FIELD: data processing.
SUBSTANCE: group of inventions relates to generation of synthetic images using machine learning algorithms for use in radiotherapy, specifically to systems and methods for generating computer tomography (CT) images from magnetic resonance imaging (MRT) images using neural networks. Computer-aided method for generating a synthetic anatomical image based on an anatomical portion of the original image obtained by a visualization apparatus using a first imaging modality, comprises receiving an initial image of an anatomical portion obtained by a visualization apparatus using a first imaging modality, receiving a model of a convolutional neural network trained to predict a synthetic image based on the original image and transforming the original image into a synthetic image by the processor using the convolutional neural network model, wherein the synthetic image is similar to the anatomical image in the second imaging modality, wherein the second imaging modality differs from the first imaging modality. System for generating a synthetic image of an anatomical portion based on an anatomical portion of the initial image obtained by a visualization apparatus using a first imaging modality, comprises input interface, storage device and processor for image processing. Non-temporal machine-readable data medium contains instructions which, being executed by the processor, cause the processor to perform a method for generating a synthetic image of the anatomical part based on the initial image of the anatomical portion obtained by the imaging device using the first imaging modality.
EFFECT: use of the group of inventions widens the range of means for generating synthetic CT images from other images, such as MRT images.
28 cl, 9 dwg
