73 results about "Multi contrast" patented technology

Techniques for reducing the scan time required for the acquisition of two or more magnetic resonance imaging images of a subject having differing contrasts. In one arrangement, a combo acquisition protocol of N sets of parameters is selected prior to imaging. In order to image a subject, a first set of parameter values is selected from the protocol, a first RF pulse and at least one gradient field are used to excite the subject, a refocusing RF pulse and at least one gradient field is applied to the subject, a phase encoding gradient field is applied to the subject, and then a measurement gradient field is applied to the subject simultaneously while an induced signal is measured. The process is repeated to obtain N measurements, which are then processed into two or more reconstructed images of differing contrast, and where some of the measurements are used during the reconstruction of two or more of the images.

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.

The invention belongs to the technical field of magnetic resonance reconstruction and provides a magnetic resonance imaging method and a device. The method comprises steps of acquiring completely acquired offline multi-contrast images of a sample object; carrying out insufficient acquisition on each completely acquired image in the completely acquired offline multi-contrast images in K space so as to acquire insufficiently acquired multi-contrast images; according to the insufficiently acquired multi-contrast images and the completely acquired offline multi-contrast images, training a depth learning network; acquiring the insufficiently acquired images of a to-be-measured object; and inputting the insufficiently acquired images of the to-be-measured object into the trained depth learning network so as to acquire an online magnetic resonance image of the to-be-measured object. According to the invention, under the condition that the original contrast is kept, the lost information in the insufficient acquisition of the magnetic resonance image is recovered, so imaging is accelerated and magnetic resonance imaging precision is improved.

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.

The invention, relates to the magnetic resonance imaging field, provides a convolutional-neural-network-based multi-contrast magnetic resonance image reconstruction method. A low-resolution image andhigh-resolution image of a multi-contrast magnetic resonance are obtained; a convolutional neural network model of multi-contrast magnetic resonance image reconstruction is established; the convolutional neural network is trained by using the multi-contrast magnetic resonance image as a training set; and then the low-resolution magnetic resonance image and a corresponding reference image are inputted into the network to reconstruct a high-resolution magnetic resonance image. On the basis of deep learning, the image reconstruction method with structural similarity between multi-contrast imageshas characteristics of high reconstruction speed and good reconstruction effect.

In a magnetic resonance apparatus and method for acquiring magnetic resonance data, a magnetic resonance data acquisition scanner executes a turbo spin echo (TSE) data acquisition sequence with simultaneous multi-slice (SMS) imaging wherein nuclear spins in two different slices of an examination subject are simultaneously excited so as to produce respective echo trains. The magnetic resonance data acquisition scanner is operated with the SMS imaging configured so that magnetic resonance signals from the respective slices have a different contrast, with the SMS being configured to allow evolution of magnetization of the nuclear spins for the second contrast while magnetic resonance signals with the first contrast are being detected. The respective magnetic resonance signals from the two different slices are detected and entered into an electronic memory organized as k-space, as k-space data.

In a magnetic resonance tomography apparatus and method, whereby an interactive contrast optimization is implemented by 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.

The invention discloses a novel visible light airport airplane detection method based on potential target points. The method mainly comprises the following steps: proposing a multi-contrast image segmentation method, calculating the position of a potential target point through a segmentation area of each segmentation image, and proposing a data structure similar to a set to complete the storage ofthe potential target points because the plurality of segmentation images are easy to generate redundant potential target points; secondly, counting the sizes of the aircraft targets, providing a self-adaptive method based on statistics to generate detection windows of different sizes, and matching the detection windows of different sizes with the sizes of the aircraft targets to a great extent; finally, training an aircraft target classifier in a classifier mode of combining Fourier Hog features and Adaboost; and taking the potential target point as a central point, intercepting a current region by detection windows with different sizes, and judging whether the current region contains the aircraft or not by means of a trained classifier, thereby completing the detection of the airport aircraft.

The contrast of an image captured by imaging using a multi-echo sequence by radial sampling is improved.

Images are simultaneously captured by using a multi-echo sequence by radial sampling, and echo signal groups of one or more blocks measured by executing the imaging using the multi-echo sequence are divided into a plurality of partial echo signal groups.

Using the partial echo signal groups, images with different contrasts are reconstructed.

The invention discloses a PET quick imaging method and system based on deep learning. The PET quick imaging method comprises the steps of obtaining the matched nuclear medicine image sets collected ina first long time and a second short time; preprocessing the nuclear medicine image group and normalizing the image size and the signal intensity; constructing a noise recognition neural network based on Res-UNet, wherein the noise recognition network uses a network output image composed of a network input image formed by randomly adding a noise signal to the acquired nuclear medical image and the noise signal; constructing an image reconstruction convolutional network which is provided with a plurality of convolutional neural network CNN levels connected in series and comprises an encoder-decoder residual depth network structure which is symmetrically connected in series, wherein the input is an image acquired in a second short time and a multi-contrast image used as multi-modal input; and training the image reconstruction convolutional network to enable the image reconstruction convolutional network to priori extract effective image features of the input nuclear medical image and the multi-contrast image so as to reconstruct a high-quality nuclear medical image acquired for a long time.

