49 results about "Dermatoscopes" patented technology

The invention provides a dermatoscope image segmentation method, and relates to the technical field of medical image processing, and the method comprises the steps: inputting a to-be-segmented image into an improved encoder module, carrying out the multi-scale feature information extraction, and obtaining a deep advanced semantic feature map; inputting the deep advanced semantic feature map into the bridging module, and performing context information modeling and cross-scale feature fusion to obtain a fused feature map; inputting the fused feature map into the decoder module, and carrying out layer-by-layer up-sampling and feature integration to obtain a reconstructed feature map; inputting the reconstructed feature map into the boundary sensing double-branch module, and performing collaborative feature processing; wherein the main branch outputs a binary segmentation mask, and the auxiliary branch is used for measuring a sign distance function diagram to provide boundary perception supervision; obtaining a binary segmentation mask output by the main branch as a segmentation result; compared with an existing model, the optimized dermatoscope image segmentation model is higher in recognition accuracy.

The invention discloses an efficient medical image segmentation method considering global modeling and local enhancement, and relates to the technical field of image segmentation. According to the method, adaptive space shift operation is executed in different directions through the AS-MLP module, the long-range dependence modeling capability is effectively enhanced, and the recognition performance of a complex structure focus is improved; the channel and space double attention mechanism and multi-scale convolution of the LMCAM module are combined, so that fine-grained feature extraction is realized, and the segmentation precision of the lesion boundary and the small-scale structure is remarkably improved; a lightweight network design is adopted, the calculation complexity is low, the reasoning speed is high, and the method is suitable for resource-limited clinical terminals and real-time diagnosis application; besides, the method has good cross-modal adaptability, can keep stable and efficient segmentation performance in various medical imaging modalities such as CT, MRI, ultrasound and dermatoscope, and has wide application value.

The invention provides a skin lesion auxiliary detection method and system based on multi-modal information fusion, and belongs to the technical field of computer vision and application. The method comprises the following steps: firstly, acquiring and constructing bimodal data including a dermatoscope image and a text of a patient, and preprocessing the data; secondly, respectively adopting an improved convolutional neural network and a text coding model, and extracting deep and high-dimensional features from the skin disease image and the self-described text of the patient; afterwards, an improved symmetric gating cross attention fusion module is constructed, fine-grained interaction and alignment between visual and text features are realized through a bidirectional cross attention mechanism, and a gating unit is utilized to adaptively adjust a fusion weight; and finally, inputting the deeply fused multi-modal feature vector into a classifier to obtain a confidence score of each skin lesion type so as to provide reference for dermatologists. According to the method, the diagnosis accuracy, robustness and interpretability are synergistically improved, and the method has great clinical application value.

The invention belongs to the technical field of medical image and big data analysis, and provides an automatic skin disease diagnosis method and system based on a multi-modal big language model, and the method comprises the steps: obtaining a dermatoscope diagnosis report of a historical patient, and extracting an original image and a diagnosis text in the dermatoscope diagnosis report; constructing a multi-modal large language model, wherein the multi-modal large language model comprises a visual encoder, a linear projection layer and a large language model; training the multi-modal large language model; obtaining an external medical knowledge base, extracting text information in the external medical knowledge base, and constructing a vector database based on the text information; in response to an input image and a query text of a user, N vectors with the highest similarity are retrieved in the vector database according to the query text, the retrieved vectors are decoded into a text, and the text and the query text jointly construct complete prompt word input; and based on the input image and cue word input, obtaining a final dermatoscope diagnosis report by using the trained multi-modal large language model. The diagnosis report is automatically generated.

