module for lesion identification.

2. The portable DR image lesion analysis method according to claim 1, characterized in that, The process involves acquiring performance status information of each imaging unit inside the detector, including dark current deviation, response sensitivity deviation, and position offset of the imaging unit.

3. The portable DR image lesion analysis method according to claim 1, characterized in that, The step of providing the processed image data to the image analysis module for lesion identification includes: Acquire sensor readings from the imaging unit; Analyze the dynamic patterns of sensor readings to identify abnormal fluctuations that are inconsistent with the drift of the detector's own performance; Electromagnetic interference feature maps are generated by quantifying the intensity, duration, and affected pixel regions of abnormal fluctuations. Electromagnetic interference feature maps, image data, and performance status information are encapsulated to form an integrated data stream; Transmitting integrated data streams; It receives the integrated data stream and extracts electromagnetic interference feature maps, image data, and performance status information from the integrated data stream; Based on the parsed performance status information, the image data is initially corrected; The preliminarily corrected image data and electromagnetic interference feature maps are provided to the image analysis module for lesion identification.

4. The portable DR image lesion analysis method according to claim 1, characterized in that, The steps for obtaining the performance status information of each imaging unit inside the detector include: Acquire sensor readings from the imaging unit; Track and analyze historical readings of the sensor to establish long-term performance drift trends; Identify systematic biases in sensor readings that do not conform to long-term performance drift trends, and these systematic biases are related to environmental changes or usage duration. Based on systematic bias, it is determined that the sensor itself has measurement errors due to performance drift or degradation; Based on the preset sensor calibration parameters, the sensor readings are compensated to correct the sensor's own measurement error; The compensated sensor readings are used as performance status information.

5. The portable DR image lesion analysis method according to claim 1, characterized in that, The step of performing gain adjustment, offset correction, and geometric correction on the image data based on the parsed performance status information to eliminate local image defects caused by detector performance drift further includes: Content encoding is performed on the performance status information and the image data acquired by the imaging unit to obtain the encoded performance status information and the encoded image data. The encoded performance status information and the encoded image data are encapsulated to form an integrated encoded data stream; Generate unique identification information related to the integrated coding data stream; The unique identification information and the encoded integrated data stream are encapsulated to form a secure integrated data stream; Transmit secure integrated data stream; Upon receiving the secure integrated data stream, the integrity of the secure integrated data stream is verified to confirm that the secure integrated data stream has not been tampered with during transmission. After the integrity verification is passed, the secure integrated data stream is decoded to recover performance status information and image data; Based on the recovered performance status information, the image data is processed by gain adjustment, offset correction and geometric correction to eliminate local image defects caused by detector performance drift.

6. The portable DR image lesion analysis method according to claim 1, characterized in that, The step of encapsulating performance status information and image data acquired by the imaging unit to form an integrated data stream, and transmitting the integrated data stream includes: The image data acquired by the imaging unit is divided into multiple image data blocks; The performance status information is divided into multiple performance status information blocks based on the image data blocks; A data packet is generated for each image data block and its corresponding performance status information block. The data packet contains a data packet header, which includes the data packet sequence number, the data block type identifier, and verification information. Data packets are prioritized, with performance status information packets marked as high priority and image data packets marked as normal priority. Data packets are transmitted to the portable computer via the detector's data interface in order of priority.

7. The portable DR image lesion analysis method according to claim 1, characterized in that, The step of encapsulating performance status information and image data acquired by the imaging unit to form an integrated data stream, and transmitting the integrated data stream includes: The performance status information and image data acquired by the imaging unit are processed in blocks, and each data block is redundantly encoded. The encoded performance status information data block and the encoded image data block are encapsulated to form multiple data packets of an integrated data stream; Based on the health status map of the storage medium, select a healthy storage area and write multiple data packets of the integrated data stream into the healthy storage area; During the writing of multiple data packets into the integrated data stream, monitor the success rate of the write operation and data integrity; When a write error is detected, the damaged data packet is rewritten to the backup storage area, and the health status map of the storage medium is updated. After multiple data packets of the integrated data stream are successfully written, the integrated data stream in the storage medium is verified to confirm the integrity of the data. Transmitted integrated data stream that has passed verification.

8. The portable DR image lesion analysis method according to claim 1, characterized in that, The step of performing gain adjustment, offset correction, and geometric correction on the image data based on the parsed performance status information to eliminate local image defects caused by detector performance drift includes: Obtain the instantaneous interference signal strength of the detector's operating environment; Compare the instantaneous interference signal strength with the preset interference threshold; When the instantaneous interference signal strength exceeds the interference threshold, the parsed performance status information is corrected in real time. Based on the corrected performance status information, the image data is subjected to gain adjustment, offset correction, and geometric correction.

9. The portable DR image lesion analysis method according to claim 1, characterized in that, The step of encapsulating performance status information and image data acquired by the imaging unit to form an integrated data stream, and transmitting the integrated data stream includes: The performance status information and image data acquired by the imaging unit are processed in blocks; Each data block is appended with a sequence number and a time identifier, and a data packet is generated. Redundant encoding is applied to the data packets; The encoded data packets are transmitted via the wireless communication module; At the receiving end, out-of-order detection and duplicate data packet removal are performed on the received data packets, and the data packets are reordered and their integrity is verified according to the sequence number and time identifier of the data packets. When a packet loss is detected, a retransmission request is triggered; Decode the successfully retransmitted data packets to recover the data blocks; The recovered data blocks are reassembled to obtain complete performance status information and image data; Complete performance status information and complete image data are encapsulated to form an integrated data stream, which is then transmitted.

10. A portable DR image lesion analysis system, characterized in that, include: The information acquisition module is used to acquire performance status information of each imaging unit inside the detector; The data encapsulation and transmission module is used to encapsulate performance status information and image data acquired by the imaging unit to form an integrated data stream and transmit the integrated data stream. The data parsing module is used to receive the integrated data stream and parse performance status information and image data from it. The image correction module is used to perform gain adjustment, offset correction and geometric correction on the image data based on the parsed performance status information, so as to eliminate local image defects caused by detector performance drift in the image data. The image analysis module provides processed image data to the image analysis module for lesion identification.

Images