37 results about "Nuclei segmentation" patented technology

Provided herein are systems and computer-implemented methods for quantitative analyses of tissue sections (including, histopathology samples, such as immunohistochemically labeled or H&E stained tissue sections), involving automatic unsupervised segmentation of image(s) of the tissue section(s), measurement of multiple features for individual nuclei within the image(s), clustering of nuclei based on extracted features, and/or analysis of the spatial arrangement and organization of features in the image based on spatial statistics. Also provided are computer-readable media containing instructions to perform operations to carry out such methods. A quantitative image analysis pipeline for tumor purity estimation is also described

A method of segmentation of cell nuclei is described that uses an active contours approach based on a Viterbi search algorithm. Also disclosed is an initialisation technique that maximises the probability of a correct segmentation. The invention also includes a method to classify cell nuclei according to the difficulty of correct segmentation. Nuclei that are difficult to segment can be rejected to minimise the probability of false positives. The probability of false positives can be driven close to 0%.

The invention provides an image segmentation method and device, a neural network model training method and device, a computer readable storage medium and electronic equipment. The image segmentation method comprises steps that a segmentation mask image of a to-be-segmented pathological image is acquired, and the segmentation mask image comprises boundary information of cell nucleuses in the to-be-segmented pathological image; a distance regression image of the to-be-segmented pathological image is obtained according to the segmentation mask image; according to the segmentation mask image and the distance regression image, the binary cell nucleus segmentation image of the to-be-segmented pathology image is acquired. The cell nucleus binary classification segmentation task can be performed by combining the cell nucleus boundary information and the distance information, the cell nucleus segmentation precision is improved, and small cell nucleuses, adhesion and overlapped cell nucleuses can be better segmented; in addition, the cell nucleuses from different organs are greatly different in appearance, form, shape, color and density, and the technical scheme provided by the embodiment ofthe invention can have good generalization ability.

The invention belongs to the field of medical image processing and especially relates to a nucleus segmentation method based on white blood cell detection. The method comprises the following steps: to begin with, summarizing the size of diameter of each kind of white blood cell to set a plurality of sizes of sliding windows; then, selecting a candidate window for each sliding window according to gradient information of each white blood cell picture; calculating characteristic value of each candidate window according to the color contrast information, boundary density contrast information and gradient contrast information of the white blood cell pictures, and inputting the characteristic values to an epsilon-SVR model, and selecting a final detection window; intercepting the region of the final detection window of each white blood cell picture as a final positioning subgraph; and then, selecting threshold segmentation nucleus through a subgraph gray statistical chart polynomial curve fitting method, and then, carrying out hole filling and isolated point removal through morphologic processing to enable the segmentation result to be more accurate.

The embodiment of the invention discloses a medical image segmentation method and device, equipment and a storage medium. The method comprises the following steps: acquiring a tissue image of a detected object, and extracting a preliminary segmentation image of a cell nucleus from the tissue image based on a preset image segmentation algorithm; and inputting the tissue image and the preliminary segmentation image into a trained cell nucleus segmentation model, and obtaining a target segmentation image of the cell nucleus according to an output result of the cell nucleus segmentation model. According to the embodiment of the invention, a traditional image segmentation algorithm and a deep learning algorithm are combined; the good segmentation performance of the deep learning algorithm can be brought into full play, the requirement of a deep learning algorithm for manually annotated perfect data can be greatly relieved, so that good segmentation performance can still be achieved under the condition that only a small amount of manually annotated perfect data exists, and the effect of accurately segmenting the cell nucleus image from the medical image with low labor cost and time costis achieved.

The invention provides a cell segmentation method and device and a computer readable storage medium. The method comprises the steps of obtaining an original image; processing the original image to obtain an image Igram; processing the image Igray by adopting a local threshold method to obtain an image ICLT; processing the image Igray by adopting a LoG algorithm to obtain an image ILoG; performingfusion operation on the image ICLT and the image ILoG to obtain an image Ifuse; and carrying out cell segmentation on the image Ifuse to obtain a cell nucleus segmentation image. According to the invention, the color image is converted into the grayscale image Igray and then is segmented, so that the problems of false cell nucleus detection and missing detection caused by non-uniform color distribution in the original image can be avoided. According to the method, the cell nucleus pixel points are obtained by using a mode of fusing a constrained local threshold method and a LoG algorithm, so that the problems of false cell nucleus detection and missing detection caused by non-uniform intensity distribution in an original image can be avoided.

The invention discloses a high-resolution microscopic endoscope image cell nucleus segmentation method based on deep learning, and the method comprises the steps: obtaining an original endoscope image, carrying out the pixel-level marking of a cell nucleus of the endoscope image, and obtaining a mask image of the cell nucleus, dividing the marked mask image and the endoscope image into a trainingset and a verification set; constructing a hierarchical multi-scale attention mechanism high-resolution convolutional neural network model; inputting a training data set into the convolutional neuralnetwork for iterative training after data enhancement, and judging whether iterative training is completed by using a verification set; when it is judged that iterative training is completed, inputting the original endoscope image into the trained convolutional neural network, outputting the prediction probability that all pixels in the endoscope image belong to cell nucleuses, obtaining a segmentation result of the cell nucleuses, and realizing accurate segmentation of the input image.