This application provides a method and apparatus for visualizing dermoscopic image features based on orientation awareness, comprising: processing a dermoscopic image through a backbone network of a segmentation network, wherein the spatial resolution of multiple hierarchical modules is progressively reduced and the number of channels is progressively increased; decoupling the orientation features of the dermoscopic image and hierarchically propagating the orientation prior through orientation-aware base blocks; extracting directional latent priors containing global response information in multiple directions from deep orientation features; fusing orientation features under the guidance of the directional latent priors through a decoder to obtain a segmentation feature map; progressively decoupling and enhancing multi-scale orientation information during feature extraction, and extracting compact directional latent priors from deep features; and using the orientation priors to guide dual attention computation within a local window, enabling the model to achieve a balance between structural consistency and detail accuracy, thereby achieving robust recovery of the continuous boundaries and topological integrity of lesions.

The invention belongs to the technical field of dermatoscope image analysis, and relates to a pathological space diagram modeling system and method for skin lesion intelligent diagnosis, the system comprises four parallel deployment visual representation learning and topological structure modeling, the modeling comprises a saliency map embedding layer and a plurality of pathological space diagram blocks, the pathological space graph block is sequentially composed of a pathological feature point digger, a layered graph connector and a geometric distribution learner; the method comprises the following steps: constructing a multi-scale saliency block embedding layer to extract initial visual representation, introducing a pathological feature point digger to realize intelligent identification of sparse key pathological points, constructing a hierarchical graph connector to organize a multi-scale topological relation, realizing hierarchical topological reasoning and vector distillation of a geometric distribution learner, and constructing a geometric distribution learner. Defining an attention fusion mechanism and a classification head to realize topological guidance diagnosis; according to the method, the defects in the aspects of structural modeling, calculation distribution, graph representation and the like in the prior art are overcome, and the diagnosis accuracy, robustness and clinical interpretability are remarkably improved.

The invention relates to the technical field of medical instruments, and discloses a rotary dermatoscope magnification adjusting ring which comprises an adjusting ring body, a lens body and a connecting ring, the connecting ring sleeves the outer side of the lens body and is in sliding connection with the lens body, and the adjusting ring body rotatably sleeves the outer side of the connecting ring; a plurality of damping grooves are formed in the inner wall of the adjusting ring body in the circumferential direction at intervals, a plurality of connecting pieces are symmetrically and fixedly connected to the surface of the connecting ring, an arc-shaped spring piece is fixedly connected to the surface of each connecting piece, and two damping columns are fixedly connected to the two sides of each arc-shaped spring piece; the damping column elastically abuts against an inner cavity of the damping groove through the elastic acting force of the arc-shaped spring piece to form a gear damping structure, through cooperation of the damping column and the triangular damping groove and the elastic acting force of the arc-shaped spring piece, clear'click 'sound and the sense of lagging are generated in the adjusting process, a clear gear switching signal is provided for a doctor, and the safety of the doctor is improved. The defect that the adjusting ring lacks positioning sensing in the prior art is overcome.

The application discloses a skin cancer multi-modal case retrieval method and device based on global-local alignment, equipment, medium and product, relates to the field of multi-modal information fusion, and comprises the following steps: constructing a case database containing dermoscope images and diagnosis labels and a composite query; extracting multi-level visual features of the query image by using a Swin Transformer; encoding the query text by using BERT, and interacting and fusing the multi-level visual features and the text features by using a cross-modal Transformer to obtain a query feature of the text-image fusion; respectively calculating global similarity and local similarity of the query feature of the text-image fusion and target image features; weighting and fusing the global and local similarity to obtain a final similarity score, and performing sorting retrieval and label consistency verification on the database based on the score. The application breaks through the limitation of traditional single-modal retrieval, realizes fine-grained semantic alignment, and significantly improves the accuracy of skin cancer case retrieval and the value of clinical decision support.

