[신간의 별자리x] 우리/미술, 그리고 ‘슬픔의 박물관’

‡We describe the use of singular value decomposition in transforming genome-wide expression data from genes 3 arrays space to reduced diagonalized ‘‘eigengenes’’ 3 ‘‘eigenarrays’’ space, where the eigengenes (or eigenarrays) are unique orthonormal superpositions of the genes (or arrays). Normalizing the data by filtering out the eigengenes (and eigenarrays) that are inferred to represent noise or experimental artifacts enables meaningful comparison of the expression of different genes across different arrays in different experiments. Sorting the data according to the eigengenes and eigenarrays gives a global picture of the dynamics of gene expression, in which individual genes and arrays appear to be classified into groups of similar regulation and function, or similar cellular state and biological phenotype, respectively. After normalization and sorting, the significant eigengenes and eigenarrays can be associated with observed genome-wide effects of regulators, or with measured samples, in which these regulators are overactive or underactive, respectively.

Singular Value Decomposition for Genome-Wide Expression Data Processing and Modeling

Magnetic resonance is an exceptionally powerful and versatile measurement technique. The basic structure of a magnetic resonance experiment has remained largely unchanged for almost 50 years, being mainly restricted to the qualitative probing of only a limited set of the properties that can in principle be accessed by this technique. Here we introduce an approach to data acquisition, post-processing and visualization—which we term ‘magnetic resonance fingerprinting’ (MRF)—that permits the simultaneous non-invasive quantification of multiple important properties of a material or tissue. MRF thus provides an alternative way to quantitatively detect and analyse complex changes that can represent physical alterations of a substance or early indicators of disease. MRF can also be used to identify the presence of a specific target material or tissue, which will increase the sensitivity, specificity and speed of a magnetic resonance study, and potentially lead to new diagnostic testing methodologies. When paired with an appropriate pattern-recognition algorithm, MRF inherently suppresses measurement errors and can thus improve measurement accuracy. A new approach to magnetic resonance, ‘magnetic resonance fingerprinting', is reported, which combines a data acquisition scheme with a pattern-recognition algorithm that looks for the ‘fingerprints’ of interest within the data. Although nuclear magnetic resonance is a powerful analytical tool for many scientific and medical disciplines, usually only a fraction of its potential power is harnessed. Most implementations are qualitative, and restricted in the range of properties that are probed. Dan Ma and colleagues introduce a new approach — termed magnetic resonance fingerprinting — aimed at greatly enhancing the amount of quantitative information that can be obtained in one measurement. Their approach combines a data-acquisition scheme that is indiscriminate in the material properties that it probes with pattern-recognition algorithms that look for the 'fingerprints' of interest within the data. Magnetic resonance fingerprinting has the potential to detect and analyse early indicators of disease or complex changes in materials, as well as increasing the sensitivity, specificity and speed of magnetic resonance studies.

/pdf/magnetic-resonance-fingerprinting-3wu7niu1v6.pdf

Magnetic resonance fingerprinting

Parametric mapping techniques provide a non-invasive tool for quantifying tissue alterations in myocardial disease in those eligible for cardiovascular magnetic resonance (CMR). Parametric mapping with CMR now permits the routine spatial visualization and quantification of changes in myocardial composition based on changes in T1, T2, and T2*(star) relaxation times and extracellular volume (ECV). These changes include specific disease pathways related to mainly intracellular disturbances of the cardiomyocyte (e.g., iron overload, or glycosphingolipid accumulation in Anderson-Fabry disease); extracellular disturbances in the myocardial interstitium (e.g., myocardial fibrosis or cardiac amyloidosis from accumulation of collagen or amyloid proteins, respectively); or both (myocardial edema with increased intracellular and/or extracellular water). Parametric mapping promises improvements in patient care through advances in quantitative diagnostics, inter- and intra-patient comparability, and relatedly improvements in treatment. There is a multitude of technical approaches and potential applications. This document provides a summary of the existing evidence for the clinical value of parametric mapping in the heart as of mid 2017, and gives recommendations for practical use in different clinical scenarios for scientists, clinicians, and CMR manufacturers.

https://link.springer.com/content/pdf/10.1186/s12968-017-0389-8.pdf

Clinical recommendations for cardiovascular magnetic resonance mapping of T1, T2, T2 and extracellular volume: A consensus statement by the Society for Cardiovascular Magnetic Resonance (SCMR) endorsed by the European Association for Cardiovascular Imaging (EACVI)

What has happened in machine learning lately, and what does it mean for the future of medical image analysis? Machine learning has witnessed a tremendous amount of attention over the last few years. The current boom started around 2009 when so-called deep artificial neural networks began outperforming other established models on a number of important benchmarks. Deep neural networks are now the state-of-the-art machine learning models across a variety of areas, from image analysis to natural language processing, and widely deployed in academia and industry. These developments have a huge potential for medical imaging technology, medical data analysis, medical diagnostics and healthcare in general, slowly being realized. We provide a short overview of recent advances and some associated challenges in machine learning applied to medical image processing and image analysis. As this has become a very broad and fast expanding field we will not survey the entire landscape of applications, but put particular focus on deep learning in MRI. 
Our aim is threefold: (i) give a brief introduction to deep learning with pointers to core references; (ii) indicate how deep learning has been applied to the entire MRI processing chain, from acquisition to image retrieval, from segmentation to disease prediction; (iii) provide a starting point for people interested in experimenting and perhaps contributing to the field of machine learning for medical imaging by pointing out good educational resources, state-of-the-art open-source code, and interesting sources of data and problems related medical imaging.

