Special Issue "Whole Body MRI: Restoration and Analysis with Signal/Image Processing Principles"
A special issue of Signals (ISSN 2624-6120).
Deadline for manuscript submissions: 31 January 2024 | Viewed by 982
2. Department of Biological Sciences, University of Cyprus, Nicosia 1678, Cyprus
Interests: medical image analysis
Interests: data science; epilepsy; Parkinson’s disease; light and fluorescent microscopy; shear-wave elastography; neuroradiology; functional neuroradiology; MR-spectroscopy
Special Issues, Collections and Topics in MDPI journals
Whole-body MRI (WB-MRI) has a large field-of-view (FoV) that covers the entire body. Due to its broad data acquisition in high anatomical resolution, WB-MRI is a competitive image expedition tool to both cover conventional clinical needs and enable novel observations. Current indications span over a spectrum of physiology and pathology. An example for physiology is the monitoring of the body elemental compartments using fat and muscle segmentation in sports medicine applications. An expanding list of indications for WB-MRI in clinical oncology and the evidence collected so far argument towards a future first-line indication of WB-MRI in cancer staging and follow up. Implementations for bone metastatic disease detection reveal equal or higher sensitivity and specificity of the WB-MRI compared to classical radiation-base methods such as the bone scans with Tc99m-based radiopharmaceuticals and positron emission tomography CT (PET-CT). Thus, MRI with T1-weighting, T2-weighting and other contrasts offers a screening solution with high anatomical resolution, free of ionizing radiation and, eventually, free of Contrast Enhancing (CE) agents.
WB- MRI is a technological and clinical hotspot of research to improve the state-of-the-art on various fronts. Large FoVs, parallel imaging with multiple coils having non-uniform sensitivities, and large scanning times prone to motion artifacts are only some of the technologies and the corresponding sources of artifacts that limit image quality and that need to be addressed. Quality improvements of the imaging data have their cornerstone in image restoration with effective and efficient pre-processing steps. A basic step is image denoising while at the same time preserving essential structures. These data often also suffer from motion artifacts. The motion can be due to discomfort, or due to involuntary physiological functions such as breathing and gastrointestinal tract (GIT) peristalsis. Voluntary moving artifacts due to, for example, the movement of extremities, are also frequently encountered. Motion correction, as well as registration, are essential steps to compensate for the displacements between scans of the same patient that are part of an imaging protocol. Hence, there is a need for WB intra-patient image registration.
WB-MRI images are large FoV data from multicoil acquisitions, inevitably suffering from extensive intensity nonuniformity artifacts. In this Special Issue, we call, among others, on research challenges for smooth signal corrections for individual coils, as well as compensatory solutions for signal intensity and chemical shifting jumps at the coil junctions. That is, these data require appropriate combined intensity uniformity restorations.
In this Special Issue, we are thrilled to host a stimulating interaction platform on innovative WB-MRI pre-processing steps to improve the quality of the data as well as to improve the conspicuity of the findings. This can perhaps obviate the need for contrast enhancement (CE) agents during imaging. Additional information that can also obviate the need for contrast enhancement can originate from diffusion-weighted imaging (DWI). Hence, we are looking forward to endorsing manuscripts on WB-DWI MRI pre-processing.
All in all, based on the MRI technical improvements of the last decade, the MRI community speculates on a future “one-stop-shop” role for WB-MRI, with implementations that go beyond bone staging and cover individual organs such as liver, pancreas, spleen or even GIT and the lungs.
The objective is to use the corrected imaging data for further analysis, i.e., for providing a clinical report or for the final diagnosis of a patient. This can be done by a trained and experienced radiologist. However, due to the size and the amount of information in the data this is not only cumbersome, but also limiting. It is preferable and sometimes even necessary to perform computer analysis of the imaging data to extract semantic information. One type of automated analysis can be along the lines of detection, segmentation and compartmentalization. An example is the compartmentalization into tissue types throughout the body, for example, into muscle, fat and other tissues with DIXON, to assess the effect of exercise on an athlete. Another example is for pathology to achieve the detection, volumetry and statistical quantitative evaluation of organ lesions, especially in disseminated expressions, which is impossible to achieve without computative support.
The time requirements both for image acquisition as well as for analysis are high for whole-body imaging. There is a need to expedite both. The whole-body imaging has been expedited with parallel imaging, partial (half) Fourier imaging, and other signal processing techniques such as compressed sensing. These imaging and reconstruction technologies must be further improved, without at the same time ignoring the requirement to maintain a high data quality.
The large size of the acquired datasets, approximately one gigabyte per patient, sets challenging space and time requirements in data processing. The rigid registration lasts about 10–15 min, which adds on to other time-demanding pre-processing and analysis steps. This time is not necessarily available considering clinical routine speed. Hence, there is a need for efficient methodologies and for their efficient implementation for all processing steps. It is necessary to emphasize the critical steps in the processing and summarize that which remains. The methods must be implemented efficiently; whenever possible, they should be parallelized and implemented with parallel acceleration hardware such as general-purpose graphical processing units (GP-GPUs).
Beyond the work of individual laboratories, the progress in this field can also benefit from collaboration between laboratories. To this end, this issue also invites manuscripts for projects for WB MRI public databases openly available to the community for analysis and evaluation. The databases can consist of the typical data modalities discussed above, which are T1 weighted (T1w), T2 weighted (T2w), inversion recovery (IR), DIXON and diffusion. Data from other modalities are also welcome. Furthermore, these databases can enable cross-sectional studies for population analysis.
The objective of this Special Issue is to emphasize the significance of WB-MRI and improve the state-of-the-art on WB MRI data processing. To this end, it intends to consolidate the problem of WB-MRI reconstruction, restoration and analysis for clinical interpretation. The methodologies to be investigated can be based on both model-based signal and image processing principles, as well as on machine learning and neural networks. This Special Issue is expected to develop a novel methodology in the various aspects of the WB processing problem that will directly benefit the extensive and efficient patient examination.
Dr. Stathis Hadjidemetriou
Dr. Ismini E Papageorgiou
Manuscript Submission Information
Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.
Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Signals is an international peer-reviewed open access quarterly journal published by MDPI.
Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 1000 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.