296 results about "Medical imaging technology" patented technology

The invention discloses a mild traumatic brain injury image evaluation system based on deep learning, and relates to the technical field of medical imaging, and the system comprises a data source module which is used for obtaining head magnetic resonance imaging data as an input data source; the preprocessing module is used for carrying out de-noising processing and image registration on the skull magnetic resonance imaging data to realize standardized preprocessing; and the feature extraction module is used for extracting a local feature map from the multi-dimensional image data by constructing a multilayer convolutional neural network, and identifying a subcortical structure change and damage suspicious region. According to the mild traumatic brain injury image evaluation system based on deep learning provided by the invention, a standardized evaluation report automatically generated by the system not only integrates clinical information of a patient, but also provides detailed injury types, severity and follow-up visit suggestions, so that the efficiency and reliability of clinical decision making are greatly improved; and powerful support is provided for formulating an individualized diagnosis and treatment scheme.

The present invention relates to the technical field of medical images. Disclosed are a deep learning-based CT artifact removal method and system. The method comprises: acquiring a CT image, and separately performing sine transform and wavelet transform processing on the CT image; constructing an image enhancement model, and performing image optimization on the processed image separately by means of the image enhancement model and a random inversion layer which are connected in sequence; coupling the optimized image and the original image and then inputting the coupled image into the image enhancement model for reprocessing; and performing element-wise addition on the reprocessed image and the optimized image to obtain an artifact-removed CT image. In the present invention, wavelet transform is introduced to process the CT image to extract the context and spatial information of the CT image, effectively extracting feature information in an artifact removal process and improving the performance of image enhancement; and a CT image resolution enhancement model based on a VMamba model is established, enhancing the long-term dependencies in network training, effectively recognizing and removing radioactive artifacts, and improving the network training efficiency.

The invention discloses a self-adaptive adjustment method and system for operating parameters of color ultrasound equipment, and relates to the technical field of medical imagines.The method comprises the steps that in the cardiac ultrasound examination process, image feature state data information is associated with echo signal state data information, and a motion artifact evaluation coefficient Xwy is obtained; after comparison, an artifact image frame set is marked, and a gain adjustment analysis instruction is sent out; the method comprises the following steps: acquiring vascular rigidity state data information according to interference of vascular instantaneous relaxation on echo signals, and associating the vascular rigidity state data information with a corresponding motion artifact evaluation coefficient Xwy to obtain an echo signal gain degree index Zzy; through comparative analysis, whether each frame of echo signal corresponding to the current artifact image frame set needs gain compensation is judged so as to generate and execute a gain adjustment instruction of a corresponding grade, and the process optimizes the gain adjustment of the echo signal, improves the image quality, ensures the definition of ultrasonic imaging, and improves the accuracy of ultrasonic imaging. Therefore, the accuracy and the diagnosis reliability of cardiac ultrasonic examination are improved.

The invention belongs to the technical field of medical images, and particularly relates to an anterior mediastinal surgery three-dimensional imaging method and system based on big data. According to the method, parallel preprocessing is carried out on the multi-modal medical image data through a distributed computing framework, the data processing speed and efficiency are effectively improved, through comparative analysis of different reference image data, an overlapping area and an independent area can be determined, corresponding partition processing is carried out on the overlapping area, and in the three-dimensional reconstruction stage, the three-dimensional reconstruction efficiency is improved. Three-dimensional reconstruction is respectively carried out on the independent region and the overlapped sub-regions, a three-dimensional image model of the independent region and a fused overlapped sub-region three-dimensional image model are generated, the spatial topological structure and the surface geometric characteristics of image data are fully considered, the accuracy of the reconstructed image is ensured, and in the image reconstruction stage, the image reconstruction efficiency is improved. And the three-dimensional image model of the independent region and the three-dimensional image model of the overlapped sub-regions are combined to generate a complete anterior mediastinal surgery three-dimensional image, so that a powerful guarantee is provided for accurate implementation of the surgery.

The invention relates to the technical field of medical images, and discloses a medical image recognition method and system based on deep learning, electronic equipment and a medium, and the method comprises the steps: carrying out the split-mode adaptive normalization of an initial medical image, carrying out the preprocessing through elastic deformation, random zooming and Gaussian noise, and obtaining a preprocessed medical image, constructing a Swinin-ResNeSt hybrid model based on the preprocessed medical image, adding an adaptive multi-scale attention module in the Swinin-ResNeSt hybrid model, designing double task heads including classification and positioning, adopting a hybrid loss function combining a Focal Loss function and a GIoU Loss function, and outputting a focus category and a bounding box; according to the invention, the identification accuracy is improved.