The invention discloses a cell nucleus segmentation method based on attention learning, and relates to the cell nucleus segmentation problem in intelligent pathological diagnosis. The intelligent pathological diagnosis technology utilizes a deep learning technology to segment and identify abnormal cells in a cell image. However, there are few models for cell nucleus segmentation in cell images. The following problems exist: (1) the problems of cell nucleus overlapping and unobvious boundary are not considered, so that the segmentation precision is low; and (2) the contextual information of thecell nucleus edge is not considered, so that under-segmentation or over-segmentation is caused, and the subsequent classification result is influenced. Therefore, the invention provides a cell nucleus segmentation method based on attention learning. Experiments show that the model can effectively solve the problem of segmentation of overlapped cells, and the problems of under-segmentation or over-segmentation of unclear boundaries and the like. The invention is applied to cell nucleus segmentation in intelligent pathological diagnosis.

The invention discloses an intelligent cervical cancer cell detection method based on deep learning. The invention relates to classification of cell nucleuses by a deep learning method. The inventionaims to solve the problems of low cancer cell detection accuracy, long consumed time and the like in the existing traditional diagnosis mode. In order to solve the problem, the invention provides an intelligent cervical cancer cell screening method based on deep learning. The method comprises the following specific steps: 1, preparing data; 2, carrying out cell nucleus segmentation; 3, carrying out cell nucleus classification; and 4, screening cancer cells. In the cell nucleus classification part, data expansion and category subdivision are carried out on the data by using an active learning method; on the model, ResNeSt is taken as a basic model, doctor diagnosis experience is introduced, and a more accurate model is trained under the combined action of diagnosis index extraction. Experiments show that the accuracy of the cell nucleus classification method is higher than that of an original model, and in addition, the invention further provides a more effective data preparation methodfor expanding data and subdividing categories. The method is applied to the field of medical image classification.

The invention provides a cell nucleus segmentation method and apparatus, an electronic device and a computer readable storage medium, and relates to the technical field of image processing, the methodcomprises the steps of obtaining a to-be-segmented original cell image; determining a to-be-segmented region in the original cell image; and carrying out cell nucleus segmentation on the to-be-segmented region by adopting a circle detection algorithm to obtain a target segmentation image. Compared with a mode of carrying out cell nucleus segmentation by using a watershed method in the prior art,the mode of carrying out cell nucleus segmentation on the to-be-segmented region in the original cell image by using the circle detection algorithm has the advantages that the segmentation accuracy ishigher, and the actual application requirements can be alleviated to a certain extent.

The invention discloses a cancer auxiliary analysis system and device based on an HE staining pathological image, and belongs to the technical field of medical imaging. The problem that an existing cell segmentation neural network model cannot accurately segment cytoplasm is solved. The system comprises a staining slice image acquisition module for acquiring a staining slice image of HE staining, a cell nucleus segmentation module for calling a cell nucleus segmentation network model to perform cell nucleus segmentation on an image block, and a cell nucleus masking module for masking a cell nucleus, a cytoplasm segmentation module used for calling a cytoplasm segmentation network model to carry out cytoplasm segmentation on the image masked by the cell nucleus masking module, a cell overall unit determination module used for mapping the segmentation results of the cytoplasm and the cell nucleus segmentation module into the same image block, and a cancer assisted analysis module for providing assisted analysis. The method is mainly used for providing auxiliary analysis for cancer recognition.

The invention provides a tissue pathology image cell nucleus segmentation method and system based on polar coordinate representation; wherein the method comprises the steps: extracting the image features of a to-be-segmented image block, and carrying out the modeling based on a polar coordinate system, and obtaining the classification data, central point coordinates and ray length of the image features; according to the classification data, the center point coordinates and the ray length, carrying out cell nucleus segmentation of the to-be-segmented image block to acquire the image block containing the center position coordinates and the edges of each cell nucleus. On the aspect of the processing effect, the invention provides the polar coordinate representation-based automatic cell nucleus segmentation method of the histopathological image for the first time. The method based on polar coordinates is very suitable for segmentation of circular objects; therefore, the pathological imagecell nucleus can be automatically segmented, the cell nucleus center and the ray length pointing to the target edge with the cell nucleus center as the starting point are obtained, and positioning andfine segmentation of the cell nucleus are achieved.

