15 results about "Small Lesion" patented technology

The invention discloses a DR focus segmentation method and system based on a dynamic loss optimization function, and the method comprises the steps: collecting image data in an IDRiD data set, and carrying out the preprocessing of the image data, and obtaining the preprocessing data; based on the preprocessed data, a visual SAM large model is constructed, a loss function weight optimization strategy is designed, and an RTSAM segmentation model is obtained; segmenting the DR eye fundus image based on an RTSAM segmentation model to obtain a DR focus segmentation result; and based on the DR focus segmentation result, providing an automatic identification and segmentation visualization result of the DR focus. Aiming at the problems of non-uniform lesion distribution, non-uniform lesion occurrence frequency and the like existing in a DR color fundus image, the segmentation effect is poor, so that a dynamic loss optimization strategy is used, weight distribution in a loss function is adaptively adjusted according to a model segmentation index, and the attention of a model to rare lesion and small lesion areas is enhanced.

The invention discloses an implantable cerebral apoplexy treatment stimulator, which comprises at least two electrode parts and at least two skull nails, the at least two electrode parts are suitable for being respectively implanted into implantation holes of at least two skull, the electrode parts are connected with wires, the wires are connected with a pulse generator, each skull nail comprises a stud part and a screwing part, and the stud part is connected with the pulse generator. The stud part is suitable for being screwed into the implantation hole, the bottom of the stud part abuts against the top of the electrode part, the screwing part is connected to the top of the stud part, an operation groove is formed in the top of the screwing part, and a screw driver can be inserted into the operation groove to rotate the skull nail. The implantable cerebral apoplexy treatment stimulator is simpler in structure and more convenient to fix on the skull, the stimulation intensity can be improved, small-range lesions can be accurately targeted, stimulation is more accurate, in addition, stimulation parameters and stimulation areas can be adjusted and controlled, the implantable cerebral apoplexy treatment stimulator dynamically adapts to cerebral injury conditions of different patients, and the cerebral apoplexy treatment effect is better.

The invention discloses an ultrasonic real-time focus positioning and benign and malignant auxiliary diagnosis platform based on deep learning, and the platform obtains an ultrasonic video stream in real time through an API or a video interface, carries out the focus detection through a CIoU loss optimization-based YOLOv8 model, and introduces a boundary perception auxiliary head and a K-means clustering anchor frame to improve the small focus positioning precision. Then, morphological features such as boundary definition, shape regularity and echo texture of the focus are extracted, benign and malignant classification is carried out based on comprehensive features, uncertainty estimation and conservative dynamic calibration are carried out by utilizing Monte Carlo Dropout, the calibrated malignant probability, confidence coefficient and focus boundary are output by the system, and the accuracy of the focus is improved. Doctor decisions are supported by structured reports and visual warning, the processing speed is larger than or equal to 15 frames per second, and high precision, real-time performance and clinical robustness are achieved.

According to the technical scheme, the invention discloses a method for accurately detecting eye CT (Computed Tomography) lesions through double-branch vision Transformer. The dual-branch ViT structure provided by the invention aims at the characteristics (1) of orbit CT (Computed Tomography): lesions are concentrated in a relatively small area, but the global perception ability still needs to be reserved; 2, the anatomical structure is fine and sensitive to down-sampling operation; however, full-size CT input easily causes unobvious features), overall and local branches can be mutually referenced and supplemented through a cross attention fusion mechanism, the feature capture ability of small focuses is significantly improved, effective cognition of the overall anatomical structure is maintained, and the probability that the features of the small focuses are not obvious is reduced. While the details of the lesion area in the orbit CT are efficiently and accurately captured, the overall anatomical background information is taken into account.

This utility model relates to the field of medical device technology, and in particular to a device for locating small lesions, comprising: a positioning structure, which includes a titanium clip, a sleeve, and a guidewire; the first end of the titanium clip is a clamping end, and the second end of the titanium clip is a traction end; the clamping end extends from the first end of the sleeve; the portion of the titanium clip between the clamping end and the traction end is slidably connected to the sleeve; the traction end is connected to the first end of the guidewire, and the second end of the guidewire extends from the second end of the sleeve; the surface of the sleeve has a luminescent material that can emit light when excited by a light source; and an insertion structure, which includes a catheter, a working handle, and a handle; the first end of the catheter is connected to the second end of the sleeve, and the second end of the catheter is connected to the working handle; the handle is mounted on the working handle and is slidable relative to the working handle; the second end of the guidewire passes through the catheter and is connected to the handle. This utility model is at least beneficial for improving the positioning accuracy of lesions in minimally invasive surgery.