The invention relates to the technical field of medical imaging, and discloses a method for automatically discriminating and enhancing phases in CT scanning by an attention mechanism based on multi-instance learning, and the method comprises the steps: extracting a CT scanning image and DICOM information from a data set, and unifying the image size after HU conversion, truncation and linear zooming; obtaining a fixed number of slices by using direct or indirect random sampling and a cubic interpolation strategy; then, a pre-trained model is adopted to extract slice features, scanning level feature vectors are formed through attention aggregation, and finally prediction probabilities of four types of phases are output through a full connection layer and a Softmax module; the model is trained by adopting cross entropy loss and an Adam optimizer, the performance is evaluated through indexes such as precision, recall rate and F1-score, and a thermodynamic diagram is generated by combining gradient back propagation, so that visual visualization of a result is realized, and clinical judgment is assisted. According to the method, the CT scanning multi-phase can be accurately, rapidly and interpretably classified in a real clinical environment.

The invention relates to the technical field of medical imaging, and discloses an image processing method and system for a dual-energy X-ray bone mineral densitometer, and the method comprises the steps: collecting a dual-energy X-ray image, and carrying out the light and dark field correction of original projection data; based on the corrected projection data, using an aluminum sheet to simulate bones and glass to simulate tissues, carrying out data fitting to obtain a bone mineral density value, and fitting a bone-aluminum conversion coefficient; preprocessing the image, wherein the preprocessing comprises Gaussian smooth filtering, threshold processing and corrosion operation; segmenting the image by using a region growing algorithm, and carrying out edge detection by using a Canny operator; extracting an ROI (Region of Interest), and determining the boundary of the ROI by searching coordinates meeting a preset threshold condition within a preset proportion range in the vertical direction of the image; and calculating the area of the ROI, the bone mineral density BMD and the bone mineral salt content. The accuracy of bone mineral density measurement is improved, the influence of noise and artifacts can be effectively reduced, and the target area is accurately extracted.

The invention provides a tumor cell assisted positioning system and method based on a composite image, and relates to the technical field of medical image.The system comprises a data acquisition module used for synchronously acquiring radio frequency echo signals, ultrasonic images and computed tomography data of target biological tissue to generate an original fusion data set; the feature processing module is used for performing spatial registration and feature extraction on the original fusion data set to obtain a multi-dimensional feature set; and the identification imaging module is used for extracting a multi-band radio frequency echo signal intensity level value of the target area based on the multi-dimensional feature set, generating a tissue radio frequency characteristic diagram through logarithmic compression, and performing pixel-level fusion on the tissue radio frequency characteristic diagram and the synchronous ultrasonic image to generate a composite image. According to the method, the three-dimensional coordinates of the tumor boundary are determined through multi-modal data fusion and precise calibration in combination with anatomical structure constraints, and the accuracy of tumor cell positioning is improved.

The invention belongs to the technical field of medical imaging, and discloses an angiography high-pressure injection system which comprises the following steps: acquiring physiological data information of a patient, and calculating an individualized contrast agent injection parameter coefficient based on the physiological data information; acquiring anatomical structure information of a blood vessel region of the target patient, and performing pressure evaluation based on hemodynamic characteristics to obtain blood vessel pressure bearing capacity and injection pressure regulation factors; performing physiological status assessment based on the physiological data information of the patient to obtain a physiological status influence factor and generate a risk marker value; performing multi-source data fusion according to the injection pressure regulation factor, the physiological status risk marker value and the historical radiography record, and constructing the radiography safety degree of the blood vessel sub-region; an optimal injection parameter path is obtained based on the radiography safety degree in combination with a deep reinforcement learning algorithm, and high-pressure injection equipment is controlled to execute injection operation; according to the invention, accurate and personalized control of contrast agent injection is realized, and the safety and imaging effect of contrast examination are remarkably improved.