Examination device for optical medical examinations, in particular dermatoscope, with optics (14) and a lighting device (16), wherein the optics (14) has an optical axis (A) and defines an examination area (U), wherein the lighting device (16) has at least one first light source (30) and at least one second light source (32), wherein the at least one first light source (30) and the at least one second light source (32) are arranged such that they jointly illuminate the area of ​​investigation (U), wherein the at least one first light source (30) is configured to emit white light with a color temperature of 5000K and the at least one second light source (32) is configured to emit light that is bluer than the white light emitted by the at least one first light source (30), wherein the light emitted by the at least one second light source (32) has a wavelength of 400 nm to 550 nm, and wherein the examination device (10) is set up such that that when the lighting device (16) is activated, at least one first light source (30) is always switched on, that at least one second light source (32) can be switched on in addition to at least one first light source (30), that the luminous intensity of the light emitted by the at least one second light source (32) is variable, and that the color temperature of the light emitted by the lighting device (16) can be set within a color temperature range between 5000K and 12000K.

This invention discloses a cross-modal medical image segmentation method based on pathological anchoring. Using pathological priors as semantic anchors, this method integrates cross-scale collaboration and uncertainty perception in a closed-loop prompting refinement process. This achieves interpretable suppression and robust learning of cross-modal domain shifts, forming a closed-loop path from structure to semantics to noise control. Under strong cross-modal zero-sample settings such as multi-source ultrasound cross-domain segmentation and color dermoscopy, stable and consistent improvements are achieved, manifested in more precise boundaries, better calibration, and more robust generalization. This provides technical support for disease image diagnosis.

The invention discloses a radiodermatitis dermatoscope imageomics prediction method and system and an application program, the system comprises a hardware module and a software module, the hardware module comprises a dermatoscope image acquisition device, and the dermatoscope image acquisition device is used for acquiring a high-resolution dermatoscope image of a radiotherapy treatment area of a patient; the software module is integrated in an application program of a mobile terminal and comprises a radiomics feature extraction module used for extracting quantitative features from dermatoscope images; the parameter calculation module is used for calculating radiomics parameters associated with the skin microstructure change; and the risk prediction module is used for constructing an interpretable machine learning risk prediction model and inputting the patient feature data and the radiomics parameters into the risk prediction model to generate a personalized radiodermatitis risk score for the patient. According to the invention, accurate prediction of acute radiodermatitis is realized.

The invention relates to the technical field of medical image processing, and particularly discloses a dermatosis auxiliary diagnosis method and system based on multi-mode optical image fusion, and the method comprises the steps: carrying out the semantic segmentation of an obtained dermatoscope image of a target skin region, and recognizing at least one target feature region; based on the semantic category and position information of the target feature region, screening out at least one associated feature region from an obtained RCM image set of a target skin region; extracting a first feature vector in the target feature region, and extracting a second feature vector in the associated feature region; and fusing the first feature vector and the second feature vector, and performing disease classification on the fused feature vector based on a trained classification model to obtain a skin disease classification result of the target skin area. According to the method, the corresponding associated region can be screened out from the RCM image set according to the semantic segmentation result of the dermatoscope, and the efficiency and reliability of multi-modal data integration are remarkably improved.

A diagnostic imaging system for analysing skin tissue image data received from a remote user for evaluating the image data for screening for a skin disease or skin condition, the system comprising: a manipulator arm and slider system of a robot, the manipulator arm moving on the slider along a patient body and comprising a plurality of elements interconnected to each other by a plurality of joints whereby each element is rotatable relative to an adjoining element of the manipulator arm; a non-invasive dermatoscope module comprising one or more dermatoscopes arranged to acquire tissue image data, the dermatoscope module being coupled to one of the elements of the manipulator arm by a coupling arrangement to allow movement of the dermatoscope module relative to said one of the elements of the manipulator arm; and a controller coupled to the manipulator arm and a coupling arrangement to employ the manipulator arm to move the dermatoscope module relative to the subject's anatomy; a server in communication with a distribution network for receiving tissue image data from the dermatoscope module, the tissue image data; a processor for executing a learning machine algorithm, wherein the algorithm is trained using image data sets having known outcomes for a plurality of skin diseases or skin conditions, the processor being operable for receiving image data from the dermatoscope module and processing the image data to produce processed image data and introducing the processed image data into the learning machine to provide an output indication of whether the tissue image data is associated with a skin condition or skin disease.

