Adaptive multi-site scanning imaging method based on deep learning and system thereof

Abstract

The invention discloses an adaptive multi-site scanning imaging method based on deep learning and a system thereof. The method is applied to a multi-modal imaging device mainly based on the nuclear medicine, and comprises the following steps of step A, for multiple target sites of an imaging target object, conducting single-modal or multi-modal reconnaissance scanning imaging; step B, utilizing image analysis software based on a deep learning technology to analyze the reconnaissance scanning imaging data in step A, and combined with relevant prior information, detecting a local area which needs further focus imaging and quantifying importance or risks thereof; step C, based on the results of step B, selecting an optimized process and parameters to perform the next scanning. The adaptive multi-site scanning imaging method maximizes the image diagnostic value contributed by unit scanning time and / or a radiation dose, achieves optimized precision image inspection for individuals through multi-site scanning imaging application on tumors to improve the diagnostic performance, and has an important practical value.

Description

technical field [0001] The invention relates to the technical field of nuclear medicine imaging, in particular to an adaptive multi-part scanning imaging method and system based on deep learning. Background technique [0002] SPECT (Single Photon Emission Computed Tomography, Single Photon Emission Computed Tomography) and PET (Positron Emission Tomography, Positron Emission Tomography) are two imaging technologies in nuclear medicine, which generate radioactive indicators through gamma photon detection and imaging technology. Static or dynamic images that track the uptake, distribution and excretion of drugs in humans or animals, so as to provide functional information of related systems, organs, and tissues in humans or animals, and in some cases, reveal the biochemical reactions of special cells at the molecular level process, so as to be used in clinical medical diagnosis and basic medical research. The main difference between SPECT and PET lies in the gamma photon coll...

AI Technical Summary

Problems solved by technology

There are two main problems with the above-mentioned principles and research work guiding the scanning protocol: the first is that it lacks an intuitive and clear connection with the clinical application goals of imaging-diagnosis or efficacy evaluation, thus making it difficult to have a truly accurate, personalized and Quantitative standards that are generally recognized by clinics are generated; on this basis, the selection of specific scanning acquisition procedures and parameters is greatly influenced by subjective factors such as user habits and operator experience in practice, which further leads to the questioning of the clinical value of personalized acquisition procedures , so that it is difficult to really play a role

On the basis of obtaining relatively sufficient input information, how to select parameters and procedures to achieve the ideal compromise of the above-mentioned comprehensive conditions is also a relatively complicated optimization problem.

Images