The invention discloses a CT image change-based lung cancer curative effect dynamic prediction method and system, and relates to the technical field of medical imaging. Clinical information, a baseline CT image and a follow-up CT image of a patient are acquired; performing feature extraction on the baseline CT image and the follow-up CT image to obtain baseline features and follow-up features; obtaining an enhanced CT dynamic feature matrix and a flat-scanning CT dynamic feature matrix based on the baseline features and the follow-up visit features; inputting the enhanced CT dynamic feature matrix, the plain-scan CT dynamic feature matrix and clinical information into a curative effect prediction model, outputting logic risk scores at different time points, and obtaining progress or death risk probabilities of the patient at different time points; according to the method, image feature changes in a baseline and follow-up visit are effectively represented through a dynamic feature matrix, correlation between the interior of the dynamic feature matrix and between different matrixes is identified through a Transform attention network to predict the curative effect, and the relationship between accumulated data at each follow-up visit time point and the curative effect is learned through a time-sharing training strategy.

The embodiment of the invention is suitable for the technical field of medical imaging and image calibration, and provides an ultrasonic probe calibration method and system and electronic equipment, and the method comprises the steps: receiving an ultrasonic probe pose sequence and a calibration model pose sequence; receiving an ultrasonic image sequence; performing mark point identification and straight line identification on the ultrasonic image sequence to determine a first noise image and a second noise image in the ultrasonic image sequence; after the second noise image is deleted from the ultrasonic image sequence, time calibration is carried out on the ultrasonic probe, and a rate difference value between an ultrasonic probe displacement signal and an ultrasonic image displacement signal is obtained; and after deleting the first noise image from the ultrasonic image sequence, performing spatial pose calibration on the ultrasonic probe based on the rate difference value to obtain a spatial pose calibration matrix of the ultrasonic probe. By adopting the method, the calibration precision of the ultrasonic probe and the stability of the calibration result can be improved.

The invention discloses a dynamic DR lung field area diagnosis system based on deep learning, and relates to the technical field of medical imaging, and the system comprises an image collection optimization module which is used for capturing a low-dose dynamic DR sequence of a complete period of calm respiration and best effort respiration of a patient, generating a dynamic DR image, and optimizing the dynamic DR image through employing a multi-mode transfer learning strategy, outputting a DR time sequence tensor; and the intelligent diagnosis module is used for carrying out disease cause classification through a gating chart convolutional network in combination with the lung field dilation function ratio and the space structure complexity index, outputting a classification result, and generating a structured report in combination with the lung field dilation function ratio, the space structure complexity index and the continuous lung field area change curve. The differential homeomorphic transformation field is generated through the Lie group exponential mapping algorithm, and the lung lobe boundary connectivity and the bronchial tree topological structure are strictly kept in the three-dimensional deformation process.

The invention relates to the technical field of medical images, in particular to a medical image intelligent detection and auxiliary diagnosis system based on deep learning, and the system comprises a data access module which is used for obtaining medical image data and clinical text data of a patient; the data fusion module is used for generating a focus feature vector and a text feature vector, and performing cross-modal alignment and fusion to generate a fusion feature vector; the diagnosis analysis module is used for executing focus detection, focus segmentation and focus classification tasks, generating a diagnosis result and generating a diagnosis label based on the diagnosis result; the decision generation module is used for mapping the generated execution result to a preset medical knowledge base and performing deep reasoning to generate an auxiliary diagnosis decision; the decision auditing module is used for performing confidence scoring on the auxiliary diagnosis decisions and selecting the auxiliary diagnosis decision with the highest confidence score as the final auxiliary decision; and the data visualization module is used for carrying out visualization processing on the auxiliary decision and the diagnosis result.

The invention relates to the technical field of medical images, in particular to a multi-modal medical image registration and fusion analysis method, which comprises the following steps of: eliminating image noise and artifacts based on a modal adaptive filtering strategy; constructing a pyramid type feature extraction network to realize multi-scale feature extraction, calculating feature matching degrees among different modal images, and dynamically adjusting matching weights by combining feature differences among modals; a focus area attention mask is constructed, targeted enhancement of registration image features is realized, and a hierarchical fusion strategy is adopted to evaluate the quality of a fused image; and constructing a multi-task deep learning model to complete focus automatic detection, segmentation and benign and malignant preliminary judgment on the fused image. According to the multi-modal medical image registration and fusion analysis method, a three-layer feature pyramid is constructed, a multi-feature fusion matching cost function is introduced, and cross-modal feature matching is optimized through an adaptive weight iteration nearest point ICP algorithm, so that the information richness, marginal definition and focus discrimination of a fused image reach the standard.