The application discloses a method for reconstructing a magnetic resonance multi-contrast image. A final reconstructed image is obtained by adding the transition image and the residual image of an image to be corrected, wherein the image to be corrected is at least one of initial images of N contrasts, the transition image has lower noise and artifacts than the initial images and loses a part of organizational structure information, the lost part of the organizational structure information can be compensated by the organizational structure information in the residual image, and the final imageobtained by the addition of the transition image and the residual image can maintain the fidelity of the image. Since the transition image and the residual image do not change the contrast and the resolution of the original image and both have low noise and artifacts, the final image obtained by the addition of the transition image and the residual image can maintain the contrast and the resolution of the image, and also has low noise and artifacts. The application also discloses a device for a reconstructing a magnetic resonance multi-contrast image.

In a method and imaging apparatus for acquiring multi-contrast magnetic resonance (MR) data, a data acquisition scanner is operated in a simultaneous multislice data acquisition sequence to radiate at least one single-band binomial radio-frequency (RF) pulse, that excites fat protons in at least some slices of an examination subject from which MR raw data are to be acquired simultaneously, and leaving water in a longitudinal plane for those at least some slices, and leaving all spin species in a longitudinal plane in others of the slices that are to be acquired simultaneously. A spoiler gradient is subsequently activated that dephases the fat protons that were excited. The scanner is then operated to execute an MR data acquisition sequence with excitation by radiation of multi-band RF pulses. MR raw data resulting from excitation of the fat protons, and MR raw data acquired with said multi-band RF excitation, are compiled in respective data files.

The invention discloses a magnetic resonance multi-contrast image reconstruction method and device. The method includes the following steps: collecting magnetic resonance signal data at N contrasts, wherein the magnetic resonance signal data includes signal data reflecting organizational structure information and signal data reflecting independent contrast information at each contrast; carrying out image reconstruction according to the signal data reflecting independent contrast information at each contrast to get contrast information images; carrying out image sharing information joint reconstruction according to the signal data reflecting organizational structure information to get a shared organizational structure information image of a multi-contrast image; and combining each contrastinformation image with the shared organizational structure information image of the multi-contrast image to generate an initial image at each contrast. The method not only can avoid repeated scanningof organizational structure information and save scanning time, but also can improve the accuracy of reconstructed images.

The invention relates to an image edge detection method based on multiple stochastic resonance mechanisms. A bistable serial-parallel network model with a stochastic resonance character is built, and edge detection based on the multiple stochastic resonance mechanisms is carried out on multi-contrast-ratio edges in an image in sequence from strong to weak. Independent stochastic resonance modulation is carried out on the edges with different contrast ratios by different levels, and finally locally optimum detection results are integrated into complete edge information. The function of noise in image edge enhancement and detection is fully utilized, and the detection thought on multilevel edge details under a single scale in a traditional method is changed.

In a method and apparatus for the automatic assignment of at least one combination image of an examination object to a spin species represented in the combination image, relationships, which were determined from an existing database and which relate to the assignment of spin species to combination images, are loaded into a computer. At least two MR datasets at one of at least two echo times in each case following an excitation by means of a multi-contrast measurement are supplied to the computer. At least one combination image is determined in the computer from the at least two MR datasets. The spin species represented in the at least one combination image are assigned in the computer on the basis of the loaded relationships. By using relationships determined from an existing database, an automatic unambiguous global assignment of the correct spin species is enabled.

The invention discloses a fast curved surface reconstruction method for a multi-modal multi-contrast medical image and a storage medium. The method comprises the following steps: importing a multi-modal multi-contrast image; selecting two adjacent points as initial seed points for the target blood vessel segment in the image; performing automatic center line search; in the searching process, taking the first initial seed point as a starting point, controlling the searching direction through the second initial seed point, and conducting point-by-point searching on the center line of the target blood vessel segment; smoothening the center line; scanning the smoothed center line along a user-defined direction to obtain an original curved surface required by curved surface reconstruction; and unfolding the original curved surface to obtain a curved surface reconstruction result. According to the method, the center lines of multiple contrast images such as a CT blood vessel image, an MRI bright blood image and an MRI black blood image can be rapidly extracted. In addition, according to the method, a new data structure is adopted to display a two-dimensional plane image, and resampling distortion of a curved surface image is avoided.

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.

The invention provides a multi-contrast MRI image reconstruction method based on deep learning, and relates to the technical field of medical image processing, and the method comprises the steps: collecting a real full-sampling MRI image and a reconstructed MRI image randomly sampled in a K space to form a training set sample, and training a deep convolutional neural network; then taking the under-sampled MRI images with different contrast ratios as input, outputting preliminary full-sampled MRI images with different contrast ratios, extracting structural features and contrast ratio features of the preliminary full-sampled MRI images with different contrast ratios by an encoder, then performing similarity constraint, and finally generating final complete MRI images with different contrast ratios by a generator. According to the method, the defects that at present, most models are reconstructed only through a single-contrast MRI image, and reconstruction lacks of utilization of multi-contrast MRI image associated information are overcome, the reconstruction quality of the MRI image is improved, and the reliability of a diagnosis result of a medical system is guaranteed.