Abstract: Disclosed is a fundus image quality evaluation method based on multi-source and multi-scale feature fusion, comprising following steps: S1, acquiring multi-source fundus images, labeling the multi-source fundus images with four evaluation dimensions of brightness, blur, contrast and overall image quality, and forming training samples with the fundus image and labeling labels; S2, constructing a fundus image quality evaluation network including a feature extraction module, a fusion module, an attention module and an evaluation module; S3, training the fundus image quality evaluation network by using training samples to obtain a fundus image quality evaluation model; and S4: inputting fundus images to be measured into the fundus image quality evaluation model, and outputting quality evaluation results through calculation. Also provided is a fundus image quality evaluation device based on above method.

Abstract: Various techniques are provided for increasing contrast of gas features in a scene. In one example, a method includes receiving a captured infrared image comprising a gas feature and a scene feature. The captured infrared image comprises a first range of pixel values associated with a first temperature range of the gas feature and the scene feature. The method also includes applying a spatial filter to the captured infrared image to provide a spatially filtered infrared image retaining the gas feature and removing the scene feature. The spatially filtered infrared image comprises a second range of pixel values associated with a second temperature range of the gas feature without the additional scene feature to exhibit increased gas contrast over the captured infrared image. Additional methods and systems are also provided.

Abstract: The present disclosure relates to a method of recommending cosmetics based on a melanin index and a hemoglobin index and a device thereof, and more particularly to a method for implementing quantification and standardization based on hemoglobin and melanin indexes from personal color consulting, which conventionally done offline based on a color index (Lab index) and has high result variability depending on personal experience or knowledge of a consultant and recommending cosmetics based on the quantification and standardization.

Abstract: Techniques are for detecting presence of a problematic cellular entity in a target. In an example, using an analysis model, a fluorescence-based image is analyzed. The analysis model is trained using a number of reference fluorescence-based images for detecting the presence of problematic cellular entities in targets. Based on the analysis, a problematic cellular entity present in the target is detected. To perform the detection, the analysis model is trained to differentiate between the fluorescence in the fluorescence-based image emerging from the problematic cellular entity and the fluorescence in the fluorescence-based image emerging from regions other than the problematic cellular entity.

Abstract: Mechanisms are provided to implement an automated echocardiograph measurement extraction system. The automated echocardiograph measurement extraction system receives medical imaging data comprising one or more medical images and inputs the one or more medical images into a deep learning network. The deep learning network automatically processes the one or more medical images to generate an extracted echocardiograph measurement vector output comprising one or more values for echocardiograph measurements extracted from the one or more medical images. The deep learning network outputs the extracted echocardiograph measurement vector output to a medical image viewer.

Abstract: Embodiments discussed herein facilitate training and/or employing a combined model employing machine learning and deep learning outputs to generate prognoses for treatment of tumors. One example embodiment can extract radiomic features from a tumor and a peri-tumoral region; provide the intra-tumoral and peri-tumoral features to two separate machine learning models; provide the segmented tumor and peri-tumoral region to two separate deep learning models; receive predicted prognoses from each of the machine learning models and each of the deep learning models; provide the predicted prognoses to a combined machine learning model; and receive a combined predicted prognosis for the tumor from the combined machine learning model.

Abstract: A computer-implemented method for generating a rotation-invariant feature descriptor for a location in an image for use in performing descriptor matching in analysing the image, extracts samples according to a descriptor pattern for the location in the image; uses the extracted samples to determine a measure of rotation for the location in the image, the measure of rotation describing an angle between an orientation of the image and a characteristic direction of the image at the location; generating a feature descriptor for the location in the image by determining a set of samples characterising the location in dependence on the determined measure of rotation and the extracted samples; and processes the determined set of samples to generate the feature descriptor for the location in the image.

Abstract: A method of tumor detection and segmentation accepts a first Whole Slide Image (WSI) having a first resolution; creates a corresponding second WSI having a second resolution lower than the first resolution; applies an adaptive thresholding technique to the second WSI to create a background removal mask background; applies the mask to the first WSI to provide a third WSI with extracted patches, characterized by a third resolution, greater than the second resolution and lower than the first resolution; uses a first machine learning system on the third WSI to create a heat map at the third resolution, indicating a subset of the patches likely to include one or more clusters of tumor cells; and uses a second machine learning system on the first WSI and the heat map to segment each patch in a corresponding output image at the first resolution, outlining one or more corresponding clusters.

Abstract: A neural network is trained to segment interferogram images. A first plurality of interferograms are obtained, where each interferograms corresponds to data acquired by an OCT system using a first scan pattern, annotating each of the plurality of interferograms to indicate a tissue structure of a retina, training a neural network using the plurality of interferograms and the annotations, inputting a second plurality of interferograms corresponding to data acquired by an OCT system using a second scan pattern and obtaining an output of the trained neural network indicating the tissue structure of the retina that was scanned using the second scan pattern. The system and methods may instead receive a plurality of A-scans and output a segmented image corresponding to a plurality of locations along an OCT scan pattern.

Abstract: A liveness test method and apparatus is disclosed. A processor implemented liveness test method includes extracting an interest region of an object from a portion of the object in an input image, performing a liveness test on the object using a neural network model-based liveness test model, the liveness test model using image information of the interest region as provided first input image information to the liveness test model and determining liveness based at least on extracted texture information from the information of the interest region by the liveness test model, and indicating a result of the liveness test.