The invention relates to an oxidase response type molecular probe as well as a preparation method and application thereof, and belongs to the technical field of medical imaging. Aiming at the problem that the inflammation activity is difficult to accurately evaluate in the prior art, the invention provides a molecular imaging probe capable of responding to the activity of in-vivo myeloperoxidase (MPO), and the molecular imaging probe can be used for real-time and non-invasive imaging monitoring of the activity of the in-vivo MPO, so that the activity of the inflammatory reaction of a focus is accurately evaluated. The invention provides a chelating agent with structural characteristics of a chelating unit and an enzyme response unit, and is characterized in that the enzyme response molecular probe is constructed by taking rigid framework cyclohexanediamine (CDTA) as a transition metal chelating unit and taking electron-rich functional groups (such as phenol, catechol or 5-hydroxytryptamine) as the enzyme response unit. The probe is chelated with paramagnetic metal (such as Mn < 2 + >) for magnetic resonance imaging (MRI) or radioactive metal (such as 68Ga) for positron emission tomography (PET), so that specific response to MPO activity is realized. The dynamic stability of the probe is remarkably improved; after the MPO is responded, the relaxation efficiency is improved by 2-3 times, and the sensitivity is optimized; the cytotoxicity is low; in an acute pancreatitis model, the MRI contrast noise ratio (delta CNR) reaches 77.39 + / -7.04 and is 4.2 times that of a control group, the PET uptake rate is remarkably improved, bimodal imaging is supported, and accurate positioning and quantitative evaluation of the inflammation activity are achieved.

Implantable transponders comprising no ferromagnetic parts for use in medical implants are disclosed herein. Such transponders may assist in preventing interference of transponders with medical imaging technologies. Such transponders may optionally be of a small size, and may assist in collecting and transmitting data and information regarding implanted medical devices. Methods of using such transponders, readers for detecting such transponders, and methods for using such readers are also described.

The invention discloses a chest image diagnosis method and system based on multi-modal sign collection, and relates to the technical field of medical image.The method comprises the steps that a chest image of a patient is obtained, electrocardiosignals and blood oxygen saturation data are synchronously collected, and an associated radiology report is obtained; the image lesion features and the frequency domain rhythm template are combined for processing, motion artifacts are eliminated through a frequency domain decoupling equation, and refined image features are output; inputting the refined image features and the text pathological semantic features into a bidirectional attention mechanism to generate fusion features, and splicing the oxyhemoglobin saturation data and the text pathological semantic features into a sign-text vector; and inputting the fusion feature and the sign-text joint vector into a multi-task loss function, and outputting a structured diagnosis report. According to the method, accurate elimination of motion artifacts is achieved through a frequency domain decoupling equation, and coupling calculation is conducted on an electrocardio rhythm template and image lesion features in a frequency domain space.

The invention relates to the technical field of medical imaging, and discloses a peripheral lung cancer pleural invasion evaluation system based on ultrasonic targeted radiography, comprising: a double-frequency ultrasonic probe configured to independently work at different frequencies to realize depth imaging and high-resolution imaging of lung tissues and suitable for different types of lung lesions; the signal processing module can receive, analyze and process echo signals transmitted by the ultrasonic probe, generate images in real time and extract feature information; and the dynamic motion compensation module is used for automatically adjusting the emission time sequence and parameter setting of the ultrasonic signal based on the real-time breathing motion state of the patient. By adopting the design of the double-frequency ultrasonic probe and the core-shell structure, high-resolution imaging and depth detection of lung tissues are realized. By dynamically adjusting the thickness of the probe shell to be matched with the frequency, the system can adapt to complex tissue environments under different physiological conditions, and image definition and stability are ensured.

The invention relates to the technical field of medical images, in particular to a CT image segmentation method and system based on deep learning. The method comprises the following steps: acquiring an original CT image; performing boundary deep learning pre-segmentation on the original CT image based on a deep learning segmentation network to obtain an initial contour probability graph; performing preliminary interface area positioning on the initial contour probability map, and performing normal vector extraction to obtain a preliminary interface point cloud and a normal vector set; performing density gradient analysis on the compressed tissue candidate area according to the initial interface point cloud and the normal vector set to obtain an interface adjacent density feature map; performing fat gap integrity evaluation according to the initial interface point cloud and the normal vector set to obtain an interface fat gap quality map; and constructing an interface mechanical characteristic graph according to the interface adjacent density characteristic graph and the interface fat gap quality graph. According to the method, the segmentation accuracy and the adaptability to a complex pathological interface are improved through the image features changed by the biomechanical characteristics.

