Abstract
With the development of big data, Radiomics and deep-learning methods based on magnetic resonance (MR) images, it is necessary to conduct large databases containing MR images from multiple centers. Having huge intensity distribution differences among images reduced or even eliminated, robust computer-aided diagnosis models could be established. Therefore, an optimized intensity standardization model is proposed. The network structure, loss function, and data input strategy were optimized to better avoid the image resolution loss during transformation. The experimental dataset was obtained from five MR scanners located in four hospitals and was divided into nine groups based on the imaging parameters, during which 9152 MR images from 499 participants were collected. Experiments show the superiority of the proposed method to the previously proposed unified model in resolution metrics including the peak signal-to-noise ratio, structural similarity, visual information fidelity, universal quality index, and image fidelity criterion. Another experiment further shows the advantage of the proposed method in increasing the effectiveness of following computer-aided diagnosis models by better preservation of MR image details. Moreover, the advantage over conventional standardization methods are also shown. Thus, MR images from different centers can be standardized using the proposed method, which will facilitate numerous data-driven medical imaging studies.
Highlights
As the most commonly used imaging method in the diagnosis of brain diseases, magnetic resonance imaging (MRI) is one of the research hotspots of computer-aided brain diagnosis in recent years [1]
Images are not needed for training and spatial intensity nonuniformities might be eliminated with the fusion of regional and global information, we propose a universal intensity standardization method based on cycle generative adversarial network [21]
We compared the proposed method in general properties with the original unified method and the representative methods of the two major types of intensity standardization methods, the histogram matching method proposed by Sun et al and the joint histogram registration method previously proposed by our group
Summary
As the most commonly used imaging method in the diagnosis of brain diseases, magnetic resonance imaging (MRI) is one of the research hotspots of computer-aided brain diagnosis in recent years [1]. Many magnetic resonance (MR) image based studies, such as computer aided diagnosis [2,3,4], differential diagnosis [5], treatment options selection [6], and prognosis estimation [7], have made great progress, which put forward higher requirements for the quantity, and the quality of the image data. To conduct a larger unified training set contains MR images obtained from different MR scanners, the scale and intensity distribution difference of such images should be suppressed.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
