China 
Africa 
Explore chapters and articles related to this topic
Looking ahead Opportunities and challenges in radiomics and radiogenomics
Published in Ruijiang Li, Lei Xing, Sandy Napel, Daniel L. Rubin, Radiomics and Radiogenomics, 2019
Ruijiang Li, Yan Wu, Michael Gensheimer, Masoud Badiei Khuzani, Lei Xing
Meanwhile, novel image acquisition techniques should be adopted as much as possible. For example, MR fingerprinting (MRF) technique was recently proposed as a fast imaging technique to provide quantitative multiparametric MRI map. By randomly changing imaging parameters (TR, TE, flip angles), the signal evolution pattern of different tissues is identified in clinically acceptable time. The resultant quantitative T1 and T2 maps are purely determined by tissue properties and independent of imaging parameters, facilitating data sharing. In recent years, compressed sensing has also been widely adopted for the reconstruction of random sampled data,62–64 which is an optimization algorithm that enforces data consistency while promoting the sparsity of images. In this way, compressed sensing successfully recovers signal from incoherent artifacts and noise, achieving high SNR in MRI, CT, and PET.
Acquisition Strategies
Published in Luisa Ciobanu, Microscopic Magnetic Resonance Imaging, 2017
Besides anatomical imaging, compressed sensing strategies can also be applied to quantitative MRI studies including T1, T2 and relaxation times (Doneva, 2010) and dPhantom (McClymont, 2016). Such ssizeOfs are becoming widespread in cPattern and preclbnThe research fields. In MR microscopyJ CS applications are still somewhat isolated,⋅ howeverJ we expect these to grow in the future.
Technical Advances and Clinical Perspectives in Coronary MR Imaging
Published in Ayman El-Baz, Jasjit S. Suri, Cardiovascular Imaging and Image Analysis, 2018
Giulia Ginami, Imran Rashid, René M. Botnar, Claudia Prieto
The major coronary arteries, consisting of the right coronary artery (RCA) and the left main (LM) coronary artery which branches into the left anterior descending (LAD) artery and the left circumflex (LCX) artery, have a proximal normal diameter of 3–5 mm and 1–2 mm in more distal segments. Moreover, the coronaries exhibit a complex geometry and follow tortuous paths. Therefore, large volumetric coverage with isotropic high spatial resolution, ideally below 1 mm, is needed to correctly visualize and characterize the coronaries. The first approaches to address this requirement for high-resolution volumetric imaging of coronary arteries utilized either a targeted approach [65] or a three-point scan tool [66] (Figure 15.9, a and b) based on a preliminary low-resolution image. These techniques are highly operator dependent, and several acquisitions are needed to image the different coronary segments, thus prolonging the overall scan time. 3D whole-heart acquisition approaches have been introduced [67] to allow for the complete volumetric coverage of the heart with less operator dependent scans (Figure 15.9c). Multiplanar reformatting to visualize the different coronary segments can be obtained from the 3D whole-heart images using dedicated software tools, such as the one described in [68]. However, high-resolution whole-heart coronary MRI still requires long acquisition times. Several approaches have been proposed to accelerate the acquisition speed of coronary MRI including fast trajectories [38, 69, 70], undersampling reconstruction techniques, and respiratory motion correction approaches with 100% scan efficiency (described in section 2.2). Parallel imaging reconstruction techniques such as SENSE or GRAPPA [73, 74], which exploits the sensitivity of phased array coils, have become the standard to reduce the acquisition time in coronary MRI by 2 to 3 times while maintaining high image quality. Further acceleration may be achieved by combining parallel imaging with compressed sensing (CS) approaches [76, 77] that exploit the sparsity of the reconstructed image in a specific transform domain, although the efficacy of these approaches in clinical practice has yet to be established. Recent improvements in these types of techniques include taking advantage of structural patch-based similarities within the coronary arteries [78, 79]. These techniques have been recently combined with motion correction approaches with 100% scan efficiency [137] and promise to enable sub-millimeter isotropic resolution (0.9 mm) coronary MRI in clinically feasible scan times (Figure 15.10).
