Compressed sensing MRI: a review of the clinical literature

Abstract

MRI is one of the most dynamic and safe imaging techniques available in the clinic today. However, MRI acquisitions tend to be slow, limiting patient throughput and limiting potential indications for use while driving up costs. Compressed sensing (CS) is a method for accelerating MRI acquisition by acquiring less data through undersampling of k-space. This has the potential to mitigate the time-intensiveness of MRI. The limited body of research evaluating the effects of CS on MR images has been mostly positive with regards to its potential as a clinical tool. Studies have successfully accelerated MRI with this technology, with varying degrees of success. However, more must be performed before its diagnostic efficacy and benefits are clear. Studies involving a greater number radiologists and images must be completed, rating CS based on its diagnostic efficacy. Also, standardized methods for determining optimal imaging parameters must be developed.

Figure 1.

The number of articles per year (a) on the topic of compressed sensing MRI and (b) met the inclusion criteria for review in this article.

Figure 2.

Scheme for categorizing the study design of each article. Studies were first grouped as to whether they evaluated image quality or diagnostic efficacy (some assessed both). These studies were then further classified as to whether they employed qualitative or quantitative assessment methods.

Figure 3.

The number of studies that have evaluated compressed sensing with each MR application: dynamic contrast-enhanced MRI (DCE-MRI);^,, paediatric MRI;^, MR angiography;^, MR spectroscopy of the brain and prostate and for the measurement of phosphocreatine regeneration in muscle; phase-contrast MRI for cardiac imaging^, and vascular flow quantification;^, sodium MRI for early detection of osteoarthritis; chemical shift imaging for fat fraction quantification in Becker's muscular dystrophy; multispectral imaging (MSI) of the spine; contrast-enhanced multiphase MRI of the liver; and brain MRI. Subjects imaged in the paediatric MRI studies were referred for abdominal, cardiac, knee or cholangiopancreatography.

Figure 4.

The number of studies that performed prospective compressed sensing acquisition, retrospective compressed sensing simulation on fully sampled data sets and both methods.

Figure 5.

The breakdown of imaging subjects across studies, grouped as healthy subjects and patients.

Figure 6.

The number of studies which utilized a given number of human evaluators to evaluate the images. Studies which did not utilize human image evaluators assessed images by measuring certain image attributes. Nine of the studies included two evaluators,^,,,–,– and one study included one evaluator.

Figure 7.

The number of image evaluators across studies who were blinded to image type or patient history or not blinded to this information. Of the 19 radiologists, 15 were blinded to image type, patient history or both. Four radiologists across two studies were not blinded with regard to image type or patient history.^,

Figure 8.

Acceleration rates tested across studies. Grey indicates the number of studies that obtained images deemed acceptable at a given acceleration rate, and black indicates the number of studies that obtained unacceptable images.

Figure 9.

Visual quantification of the image assessment methods implemented across the studies. Studies were divided based on the assessment of image quality or diagnostic efficacy and further classified based on quantitative or qualitative methodology. 17 articles assessed image quality, 16 quantitatively^,–,– and 1 qualitatively. Three articles assessed the diagnostic efficacy of compressed sensing MRI, two qualitatively^, and one quantitatively. Note that some studies fall under multiple categories.