The invention relates to the technical field of medical apparatus and instruments, and discloses a foldable telescopic portable dermatoscope main body structure which comprises a holding seat, the upper surface of the holding seat is fixedly connected with a dermatoscope body, a pressing block drives an ejector rod to ascend, a telescopic spring is extruded, then a movable block and a mounting seat are driven to move, and the dermatoscope body is fixed to the holding seat. The length of the holding base can be flexibly adjusted according to the holding habit of a user, more comfortable and stable holding experience is provided, the telescopic block can be stored in the holding base when the dermatoscope is not used, the size of the dermatoscope is greatly reduced, the dermatoscope is more convenient to carry, and the dermatoscope is more convenient to use. The dermatoscope can be easily placed in a bag or other narrow spaces, carrying inconvenience and possible damage caused by the too large size are avoided, the sealing shell is connected with the dermatoscope body, the protective pad on the outer surface and the buffer plate on the inner wall form inner and outer double-layer protection, dust and impurities are prevented from being attached, and the definition of the dermatoscope is ensured.

The invention relates to the technical field of medical instruments, and discloses a focal length-adjustable portable dermatoscope device, which comprises a dermatoscope body, the bottom of the dermatoscope body is connected with a focal length adjusting structure through a detachable connecting structure; the focal length adjusting structure comprises an inner ring connected to the bottom of the dermatoscope body through a detachable connecting structure, the surface of the inner ring is rotationally connected with an adjusting ring, an inner cavity of the inner ring is slidably connected with a sliding ring, axial sliding grooves are symmetrically formed in the surface of the inner ring in a penetrating mode, and spiral guide grooves are symmetrically formed in the inner wall of the adjusting ring. A connecting pin is arranged at the intersection of the axial sliding groove and the spiral guide groove in a penetrating mode, one end of the connecting pin is fixedly connected with the surface of the sliding ring, the other end of the connecting pin extends into the spiral guide groove, and through transmission fit of the adjusting ring and the spiral guide groove, rotary motion is converted into axial displacement of the sliding ring; the relative distance between the lens group structure and the dermatoscope body can be flexibly changed, and the focal length can be continuously adjusted.

The invention relates to the technical field of medical images, and particularly discloses a skin cancer artificial intelligence diagnosis system based on medical images, comprising: an image preprocessing module for acquiring and preprocessing a dermatoscope single-mode image; the feature extraction and compensation module is used for performing visual feature extraction on the preprocessed image and performing skin color deviation compensation by adopting a skin color perception attention mechanism; the classification and visualization module is used for performing classification based on the corrected features, generating a focus heat map and calculating a benign and malignant classification probability; the report generation module is used for outputting a structured report containing the result, a skin color sensitive channel attention layer (SACA) and skin color balance loss joint mechanism is embedded in the dermatoscope single-mode network, and the misdiagnosis rate is reduced; under the condition that the system only uses dermatoscope images, the overall accuracy of the HAM10000 data set is improved; and calibrating a curve according to the skin color classification of the current case, the baseline misdiagnosis rate of the corresponding crowd and the skin color specific confidence coefficient predicted by the model.

This invention discloses a dermoscopic hair loss classification method, device, and medium using expert-guided attention and multi-model fusion. The method includes: acquiring a dermoscopic image to be detected and performing standardized preprocessing; inputting the preprocessed image into a pre-constructed hair loss subtype auxiliary classification model, wherein the hair loss subtype auxiliary classification model outputs the classification result and its confidence score. This invention introduces an expert-guided attention optimization mechanism and confidence score calibration under image-level weak supervision to improve the accuracy of hair loss subtype model classification.

1. The name of the design product: dermatoscope. 2. The use of the design product: mainly used for replacing different lenses to observe local skin and assist in diagnosis. 3. The design points of the design product: in shape. 4. The picture or photo that best indicates the design points: perspective drawing.