The invention discloses a cell nucleus segmentation method and system of fluorescence in situ hybridization images. The method realizes the recognition and segmentation of cell nucleus in FISH imagesthrough pretreatment, binarization, hole filling, morphological operation, secondary watershed and other processing processes. The system comprises a FISH image receiving module, an image preprocessing module, a cell nucleus detecting module and a report generating module. When the FISH images of breast cancer, gastric cancer and other cancer cells are segmented to obtain the cell nucleus region,the invention can reduce the image fluorescence scattering effect through multiple pretreatments, can effectively improve the over-segmentation problem easily caused by the direct watershed, and thusimprove the segmentation quality of the FISH image recognition cell.

The invention discloses a semi-supervised learning method special for performing cell nucleus segmentation on a histopathological image dyed by hematoxylin eosin. According to the cell nucleus segmentation method provided by the invention, according to the characteristics of the histopathological image and cell nucleus segmentation, the two dyes of hematoxylin and eosin in the histopathological image are separated by adopting non-negative matrix factorization with sparse constraint, and then the eosin dye in the histopathological image is replaced by the eosin dye in other histopathological images, so that the segmentation efficiency of the cell nucleus is improved. Therefore, a group of positive example samples can be prepared, and the positive example samples have the same hematoxylin staining agent, so that the positive example samples have interpretable invariance. And inputting the multiple groups of positive example samples into an encoder, and outputting a corresponding embedded representation vector by the encoder. And constraining the model by adopting a contrast learning loss function, so that the model can learn invariance in a positive example sample, namely the hematoxylin staining agent. The hematoxylin stain can stain the cell nucleus and other nucleic acid-rich parts, such as ribosome, so that the hematoxylin stain and the cell nucleus have relatively high correlation. When the model learns the characteristics of the hematoxylin stain, the characteristics accord with the characteristics of a cell nucleus segmentation task, so that the training of the downstream cell nucleus segmentation task is facilitated. As positive example sample construction and pre-training do not need labels, a large amount of unlabeled data can be utilized for training in the mode. And finally, the pre-trained encoder is added into the segmentation model, and fine adjustment is performed on a very small amount of labeled data, so that an effect better than supervised learning on a small amount of samples can be achieved. Therefore, the demand of annotation data is also reduced, and the labor cost is greatly reduced.

The invention discloses a fine-grained cancer subtype classification method. The method comprises the following steps: step 1, obtaining a cell nucleus segmentation and classification result in a pathological picture; step 2, extracting an instance patch according to the cell nucleus segmentation and classification prediction result; step 3, performing instance patch convolution feature extraction; and step 4, generating picture-level features of the used instance patch through a Transform model to complete classification of cancer subtypes. The method can assist pathologists to classify cancer subtypes, and the working efficiency of the doctors is improved.

The invention provides a cell nucleus segmentation method and device for a pathological image, and the method comprises the steps: inputting a to-be-segmented pathological image into a cell segmentation model, and obtaining a cell segmentation mask outputted by the cell segmentation model; the cell segmentation mask comprises a semantic segmentation mask used for identifying each cell nucleus region, a normal edge mask used for identifying each cell nucleus boundary and a nucleus edge mask used for identifying a plurality of adhered cell nucleus edges; and determining a cell nucleus segmentation result of the pathological image based on the semantic segmentation mask, the normal edge mask and the nucleus edge mask. The method can greatly improve the precision of the cell nucleus segmentation result.

The invention discloses an osteosarcoma recurrence risk prediction model based on tissue morphological analysis. The method belongs to the field of machine learning and image processing. The method comprises the following specific steps: 1, segmentation of multiple types of tissues and segmentation of cell nucleuses in a tumor region; 2, feature extraction and feature selection; and 3, classifierconstruction. The invention provides a quantitative computer-assisted osteosarcoma recurrence risk prediction model established by using cell nucleus characteristics in a tumor region. Experimental results show that the image features derived from tumor cell nucleuses can be used as new markers independent of traditional standard clinical diagnosis features for prognosis analysis, so that patientscan be helped to carry out personalized treatment, and the development of precise oncology is promoted.

The invention discloses a cell nucleus segmentation method, system and device and a cancer auxiliary analysis system and device based on a pathological image, and belongs to the technical field of medical images. The invention aims to solve the problem that the edge segmentation accuracy needs to be improved in the process of segmenting a feature map by a neural network at present. The cell nucleus segmentation method provided by the invention comprises the following steps: aiming at a to-be-detected sample, preparing a slice and dyeing to obtain a slice dyeing image; performing image block segmentation on the section dyeing image; and then performing nucleus segmentation on an image block corresponding to the section dyeing image of the to-be-detected sample by using the nucleus segmentation network model to obtain a nucleus boundary segmentation image. According to the cancer auxiliary analysis system based on the pathological image, a cancer auxiliary analysis module is additionally arranged on the basis of cell nucleus segmentation, cancerous cells are recognized and classified according to the segmentation result of the image segmentation module based on the expert database, and therefore cancer auxiliary analysis is achieved. The method is mainly used for nucleus segmentation and cancer auxiliary analysis.