Koninklijke Philips Corporation N.V.

Andre Frank Salomon

Andreas Goedicke

Chuanyong Bai

Andriy Andreyev

Bin Zhang

Zhiqiang Hu

Volkmar Schulz

Jinghan Ye

Xiyun Song

Sven Kabus

Bernd Schweizer

Patrick Olivier

Manoj Narayanan

Yu-Lung Hsieh

Yanfei Mao

Thomas Dey

A non-transitory computer-readable medium stores instructions readable and executable by at least one electronic processor () to perform an imaging method (). The method includes: reconstructing emission imaging data to generate an emission image of a lesion; converting intensity values of the emission image to at least one standardized uptake value (SUV value) for the lesion; processing input data using a regression neural network (NN) () to output an SUV correction factor for the lesion, wherein the input data includes at least two of (i) image data comprising the emission image or a feature vector representing the emission image, (ii) the at least one SUV value, (iii) a size of the lesion, and (iv) reconstruction parameters used in the reconstructing; and controlling a display device () to display at least one of (I) the SUV correction factor and (II) a corrected SUV value generated by applying the SUV correction factor to the at least one SUV value.

Artificial intelligence (AI)-based standardized uptake value (SUV) correction and variation assessment for positron emission tomography (PET)

Imaging data () are acquired by a PET scanner () or other imaging device. Iterative image reconstruction of the imaging data is performed to generate a reconstructed image (). The iterative image reconstruction includes performing an update step () that includes an edge preserving prior () having a spatially varying edge preservation threshold () whose value at each image voxel depends on a noise metric () in a local neighborhood of the image voxel. The noise metric may be computed as an aggregation of the intensities of neighborhood image voxels of the reconstructed image in the local neighborhood of the image voxel. The edge preserving prior may be a Relative Difference Prior (RDP). For further noise suppression, during the iterative image reconstruction image values of image features of the reconstructed image that have spatial extent smaller than a threshold () may be reduced.

Information-adaptive regularization for iterative image reconstruction

The invention relates to a system and a method for assisting in attenuation correction of gated PET data of a moving object (). In the system, an evaluation unit () is configured to (i) receive a CT image of the object () and to segment the CT image into a plurality of CT sub-images, each CT sub-image correspond to an axial segment of an imaged volume, (ii) to determine for each CT sub-image a gate including PET data having a greatest correspondence with the CT sub-image, (iii) to construct, for each CT sub-image, a PET sub-image from the PET data included in the gate determined for the CT sub-image, the PET sub-image substantially corresponding to the same axial segment of the imaged volume as the CT sub-image, and (iv) to combine the PET sub-images to form a PET reference image of the object ().

Attenuation correction of PET data of moving object

A respiratory motion signal generation method operates on emission data () of an imaging subject in an imaging field of view (FOV) acquired by a positron emission tomography (PET) or single photon emission computed tomography (SPECT) imaging device (). An array of regions () is defined in the imaging FOV without reference to anatomy of the imaging subject. For each region of the array of regions defined in the imaging FOV, an activity position versus time curve () is computed from the emission data acquired by the PET or SPECT imaging device. Frequency-selective filtering of the activity position versus time curves is performed to generate filtered activity position versus time curves. At least one motion signal () is generated by combining the filtered activity position versus time curves of at least a selected sub-set of the regions.

Noise-robust real-time extraction of the respiratory motion signal from PET list-data

A time of flight (TOF) positron emission tomography (PET) image () is generated from TOF PET imaging data () acquired of a subject using a TOF PET imaging data acquisition device (). Iterative image reconstruction () of the TOF PET imaging data is performed with TOF localization of counts along respective lines of response (LORs) to iteratively update a reconstructed image (). Values for at least one regularization or filtering parameter are assigned to the TOF PET imaging data or to voxels of the reconstructed image based on an estimated TOF localization resolution for the TOF PET imaging data or voxels. Regularization () or filtering () of the reconstructed image is performed using the assigned values for the at least one regularization or filtering parameter. In some embodiments, the varying TOF localization resolution for the TOF PET imaging data or voxels is estimated based on related acquisition characteristics such as count rates or operating temperature of the detectors.

Time-of-flight resolution-adaptive image regularization and filtering in positron emission tomography

Patient imaging systems, such as PET imaging systems, for example, may suffer from the introduction of artificially introduced noise. This noise is, typically, introduced during iterations of reconstruction algorithms, such as the least-squares algorithms, which attempts to recreate a 2D or a 3D image from raw acquisition information. The noise appears as 'hot-spots' in the reconstructed image. Approaches to address these artefacts use filtering approaches. Typically, a least-squares reconstruction is supplemented with a penalty term, an approach known as 'Relative Difference Penalty'. The penalty parameter causes the reconstruction algorithm to filter more or less strongly at certain regions of the reconstruction. The present application proposes an approach which supplements the penalty term with continuous probability information about the likelihood of an edge being present in a portion of an image.

Medical image processing

A method of image reconstruction corrected for attenuation is provided for use with radioemission-based imaging, such as SPECT and PET. This method includes collecting measured emission projection data. The emission projection data, a reconstruction of the emission projection data, and a priori organ information are collectively analyzed to generate a body region estimate of the imaged subject. Each voxel of the body region estimate is then homogenously assigned an attenuation coefficient to generate an initial attenuation map estimate. An initial emission assumption is also generated based on a reconstruction of the emission projection data. The initial emission assumption and initial attenuation map estimate are then processed and refined to produce an image reconstruction.

Apparatus and method for generation of attenuation map

An imaging system includes a magnetic resonance portion that produces an electric field and a second imaging portion, including a detector with a two dimensional array of detector tiles, wherein adjacent tiles along an axial direction are spaced apart by an electrically conductive material, which shields the tiles from the electric field. An imaging system includes a first imaging portion having a detector, which includes an array of scintillation crystals and a photo-sensor coupled to the array of scintillation crystals, wherein an output of the photo-sensor includes a unique ratio of information that identifies each crystal.

Positron emission tomography and/or single photon emission tomography detector

A magnetic resonance (MR) image segmentation processor () is configured to identify one or more geometrical regions of a subject using an MR image of the subject. An emission data reconstruction processor () is configured to generate an attenuation map () of the subject by assigning initial attenuation values () to the geometrical regions of the subject, and to: (i) process () emission data acquired from the subject to generate an emission image () of the subject, the processing employing the attenuation map of the subject; (ii) update () the attenuation map based on corrections calculated using the emission image of the subject; and (iii) iterate operations (i) and (ii) to iteratively generate a reconstructed emission image of the subject.

Growing community of inventors

Aachen, Germany

Andre Frank Salomon