Proceedings of the World Molecular Imaging Congress 2021, October 5-8, 2021: General Abstracts
Molecular Imaging and Biology
Body : Purpose: To demonstrate an application of fast iterative method to enhance contrast and image resolution in clinical MRI imaging. An efficient rapidly converging deconvolution algorithm with a novel resolution subsets-based approach RSEMD for improving the quantitative accuracy of previously reconstructed clinical MRI images by commercial system has been evaluated. Materials and Methods: The method was tested on ACR MRI phantom and DICOM clinical MRI data. Data acquisition was performed
... n a commercial Siemens MRI system. The method was applied to MRI images previously processed with clinical MRI software to determine improvements in resolution and contrast to noise ratio. Results: In all of the phantom and patients' MRI studies the post-processed images proved to have higher resolution and contrast as compared with images reconstructed by conventional methods. In general, the values of CNR reached a plateau at around 8 iterations with an average improvement factor of about 1.7 for processed MRI images. Improvements in image resolution after the application of the method have also been demonstrated. Conclusions: An efficient, iterative deconvolution algorithm with a novel resolution subsets-based approach that operates on patient DICOM images has been used for quantitative improvement in MRI clinical imaging. The method can be applied to clinical MRI images and will be crucial in order to facilitate diagnosis of tumor progression at the earliest stages. The method can be considered as an extended blind deblurring or Richardson-Lucy like algorithm with multiple resolution levels. The uncertainty caused in the system was modeled as an iterative deconvolution with resolution subsets to denoise and enhance image resolution. This efficient extension of the blind deblurring algorithm iterates the MRI clinical image with different resolution parameters sigma and a corresponding number of iterations n (sigma) for each subset are taken in turn. In this case RSEMD method look similar to an extended Richardson-Lucy algorithm with multiple resolution levels (resolution subsets): RSEMD algorithm iterates the clinical image consistently with different resolution parameters (to maximize SNR/CNR) and a corresponding number of iterations for each resolution subset are taken in turn. During the iteration procedure the SNR is checked in each iterative step and this process can be repeated until the enhancement procedure reaches the highest SNR and resolution. The first parameter can be set as a small fraction of initial SNR. The second parameter is an initial resolution parameter (width). The original image is never revisited after the first iteration. For most clinical MRI cases, the total number of iterations for enhanced image quality is around 8 with a total number of resolution subsets around 4. References:  "Magnetom Aera, transforming 1.5T economics". Siemens AG, www.siemens.com/aera (2012).  "Syngo MR E11, increase your efficiency, expand your MRI services " , www.usa.siemens.com/E11, (2015)  Robson MD, Gore JC, Constable RT, "