The utility model discloses a dermatoscope with a protective structure, which belongs to the technical field of dermatoscopes and comprises a lens barrel, an objective lens is fixedly mounted in the lens barrel close to a first end of the lens barrel, an ocular lens is fixedly mounted in the lens barrel close to a second end of the lens barrel, and a protective sleeve is detachably mounted on the lens barrel. The protecting sleeve is provided with an annular groove matched with the first end of the lens cone, so that the section of the protecting sleeve is J-shaped, and the first end of the lens cone can be inserted into the annular groove. According to the dermatoscope, the protective sleeve is arranged, so that the work of disinfecting and wiping after the dermatoscope is used is relieved.

The utility model relates to a performance real shooting detection device of a dermatoscope lens. The device comprises a detection workbench. A manual lifting platform is arranged on the detection workbench; an aluminum profile underframe is arranged on the manual lifting platform; a camera is fixed on the aluminum profile underframe; a detection adapter ring is screwed in the interface C on the camera in a threaded manner; a detection focusing ring is screwed into the detection adapter ring; a dermatoscope lens to be detected is placed on the detection focusing ring; the camera is connected with a display through a data transmission line; the display is arranged on the detection workbench; a lifting frame is arranged on the manual lifting platform; a vertical adjusting screw rod is arranged on the lifting frame; an extending bracket is arranged on the adjusting screw rod; and the lower surface of the extending bracket is provided with a back light opposite plate. Compared with the prior art, the device has the advantages of good detection effect, high detection efficiency and the like.

The invention discloses a medical image segmentation method based on a double-branch structure and multi-frequency multi-scale feature fusion, and belongs to the field of medical image segmentation, and the method comprises the steps: carrying out the parallel extraction of the multi-scale features of an input medical image through a convolutional neural network and a Transform, and making full use of local details and global context information; then, a multi-frequency feature fusion module is introduced, discrete cosine transform is utilized to decompose frequency domain components, and cross-modal feature enhancement and complementary fusion are realized through multi-view pooling and a space attention mechanism; and finally, multi-level features are aggregated through a pyramid multi-scale fusion module, feature representation is gradually optimized by means of a receptive field enhancement module and a gating decoding structure, and an accurate segmentation mask is generated. The method shows excellent segmentation performance on dermatoscope, CT and other medical images, can effectively process focus areas with fuzzy boundaries and changeable forms, and has good generalization and clinical applicability.

This application provides a model training and image classification method, apparatus, electronic device, and storage medium. The method includes: acquiring multiple training samples, each training sample including dermoscopic images of a target part of a target object and metadata of the dermoscopic images; reconstructing the dermoscopic images using a base model to obtain visual feature vectors and reconstructed images corresponding to the dermoscopic images; training the base model using the multiple dermoscopic images and their corresponding reconstructed images to obtain a temporary model; determining the target feature vector corresponding to the dermoscopic image in each training sample based on the metadata and visual feature vectors; inputting the target feature vectors corresponding to the multiple dermoscopic images into the temporary model to obtain multiple enhanced images; and training the temporary model using the multiple training samples and the multiple enhanced images to obtain a classification model. This application can improve the accuracy of image classification results.

The invention provides a deep learning-based scalp health condition detection method and system, and relates to the technical field of image data processing, and the method comprises the steps: obtaining dermatoscope images and optical coherence tomography data of different scalp regions; constructing a voxel model with epidermis morphological characteristics and subcutaneous tissue scattering characteristics according to optical coherence tomography data; performing multi-layer convolution and pooling operation on the voxel model by using a pre-trained convolutional neural network to extract a high-dimensional feature vector; mapping the high-dimensional feature vector to a low-dimensional space by using a manifold learning algorithm, and calculating a target distance; on the basis of the target distance, the scalp area where the tissue type is the scar lesion type is marked as the irreversible alopecia area, so that detection of the scalp health condition is completed, subcutaneous hair follicle structure disappearance and fibrosis scar tissue are accurately distinguished under the non-invasive condition, and the detection accuracy is improved. The accuracy of scar alopecia judgment and the reliability of irreversible alopecia area recognition are improved.