The utility model relates to a mammary gland tiny focus positioning bearded needle guide wire capable of enhancing ultrasonic echo imaging, which is provided with a mammary gland positioning bearded needle, the surface of the mammary gland positioning bearded needle is provided with a layer of porous microcarrier microbubble coating capable of enhancing ultrasonic echo imaging, and a layer of surface super-smooth coating is arranged outside the porous microcarrier microbubble coating. The porous microcarrier microbubble coating is composed of a micron-sized porous microcarrier base body and nano-sized liquid fluorocarbon microbubbles embedded in holes, and is located between the mammary gland positioning crochet hook and the surface super-smooth coating to form a three-layer needle tube structure. The mammary gland positioning bearded needle guide wire not only can enhance ultrasonic development and guide accurate positioning, but also can be safely placed in a human body, and adverse reaction caused by a coating is not generated. According to the mammary gland tiny focus positioning bearded needle guide wire capable of enhancing ultrasonic development, the positions of the positioning bearded needle and the needle point can be clearly observed when a patient is subjected to ultrasonic guided tiny focus positioning or breast cancer operation, and accurate positioning is facilitated.

This invention relates to the field of medical imaging diagnostic technology, specifically to a method and system for precise localization of small lesions in the breast and thyroid glands based on AI multimodal imaging. The method includes the following steps: acquiring ultrasound, X-ray, and magnetic resonance images of the breast and thyroid glands; extracting anatomical spatial location and imaging direction; calculating and mapping multimodal pixel spatial offsets; calculating scale consistency and normalizing response intensity based on co-located pixels; comparing multimodal intensity differences to assess consistency and form a scoring distribution; combining the score and intensity to adjust conditional probabilities; calculating joint probabilities and selecting spatial aggregates to locate small lesions. In this invention, by uniformly calibrating the spatial location and imaging direction of multimodal images, precise correspondence between different images at the same anatomical location is achieved. Multimodal response intensities are fused and consistency assessment is performed. Combining probability weight adjustment and spatial aggregation weakens the influence of noise in single images, enhances the common expression of small abnormal areas, and improves the stability and consistency of lesion localization.

The invention relates to the technical field of medical images, and discloses an organ focus detection method and device, and electronic equipment, and the method comprises the steps: obtaining a scanning image sequence of an organ, the scanning image sequence comprises a plurality of two-dimensional slices; inputting each two-dimensional slice into a slice feature encoder of the focus classification model, and obtaining an initial feature vector of each two-dimensional slice by using the slice feature encoder; performing position coding on the initial feature vector of each two-dimensional slice, inputting the initial feature vector into a self-attention module of a focus classification model, and obtaining an enhanced feature vector, fused with global context information, of each two-dimensional slice by using the self-attention module; and obtaining a focus position detection result based on the enhanced feature vector of each two-dimensional slice. According to the method, through two-stage training, slice feature coding and a self-attention mechanism are combined, the global context is fused while the space sequence is reserved, the tiny focus is effectively detected, the calculation overhead is reduced, and performance, interpretability and practicability are considered.

The utility model provides an endoscope snare structure which comprises a handle, the front section of the handle is connected with an outer sheath tube, a pull wire is arranged in the outer sheath tube, a flushing assembly is further arranged in the outer sheath tube and comprises a flushing tube, the flushing tube is located in the outer sheath tube, the front section of the outer sheath tube is sealed through an end plate, one end of the flushing tube is connected to the end plate, and the other end of the flushing tube is connected to the end plate. The other end of the flushing pipe is connected to the bottom of the liquid adding opening. By means of the flushing assembly, flushing liquid can be injected through the liquid adding opening to flush the surgical site in the surgical process, the surgical visual field is kept clear, the situation that the operation time is prolonged due to the fuzzy visual field is reduced, and the surgical efficiency is improved. The cutting knife and the cutting groove are arranged on the snare, so that a small focus can be more efficiently cut, and for a large focus, the high-frequency electric knife can be connected through the electric knife interface, the traction wire and the snare are electrified, the focus can be more effectively cut, and electric coagulation hemostasis is carried out after the focus is cut. The rotating head is rotationally connected with the bottom of the mounting pipe through the bearing, the angle of the ring can be flexibly adjusted according to the operation requirement, the to-be-cut focus can be conveniently sleeved with the ring, and operation is convenient.