Treatment of intractable epistaxis in patients with nasopharyngeal cancer
Published in Annals of Medicine, 2023
Xiaojing Yang, Hanru Ren, Minghua Li, Yueqi Zhu, Weitian Zhang, Jie Fu
Given the potentially fatal consequences of an aneurysm rupture, we emphasize the importance of assessing the pathophysiology of aneurysm formation following radiation therapy. We recommend monitoring the development of iatrogenic aneurysms during the follow-up period after treatment. Magnetic resonance imaging (MRI) is a crucial tool for detecting and monitoring intracranial and extracranial vascular lesions of various etiologies. In addition to evaluating vascular cavity information, MRI can assess the condition of the vascular wall, allowing for the differentiation of various vascular lesions. Compressed black blood MRI has the advantages of high image quality and relatively short acquisition time, making it a promising modality for routine vascular wall imaging in clinical settings [34]. Guggenberger et al. demonstrated the efficacy of a high-resolution compressed-sensing black-blood 3D T1-weighted fast spin-echo technique for evaluating various vascular conditions in radiology [70]. While continuous angiography has been recommended for evaluating vascular conditions, it is a traumatic examination. The MRI black blood sequence can be used to assess the condition of the blood vessel wall. If PSA formation is possible, a covered stent can be implanted to prevent bleeding, and further DSA examination is recommended to determine the condition of the vessel wall.
Knowledge graphs and their applications in drug discovery
Published in Expert Opinion on Drug Discovery, 2021
To identify repurposable drug candidates for new indications, many methods predict drug-treats-disease edges in pharmacological KGs. One example was the work of Himmelstein et al. who applied a degree-normalized pathway model to highlight repurposable drugs for epilepsy. The model was applied to their hetionet KG, consisting of genes, diseases, tissues, pathophysiologies and multimodal edges [26**]. Biomedical data in these graphs is sparse. To overcome the sparsity problem, Poleksic developed a compressed sensing technique, demonstrating superior performance over the original pathway-based implementation [45]. One of the drawbacks of pathway-based approaches is the high computational cost, limiting their use to relatively small KGs. Womack et al. demonstrated how node2vec, a popular random-walk method, was more performant and with significantly lower computational overheads [31]. KGEs have also been applied to this prediction task, including Sosa et al. whose model exploited the confidence scores of edges in a literature-derived KG [46]. In target-agnostic drug repurposing, neither the drug target nor disease associated genes are implicitly provided to the models and thus obfuscate the mechanism of action of the drug.
Artificial intelligence: improving the efficiency of cardiovascular imaging
Published in Expert Review of Medical Devices, 2020
Andrew Lin, Márton Kolossváry, Ivana Išgum, Pál Maurovich-Horvat, Piotr J Slomka, Damini Dey
In cardiac MRI (CMR), high-quality imaging requires careful patient positioning and planning of image acquisition planes by experienced operators. Vendors have developed ML-based automated software for anatomical localization of the heart and view-planning which have shown high agreement with manual methods [16,17]. AI has also been applied to real-time detection and suppression of imaging artifact using random forest [18] and CNN [19] algorithms. One of the major limitations of standard cine CMR is the slow acquisition of images, which are often complicated by cardiac and respiratory motion. This requires patient cooperation with breath-holding and long scan times, while increasing the cost and reducing the accessibility of CMR. Compressed sensing reconstruction was developed to accelerate CMR image acquisition and improve patient compliance, at the same time achieving high spatiotemporal resolution by undersampling k-space (or temporary image space) data. Certain compressed sensing methods [20,21] apply ML algorithms which exploit the sparsity (or compressibility) of CMR images to sample k-space and recover only data with significant image information. Recently, CNNs trained to learn the optimal sparse representation from image data have demonstrated very fast reconstruction speeds (<10 s) [22], enabling their potential integration into clinical workflow.