The invention discloses a complex medical image intelligent recognition system and method based on deep learning, and relates to the technical field of medical image.The system comprises an anatomical image data acquisition module, an anatomical image data analysis module, a functional image analysis module and a medical image evaluation module. By combining the anatomical image, collecting the anatomical information, evaluating the image evaluation index of each anatomical tissue of the medical image and matching each risk tissue, subjective deviation of manual measurement is avoided, the risk tissue detection efficiency and positioning precision are improved, and the processing efficiency is improved. According to the method, the cell metabolic activity of each risk tissue is captured, the functional influence assessment index of each risk tissue of the medical image is assessed, the medical image recognition coincidence value is assessed and optimized, the limitation that only the morphological structure is observed in the traditional anatomical image is broken through, the diagnosis accuracy and reliability are improved, and the medical image diagnosis development is promoted.

The invention discloses a default mode network construction method, system and device and a storage medium, and relates to the technical field of medical imaging, and the method comprises the following steps: selecting a plurality of seed points in an fMRI image, constructing an original default mode network based on the seed points, and obtaining a corresponding function connection diagram; overlapping the function connection diagrams generated by the seed points to obtain an overlapping graph; screening the functional connection strength of each voxel in the overlapped graph based on a correlation threshold, and obtaining a plurality of new ROIs according to a screening result; and extracting a new ROI time sequence, calculating Pearson correlation coefficients of the time sequence, and generating a function connection matrix based on a plurality of Pearson correlation coefficients, namely generating a new default mode network. According to the method, the DMN network structure can be reflected more comprehensively through the seed points at different positions, the one-sidedness of a single seed point is made up, and a more accurate DMN network can be constructed.

The invention provides a liver cancer onset risk prediction method and system, and belongs to the technical field of liver cancer onset risk prediction. The method comprises the following steps: carrying out longitudinal multi-omics data acquisition on a target individual to generate an individualized multi-omics time sequence data set; constructing an individualized liver cancer evolution graph based on the data set; according to the individualized liver cancer evolution graph, quantifying a liver microenvironment pressure field in combination with a medical imaging technology, and performing space-time coupling on a pressure field quantification result and a multi-omics time sequence data set to generate a liver cancer dynamic risk evolution trajectory; through longitudinal multi-omics data acquisition and individualized liver cancer evolution graph construction, a liver microenvironment pressure field is quantified in combination with a medical imaging technology, dynamic and accurate prediction of the liver cancer occurrence risk is realized, and the prediction accuracy is remarkably improved.

The invention is suitable for the technical field of medical imaging, and provides an ultrasonic signal processing system and method and a brain-like calculation module.The system comprises an ultrasonic transducer, an ultrasonic transmitting module and the brain-like calculation module.The ultrasonic transducer is coupled with the ultrasonic transmitting module and the brain-like calculation module; the ultrasonic transmitting module is used for controlling the ultrasonic transducer to transmit ultrasonic waves; and the brain-like calculation module is used for processing the echo signal received by the ultrasonic transducer based on brain-like calculation.

The invention discloses a quality control method and system for an intra-frame weighted medical image with similar specific attention, relates to the technical field of medical images, and solves the technical problems that when an existing medical image processing method is used for processing a 3D medical image, the discrimination capability is weak, and the sensitivity to subtle artifacts or body position deviation is low, so that the recognition capability is weak. Comprising the steps that medical image data are acquired, and the medical image data comprise a 3D medical image and an evaluation system corresponding to the 3D medical image; performing data labeling on the 3D medical image based on an evaluation system to obtain a plurality of sequence data; dividing the plurality of sequence data into a training set, a verification set and a test set; training the medical image quality control model based on the training set, the verification set and the test set; the training process comprises data preprocessing, spatial feature extraction, time sequence enhancement feature extraction, class-guided intra-frame weighted score calculation, model optimization and class activation graph generation; and the model identification capability is enhanced.