Abstract: A method and a device of extracting a label in a medical image are provided. The method includes: performing an edge detection on the medical image by using an edge detection algorithm, to acquire edge information in the medical image; determining at least one target area defined by the edge information; performing a fitting process on the at least one target area, to obtain a fitting area; and extracting the label in the medical image by selecting, from the fitting areas, at least one target fitting area matching a preset condition, wherein the preset condition is set based on a characteristic of the fitting area.

Abstract: A method for predicting the morphological changes of liver tumor after ablation based on deep learning includes: obtaining a medical image of liver tumor before ablation and a medical image of liver tumor after ablation; preprocessing the medical image of liver tumor before ablation and the medical image of liver tumor after ablation; obtaining a preoperative liver region map, postoperative liver region map, and postoperative liver tumor residual image map; obtaining a transformation matrix by a Coherent Point Drift (CPD) algorithm and obtaining a registration result map according to the transformation matrix; training the network by a random gradient descent method to obtain a liver tumor prediction model; using the liver tumor prediction model to predict the morphological changes of liver tumor after ablation. The method provides the basis for quantitatively evaluating whether the ablation area completely covers the tumor and facilitates the postoperative treatment plan for the patient.

Abstract: Disclosed is an automatic generation method of a fine-labeled digital pathological data set based on hyperspectral imaging, comprising following steps: obtaining reference histological stained slides and double-stained slides based on pathological samples; obtaining two-dimensional color reference whole slide images based on the reference histological stained slides, and obtaining double-stained hyperspectral images based on the double-stained slides; establishing virtual staining models based on the two-dimensional color reference whole slide images and the double-stained hyperspectral images; establishing a segmentation model for automatically generating labeling information based on the double-stained hyperspectral images; and obtaining the fine-labeled digital pathological data set based on the double-stained hyperspectral images and the virtual staining models, the double-stained hyperspectral images and the segmentation model.

Abstract: A method of feature matching in images captured from camera viewpoints uses the epipolar geometry of the viewpoints to define a geometrically-constrained region in a second image corresponding to a first feature in a first image; comparing the local descriptor of the first feature with local descriptors of features in the second image to determine respective measures of similarity; identifying, from the features located in the geometrically-constrained region, (i) a geometric best match and (ii) a geometric next-best match to the first feature; identifying a global best match to the first feature; performing a first comparison of the measures of similarity for the geometric best match and the global best match; performing a second comparison of the measures of similarity for the geometric best match and the geometric next-best match; and, if thresholds are met, selecting the geometric best match feature in the second image.

Abstract: A method for indicating how a cancer patient will respond to a predetermined therapy relies on spatial statistical analysis of classes of cell centers in a digital image of tissue of the cancer patient. The cell centers are detected in the image of stained tissue of the cancer patient. For each cell center, an image patch that includes the cell center is extracted from the image. A feature vector is generated based on each image patch using a convolutional neural network. A class is assigned to each cell center based on the feature vector associated with each cell center. A score is computed for the image of tissue by performing spatial statistical analysis based on classes of the cell centers. The score indicates how the cancer patient will respond to the predetermined therapy. The predetermined therapy is recommended to the patient if the score is larger than a predetermined threshold.

Abstract: Provided is an image analysis apparatus including a hardware processor that analyzes at least one dynamic radiograph formed from a plurality of two-dimensional images acquired by radiographing dynamics of a subject including a trachea and/or a bronchus to measure a feature amount representing a stenotic state of the trachea and/or the bronchus, and estimates the stenotic state of the trachea and/or the bronchus based on a result of the measurement.

Abstract: Methods and systems for characterizing tissue of a subject are disclosed. The method includes retrieving a time series of angiography images of tissue of a subject, defining a plurality of calculation regions, generating a time-intensity curve for each respective calculation region, calculating a rank value for each respective calculation region based on one or more parameters derived from the time-intensity curve; and generating a viewable image in which on the image position of each calculation region an indication is provided of the calculated rank value for that calculation region. Also disclosed are methods and systems for generating first and second time-intensity curves for respective first and second calculation regions, calculating first and second rank values for the respective calculation regions based on first and second pluralities of parameters selected to approximate the respective time-intensity curves, and generating a spatial map of the first and second calculated rank values.

Abstract: A system and method for estimating a pose of an imaging device for one or more images is provided.

Abstract: Various example embodiments pertain to processing images that depict tissue samples using a neural network algorithm. The neural network algorithm includes multiple encoder branches that are copies of each other that share the same parameters. The encoder branches can, accordingly, be referred to as Siamese copies of each other.

Abstract: A mechanism is provided to implement a trained machine learning computer model for determining z-wise lesion connectivity. The mechanism identifies, for a given slice in a three-dimensional medical image, a first lesion in the given slice and a second lesion in an adjacent slice in the three-dimensional medical image. The mechanism determines a first intersect value between the first lesion and the second lesion with respect to the first lesion and determines a second intersect value between the first lesion and the second lesion with respect to the second lesion. The mechanism determines whether the first lesion and the second lesion belong to the same three-dimensional lesion based on the first and second intersect values.