73 results about "Multi contrast" patented technology

A method of identifying and localizing objects belonging to one of three or more classes, includes deriving vectors, each being mapped to one of the objects, where each of the vectors is an element of an N-dimensional space. The method includes training an ensemble of binary classifiers with a CISS technique, using an ECOC technique. For each object corresponding to a class, the method includes calculating a probability that the associated vector belongs to a particular class, using an ECOC probability estimation technique. In another embodiment, increased detection accuracy is achieved by using images obtained with different contrast methods. A nonlinear dimensional reduction technique, Kernel PCA, was employed to extract features from the multi-contrast composite image. The Kernel PCA preprocessing shows improvements over traditional linear PCA preprocessing possibly due to its ability to capture high-order, nonlinear correlations in the high dimensional image space.

A method of identifying and localizing objects belonging to one of three or more classes, includes deriving vectors, each being mapped to one of the objects, where each of the vectors is an element of an N-dimensional space. The method includes training an ensemble of binary classifiers with a CISS technique, using an ECOC technique. For each object corresponding to a class, the method includes calculating a probability that the associated vector belongs to a particular class, using an ECOC probability estimation technique. In another embodiment, increased detection accuracy is achieved by using images obtained with different contrast methods. A nonlinear dimensional reduction technique, Kernel PCA, was employed to extract features from the multi-contrast composite image. The Kernel PCA preprocessing shows improvements over traditional linear PCA preprocessing possibly due to its ability to capture high-order, nonlinear correlations in the high dimensional image space.

The invention provides a vascular plaque composition recognition method (300) based on a multi-contrast magnetic resonance image. The method comprises the following steps: implementing composition labeling (S310) on the to-be-trained multi-contrast vascular plaque magnetic resonance image; inputting the labeled multi-contrast vascular plaque magnetic resonance image into a convolutional neural network and implementing network model training (S320); and inputting the to-be-recognized multi-contrast vascular plaque magnetic resonance image into a trained network model and predicting the image, so as to output a vascular plaque composition recognition result (S330). According to the vascular plaque composition recognition method, the multi-contrast vascular plaque magnetic resonance image undergoes learning and modeling via the convolutional neural network, so that a new sample can be effectively recognized to assist a doctor in a diagnosis process, and working efficiency of the doctor can be greatly improved. The technical scheme can be conveniently promoted to magnetic resonance image assisted diagnosis processes of other organs.

The present invention relates to a nuclear spins tomography photograph (synonym: magnetic resonance tomography photograph, MRT), for examining patient in medical, especially relates to a magnetic resonance image forming method than can realize interactive contrast optimization and a nuclear spins tomography photograph device for actualizing the method. The optimization MRT image method has following steps: defining a multi-contrast sequence for exciting nuclear spins in a slice of a subject to be measured; defining parameters characteristic of the sequence; measuring a number of contrast versions of the slice in the form of raw data with the previously defined sequence; processing the raw data of the slice, and thereby generating a number of images of the slice that differ in contrast from one another; and generating an image with improved contrast on the basis of an interactive real-time contrast variation of the images acquired by the multi-contrast sequence, via a user interface.

A method of improving diagnostic and functional imaging is provided by obtaining at least two input images of a subject, using a medical imager, where each input image includes a different contrast, generating a plurality of copies of the input images using non-local mean (NLM) filtering, using an appropriately programmed computer, where each input image copy of the subject includes different spatial characteristics, obtaining at least one reference image of the subject, using the medical imager, where the reference image includes imaging characteristics that are different form the input images of the subject, training a deep network model, using data augmentation on the appropriately programmed computer, to adaptively tune model parameters to approximate the reference image from an initial set of the input and reference images, with the goal of outputting an improved quality image of other sets of low SNR low resolution images, for analysis by a physician.

To perform Dixon MRI, generate multi-contrast images, and extract multi-parametric maps, this invention presents a multi-echo gradient echo protocol with two sets of echo trains. An example implementation of the invention at 3 T acquires a short-TE train (ΔTE˜1.2 ms, TE <10 ms), which is used to map B0 inhomogeneity and proton density fat fraction (FF), and a second—susceptibility sensitive—long-TE train (16 ms<TE<45 ms) will enable quantification of local frequency shift (LFS) and susceptibility. The presented pipeline automatically generates co-registered images and maps with / without fat-suppressed, including magnitude- and complex-based FF map, B0 map, anatomical images, brain mask, R2* map, unwrapped phase maps for each echo, susceptibility-sensitive images (SWI, LFS and quantitative susceptibility) for each echo, mean susceptibility-sensitive images for each echo-train. The invention is directly applicable to whole head / neck, liver, knee or even whole body scans with sliding table.

The invention provides a magnetic resonance image generation method and system, a terminal and a storage medium. The method comprises the steps of constructing a magnetic resonance image data set; constructing a convolutional neural network model; using a magnetic resonance image data set to train the convolutional neural network model until a cost function of the neural network converges; and inputting scanning data of a group of magnetic resonance imaging instruments into the trained convolutional neural network model to generate a multi-contrast magnetic resonance imaging result. Through deep learning convolutional neural network model U-Net, multi-contrast magnetic resonance images can be generated from magnetic resonance fingerprints (MRF) scanned in a single sequence, that is, the images with multiple contrast ratios can be generated at a time through scanning. Therefore, not only can the scanning time be greatly reduced, but also the cost of magnetic resonance imaging can be reduced, so that the clinical application of the technology is wider.

A denoising method and apparatus for multi-contrast MRI's are disclosed. An aspect of the invention provides a denoising method for an MRI that includes: acquiring multiple MRI's having different contrast levels for the same site; determining pixels corresponding to the same tissue by using the MRI's; and calculating a new luminance value for the pixels by using luminance values of the pixels that are determined to be belonging to the same tissue.

A magnetic resonance (MR) method and apparatus use simultaneous multislice imaging, with different excitations being effective for different slices in respective iterations of a single scanning sequence, in order to acquire raw MR data from different multiple slices, with respectively different contrasts, in the single scanning sequence. Single band excitation of a first slice among the multiple slices takes place in a first iteration of the single scanning sequence, with multi-band excitation then occurring for all of the multiple slices. Raw data are then acquired from the first slice, and at least one other slice among the multiple slices, that respectively exhibit different contrasts due to only the first slice being affected by the single band excitation. In a second iteration of the single scanning sequence, another slice is excited with single band excitation, and the first slice is among the multiple slices excited with multi-band excitation. Raw data are then acquired from the first slice and at least one other slice among the multiple slices that have the respective contrasts swapped in comparison to the first iteration.