The invention relates to the technical field of medical images, and discloses a medical image zero-watermark protection method and device based on depth frequency domain fusion features, and the method comprises the steps: obtaining carrier image data; preprocessing the carrier image data to obtain a standardized carrier image with a preset size; performing depth frequency domain fusion feature extraction on the standardized carrier image to obtain a binary feature vector; carrying out encryption processing on the original watermark by adopting a logic mapping chaotic system with an optimized control parameter range to obtain an encrypted watermark; performing bit-by-bit XOR operation on the encrypted watermark and the binary feature vector to obtain a zero-watermark key, and uploading the zero-watermark key to a third-party authentication center; when copyright authentication or tracing needs to be carried out on the medical image, a binary feature vector is extracted from the medical image to be verified, and a zero-watermark key stored in a third-party authentication center is downloaded; performing bit-by-bit XOR operation on the binary feature vector of the to-be-verified medical image and the zero watermark key to obtain a to-be-decrypted watermark; and decrypting the watermark to be decrypted through the logic mapping chaotic system to obtain an original watermark.

The invention relates to the technical field of medical imaging, in particular to a pelvic stability dynamic detection method and stress analysis system based on dynamic DR. The method comprises the steps that on the basis of sign information of a user, parameters of a three-dimensional infrared motion capture system, a force measuring table and a pressure sensor are adapted to obtain static CT or MRI image data of a pelvis of a patient, and then the pelvic stability dynamic detection method and stress analysis system are obtained; establishing a pelvis coordinate system and defining geometric features of key points; collecting to-be-detected data of a user in the load bearing state; preprocessing the collected user data, and reconstructing a three-dimensional pelvis model of the user; performing biomechanical analysis based on the reconstructed three-dimensional pelvis model of the user to generate a pelvis stability index; and according to the pelvic stability index, automatically generating a pelvic stability risk index, and outputting a clinical decision suggestion. The problems that a traditional pelvis detection method still depends on a static image, dynamic stress distribution and stability evaluation cannot be effectively quantified, and an evaluation result is not accurate enough are solved.

The invention discloses a non-invasive pelvic bone tracking system for hip joint surgery, and belongs to the technical field of intelligent surgical navigation and medical imaging. A non-invasive pelvic bone tracking system for hip joint surgery comprises a multi-modal sensing module, a dynamic resolution optimization module, an AI correction module and a multi-region parallel tracking engine, and the multi-modal sensing module comprises a 3D ultrasonic unit, a near-infrared optical unit and an inertial measurement unit. The problems of positioning error and limitation of scanning speed and coverage range caused by fat interference in the prior art are solved. According to the method, low-frequency ultrasonic, near-infrared optical and IMU data are fused, scanning modes are intelligently switched based on fat thickness, the speed and precision are balanced, ultrasonic skeleton features and optical body surface topology are jointly analyzed, fat attenuation dynamic compensation is achieved, multi-region GPU parallel tracking supports intra-operative real-time multi-target collaborative navigation is achieved, and multi-target collaborative navigation is achieved. Millimeter-level real-time tracking of obese patients during operation is realized.

The invention relates to the technical field of medical images, and discloses a fusion method for a CT image and a CT-like image, and the method comprises the steps: obtaining a preoperative CT image and an intraoperative CT-like image of a patient; acquiring a first spine key point label of a spine key point in the CT image and a second spine key point label of a spine key point in the CT-like image by using a spine key point segmentation model; based on the first spine key point mark number and the second spine key point mark number, a spine pairing classification model is used for pairing spines in the CT image and the CT-like image, and a spine pairing result is obtained; and according to a spine pairing result, performing registration fusion display on the CT image and the CT-like image. According to the method, the spine key point mark number is accurately extracted by using the spine key point segmentation model, intelligent pairing and registration fusion based on the spine key point mark number are performed by using the spine pairing classification model, and the registration fusion precision of the CT image and the CT-like image is improved. The invention further discloses a fusion device for the CT image and the CT-like image and CT equipment.

The invention relates to the technical field of medical imaging, in particular to a breast cancer tissue HER2 automatic scoring method and device based on a characteristic curve, and the method comprises the steps: carrying out the reading operation of a full-slice image of immunohistochemical staining; extracting a target region of interest of the full slice image; slicing and screening the target region of interest to obtain a plurality of image slices; calculating a percentage-saturation characteristic curve of each image slice, and extracting characteristic information of the percentage-saturation characteristic curve; based on the feature information of the percentage-saturation feature curve, performing preliminary HER2 scoring on each image slice in combination with a rule base and a logistic regression classification algorithm; and integrating the scoring results of the image slices, and outputting a final HER2 score and a characteristic curve visualization graph. According to the method, the feature curve is combined with the rule base and the logistic regression classification algorithm, so that the objectivity of HER2 scoring is improved, and the problems of subjectivity and low efficiency of manual evaluation are reduced.