An overview of deep learning in medical imaging focusing on MRI

Purpose

This study explores the possibility of using gradient echo-based sequences other than balanced steady-state free precession (bSSFP) in the magnetic resonance fingerprinting (MRF) framework to quantify the relaxation parameters 



Methods

An MRF method based on a fast imaging with steady-state precession (FISP) sequence structure is presented A dictionary containing possible signal evolutions with physiological range of T1 and T2 was created using the extended phase graph formalism according to the acquisition parameters The proposed method was evaluated in a phantom and a human brain T1, T2, and proton density were quantified directly from the undersampled data by the pattern recognition algorithm



Results

T1 and T2 values from the phantom demonstrate that the results of MRF FISP are in good agreement with the traditional gold-standard methods T1 and T2 values in brain are within the range of previously reported values



Conclusion

MRF-FISP enables a fast and accurate quantification of the relaxation parameters It is immune to the banding artifact of bSSFP due to B0 inhomogeneities, which could improve the ability to use MRF for applications beyond brain imaging Magn Reson Med 74:1621–1631, 2015 © 2014 Wiley Periodicals, Inc

/pdf/mr-fingerprinting-using-fast-imaging-with-steady-state-5565li3a40.pdf

MR fingerprinting using fast imaging with steady state precession (FISP) with spiral readout

Magnetic resonance (MR) fingerprinting is a technique for acquiring and processing MR data that simultaneously provides quantitative maps of different tissue parameters through a pattern recognition algorithm. A predefined dictionary models the possible signal evolutions simulated using the Bloch equations with different combinations of various MR parameters and pattern recognition is completed by computing the inner product between the observed signal and each of the predicted signals within the dictionary. Though this matching algorithm has been shown to accurately predict the MR parameters of interest, one desires a more efficient method to obtain the quantitative images. We propose to compress the dictionary using the singular value decomposition, which will provide a low-rank approximation. By compressing the size of the dictionary in the time domain, we are able to speed up the pattern recognition algorithm, by a factor of between 3.4-4.8, without sacrificing the high signal-to-noise ratio of the original scheme presented previously.

SVD Compression for Magnetic Resonance Fingerprinting in the Time Domain

Purpose
To introduce a two-dimensional MR fingerprinting (MRF) technique for quantification of T1, T2, and M0 in myocardium.

Methods
An electrocardiograph-triggered MRF method is introduced for mapping myocardial T1, T2, and M0 during a single breath-hold in as short as four heartbeats. The pulse sequence uses variable flip angles, repetition times, inversion recovery times, and T2 preparation dephasing times. A dictionary of possible signal evolutions is simulated for each scan that incorporates the subject's unique variations in heart rate. Aspects of the sequence design were explored in simulations, and the accuracy and precision of cardiac MRF were assessed in a phantom study. In vivo imaging was performed at 3 Tesla in 11 volunteers to generate native parametric maps.

Results
T1 and T2 measurements from the proposed cardiac MRF sequence correlated well with standard spin echo measurements in the phantom study (R2 > 0.99). A Bland-Altman analysis revealed good agreement for myocardial T1 measurements between MRF and MOLLI (bias 1 ms, 95% limits of agreement −72 to 72 ms) and T2 measurements between MRF and T2-prepared balanced steady-state free precession (bias, −2.6 ms; 95% limits of agreement, −8.5 to 3.3 ms).

Conclusion
MRF can provide quantitative single slice T1, T2, and M0 maps in the heart within a single breath-hold. Magn Reson Med 77:1446–1458, 2017. © 2016 International Society for Magnetic Resonance in Medicine

/pdf/mr-fingerprinting-for-rapid-quantification-of-myocardial-t1-31ibcljw3v.pdf

MR fingerprinting for rapid quantification of myocardial T1 , T2 , and proton spin density.

A rapid technique for quantitative abdominal imaging was developed by using a fast imaging with steady-state free precession MR fingerprinting acquisition in combination with the Bloch-Siegert B1 mapping method, allowing simultaneous quantification of T1 and T2 in the abdomen within a 19-second breath hold.

https://pubs.rsna.org/pb-assets/Radiology/podcasts/transcripts/2016/2016-04-transcript.pdf

Dan Ma

Papers

Magnetic resonance fingerprinting

MR fingerprinting using fast imaging with steady state precession (FISP) with spiral readout

SVD Compression for Magnetic Resonance Fingerprinting in the Time Domain

MR fingerprinting for rapid quantification of myocardial T1 , T2 , and proton spin density.

MR Fingerprinting for Rapid Quantitative Abdominal Imaging