This invention provides a dermoscopic image segmentation method based on edge feature enhancement and depth supervision, belonging to the field of medical image segmentation technology. It solves the technical problems of insufficient edge feature representation and difficulty in balancing segmentation accuracy and robustness in traditional dermoscopic image segmentation. The technical solution includes the following steps: S1, preprocessing and expanding the dermoscopic image training samples; S2, constructing a dermoscopic image segmentation network based on edge feature enhancement and depth supervision; S3, training the network and outputting the segmentation result image. This invention significantly enhances the edge feature representation capability of dermoscopic images, improves the edge accuracy and overall precision of the segmentation results, and has stronger robustness for clinical applications.

The application discloses a skin lesion image segmentation method based on feature interaction fusion, and relates to the technical field of medical image segmentation. The method comprises the following steps: firstly, performing size normalization and noise suppression on a dermoscope image, and expanding a data set through random rotation and flipping operations; then inputting training data into a network model for training to obtain model weights after training; finally, inputting a skin lesion image to be segmented into the network model after training to realize automatic segmentation of a skin lesion area. The application designs a focused feature interaction fusion module to realize deep complementary fusion of CNN and Transformer features; a multi-scale context attention module is constructed to realize multi-scale accurate feature extraction by extracting spatial information from local to global scales while maintaining computational efficiency. The application aims to solve the problems of complex shape, variable size and low contrast of the lesion area of a skin lesion image, and improve the segmentation accuracy and robustness of the skin lesion image.

The invention discloses a large model skin disease diagnosis system based on two-stage interpretable semantic enhancement, and relates to the technical field of artificial intelligence, and the system comprises the steps: obtaining a dermatoscope image marked with a first concept label, a first concept interpretation text, a first disease label and a first disease interpretation text; training a medical multi-modal large model by using the dermatoscope image, the first concept label and the first concept interpretation text to obtain a concept recognition model; training a medical multi-modal large model by using the dermatoscope image, the first disease label and the first disease interpretation text to obtain a disease diagnosis model; inputting the dermatoscope image and the concept problem text into a concept recognition model to obtain a second concept label and a corresponding second concept interpretation text; and inputting the dermatoscope image, the second concept label, the second concept interpretation text and the disease problem text into a disease diagnosis model to obtain a second disease label and a corresponding second disease interpretation text. The diagnosis accuracy of skin diseases can be improved.

The application relates to a split portable dermatoscope and Wood lamp combined device, which comprises a receiving plate, a U-shaped plate one rotatably connected to the receiving plate, a U-shaped plate two rotatably connected to the U-shaped plate one, a U-shaped plate three rotatably connected to the U-shaped plate two, a long plate rotatably connected to the U-shaped plate three, a sliding sleeve fixedly connected to the outside of the long plate, a threaded sleeve slidably connected to the inside of the sliding sleeve, a fixing sleeve fixedly connected to the outside of the threaded sleeve, a limiting groove formed in the inside of the fixing sleeve, and an abutting rod slidably connected to the inside of the limiting groove. Through the elastic clamping cooperation of the clamping plate and the spring two, the detection mechanism and the fixing sleeve are quickly disassembled, meanwhile, the detection mechanism is driven to be horizontally fine-adjusted through the transmission cooperation of the threaded rod and the threaded sleeve, the quick switching of the detection mode and the accurate positioning of the detection position are realized, the preparation process before detection is simplified, the detection mechanism can be ensured to be aligned with the area to be detected, and the efficiency and accuracy of skin detection are improved.

The invention discloses a melanoma screening method and device based on resource-saving deep learning, and the method comprises the following steps: S1, segmenting a dermatoscope image, outputting a binary mask, and repairing a hair region labeled in the binary mask through employing a classical internal drawing technology realized by OpenCV, so as to restore an original image; s2, performing pathological segmentation on the restored original image; and S3, classifying the dermatoscope images after pathological segmentation. According to the method, the calculation and storage overhead is remarkably reduced while high precision is kept, and the method has extremely high edge device deployment potential.