The invention discloses a diabetic retinopathy segmentation system and method based on deep blood vessel feature fusion. Transfer learning is introduced, blood vessel features are deeply fused into lesion features in multiple stages by using blood vessel segmentation data of a real diabetic retinopathy data set, a blood vessel structure is utilized to assist DR segmentation, an MDR-Net framework is adopted, and the segmentation accuracy of the diabetic retinopathy is improved. Through two branches of blood vessel segmentation and lesion segmentation, respective features are extracted respectively, and meanwhile, blood vessel features are fused into lesion branches to assist segmentation. Three key modules of the framework are as follows: a blood vessel guidance progressive fusion module generates a dynamic weight map through space attention to guide lesion extraction, and combines Mama modeling global blood vessel distribution to enable blood vessel priori DR segmentation to play a role in a whole stage; the multi-scale feature fusion module integrates features of different scales of the encoder in the lesion branch, and takes small lesions and large-area lesions into consideration; the lesion specificity attention module is used for constructing an independent feature processing branch for each lesion aiming at the characteristic that the morphological difference of the four lesions is obvious.

The application discloses a kind of orange leaf lesion classification methods based on deep learning model, based on the high-performance classification model DF-SE-ResNet of deep learning, in the feature extraction stage of model, the optimized ResNet50 is combined with FPN (Feature Pyramid Networks), and improved pass attention mechanism module is added, so that model is effectively focused on small lesion plaque while retaining more semantic information around it, lay a solid foundation for subsequent accurate identification.To solve the accurate positioning problem caused by random distribution of lesion plaque, a region feature weighting fusion method based on permutation invariance is introduced after the feature extraction network, which improves the positioning ability of lesion plaque and significantly enhances the lesion features.Finally, in the experiment stage, the accuracy of DF-SE-ResNet reaches 95.4%, and compared with other deep learning models, the performance of the model proposed in the application is better.

The invention relates to the technical field of image recognition, and provides a two-stage full-mouth focus detection method and system, electronic equipment and a medium, and the method comprises the following steps: obtaining an oral CT image, processing the oral CT image through a pre-trained target detection model, positioning and detecting a single tooth, and obtaining a single tooth region image; utilizing the target detection model to continuously identify each single tooth area image to obtain the position and the category of a focus area; wherein a loss function adopted by the target detection model in the training process is formed by weighting a position loss function, a classification loss function and a target confidence loss function. According to the method, the single tooth is firstly positioned, then the focus area is finely identified, the multi-task weighted loss function is combined, the position, the classification and the confidence coefficient are optimized, and the accuracy and the robustness of small focus detection are remarkably improved.

The invention relates to a retinopathy segmentation method, device and system based on a threshold channel regulation mechanism, and the method comprises the steps: collecting a preset number of color fundus images of diabetic retinopathy of different patients, carrying out the preprocessing, and generating a corresponding multi-valued mask image containing a focus condition; constructing a RetinaSAM large model, and introducing a threshold channel regulation and control mechanism into the model; dividing a training set and a test set based on the preprocessed color eye fundus image and the corresponding multi-valued mask image, and then performing model training and testing; and inputting a to-be-diagnosed color eye fundus image into the large model to obtain a focus segmentation result. According to the method, the large model introducing the threshold channel regulation and control mechanism is constructed, and the database containing different types of lesion information is used for model training and testing, so that the recognition capability of the model on small lesion, weak features and complex lesion areas is improved, and the model segmentation precision is improved.

The utility model discloses a small-focus lung cancer tissue fixator which comprises a fixing support and a fixing set. The fixing support comprises a frame and a rotating shaft. At least two rotating shafts are fixedly connected to the interior of the frame; the fixing assembly comprises a rotating sleeve and a crochet hook; the rotating sleeve is sleeved on the rotating shaft; the number of the rotating sleeves is the same as that of the rotating shafts; the crochet hooks are linearly arranged on the outer wall of the rotating sleeve in the axis direction of the rotating sleeve. The device has the beneficial effects that after the small-focus lung cancer tissue is cut open, the puncture section of the crochet hook is inserted into the periphery of the lung cancer tissue, and then the rotating sleeve is rotated, so that the focus tissue is stretched under the pulling of the arc-shaped limiting structure, the peripheral lung tissue is prevented from extruding the focus, the focus tissue is clear and visible, and the misdiagnosis probability is reduced.

The invention discloses a cross-modal breast image segmentation and classification method based on X-ray and clinical medicine texts. The method comprises the following steps: firstly, constructing a cross-modal breast image segmentation and classification model; performing semi-supervised multi-task learning on the model by using the acquired mammary gland X-ray medical image data set and the corresponding mammary gland clinical text report data set; and realizing breast image segmentation and classification by using the trained model. According to the method, medical images and clinical text information can be deeply fused, pixel-level refined fusion and global context perception are realized through an innovative model architecture, and finally, the segmentation precision and robustness of breast cancer lesions, especially tiny lesions, are remarkably improved.