A method is described for determining a survival prognosis of a patient suffering from a breast tumor, using processing carried out by electronic processing and/or calculation means. The method first comprises step (a) of defining a biological network representative of a particular biological process associated with the breast tumor. The biological network comprises a plurality of nodes, a set of directional relationships between these nodes and a set of genes associated with these nodes. The method also includes step (b) of accessing a data set related to the patient, comprising gene expressions in a biological sample of the tumor isolated from the patient; and step (c) of calculating a continuous expression value for the aforesaid nodes of the biological network. If the node is associated with only one gene and it is found that the gene is present in the biological sample, the continuous expression value of the node is calculated as the expression of the associated gene detected in the biological sample. If the node is associated with multiple genes, and it is found that at least one of the aforesaid genes is present in the biological sample, the continuous expression value of the node is calculated based on the expressions of the associated genes, present in the biological sample. If the node is not associated with any gene, or the associated gene is not found in the biological sample, the node is marked as a node not associated with a continuous expression value. The method then comprises the following steps, carried out by the electronic processing and/or calculation means: (d) binarizing the data set of continuous expression values calculated for each node of the biological network to which a continuous expression value is associated, based on a comparison of the continuous expression value with a respective threshold, to thus obtain a first binarized data set of the nodes, obtained based on the detections made; (e) calculating an aggressiveness score based on the aforesaid first binarized data set of the nodes; and finally (f) determining a survival prognosis result based on the aforesaid aggressiveness score calculated.

A method for generating a synthetic dataset from an original dataset includes encoding categorical features of the original dataset, embedding the encoded dataset in a low-dimensional space, selecting a seed record from the embedded dataset, identifying a plurality of nearest neighbor records to the seed record, generating a new record by randomly selecting features from the plurality of nearest neighbor records, and concatenating the new record into the synthetic dataset. For a synthetic dataset that contains N records, which may be the same as or different from the number of records in the original dataset, the selecting, identifying, generating, and concatenating operations operate a total of N times on the records in the embedded dataset.

A method for generating a synthetic dataset from an original dataset includes encoding categorical features of the original dataset, embedding the encoded dataset in a low-dimensional space, selecting a seed record from the embedded dataset, identifying a plurality of nearest neighbor records to the seed record, generating a new record by randomly selecting features from the plurality of nearest neighbor records, and concatenating the new record into the synthetic dataset. For a synthetic dataset that contains N records, which may be the same as or different from the number of records in the original dataset, the selecting, identifying, generating, and concatenating operations operate a total of N times on the records in the embedded dataset.

A method of distinguishing between lung squamous cell carcinoma and head and neck squamous cell carcinoma using a 22-gene biomarker signature is presented.

A method of distinguishing between lung squamous cell carcinoma and head and neck squamous cell carcinoma using a 22-gene biomarker signature is presented.

A system and method of calibrating moving cameras capturing a sporting event is disclosed herein. A computing system retrieves a broadcast video feed for a sporting event. The broadcast video feed includes a plurality of video frames. The computing system labels, via a neural network, components of a playing surface captured in each video frame. The computing system matches a subset of labeled video frames to a set of templates with various camera perspectives. The computing system fits a playing surface model to the set of labeled video frames that were matched to the set of templates. The computing system identifies camera motion in each video frame using an optical flow model. The computing system generates a homography matrix for each video frame based on the fitted playing surface model and camera motion. The computing system calibrates each camera based on the homography matrix generated for each video frame.

A system and method of calibrating moving cameras capturing a sporting event is disclosed herein. A computing system retrieves a broadcast video feed for a sporting event. The broadcast video feed includes a plurality of video frames. The computing system labels, via a neural network, components of a playing surface captured in each video frame. The computing system matches a subset of labeled video frames to a set of templates with various camera perspectives. The computing system fits a playing surface model to the set of labeled video frames that were matched to the set of templates. The computing system identifies camera motion in each video frame using an optical flow model. The computing system generates a homography matrix for each video frame based on the fitted playing surface model and camera motion. The computing system calibrates each camera based on the homography matrix generated for each video frame.

An apparatus for estimating a target component may include: a spectrometer configured to acquire training spectra; and a processor configured to extract a set number of principal components from the training spectra, determine whether the set number of principal components is appropriate based on randomness of residual in the set number of principal components, and based on the set number of principal components being determined to be appropriate, generate a target component estimation model based on the set number of principal components.

An apparatus for estimating a target component may include: a spectrometer configured to acquire training spectra; and a processor configured to extract a set number of principal components from the training spectra, determine whether the set number of principal components is appropriate based on randomness of residual in the set number of principal components, and based on the set number of principal components being determined to be appropriate, generate a target component estimation model based on the set number of principal components.

This invention relates generally to the detection of objects, such as stents, within intraluminal images using principal component analysis and/or regional covariance descriptors. In certain aspects, a training set of pre-defined intraluminal images known to contain an object is generated. The principal components of the training set can be calculated in order to form an object space. An unknown input intraluminal image can be obtained and projected onto the object space. From the projection, the object can be detected within the input intraluminal image. In another embodiment, a covariance matrix is formed for each pre-defined intraluminal image known to contain an object. An unknown input intraluminal image is obtained and a covariance matrix is computed for the input intraluminal image. The covariances of the input image and each image of the training set are compared in order to detect the presence of the object within the input intraluminal image.