Signal abnormalities on 1.5 and 3 Tesla brain MRI in multiple sclerosis patients and healthy controls. A morphological and spatial quantitative comparison study
Introduction
Magnetic resonance imaging (MRI) has a unique sensitivity for detecting tissue abnormalities in the central nervous system (CNS). Therefore, its use has been progressively increased in recent decades, enabling better diagnosis and prognosis of several neurological diseases.
In particular, MRI is the most sensitive diagnostic method for detection of inflammatory lesions in the CNS in patients with multiple sclerosis (MS) (Zivadinov, 2007). MRI plays an important role in diagnosis and prognosis of MS (McDonald et al., 2001) and is also commonly used as a surrogate marker to monitor disease activity in clinical trials (Miller, 1995, Paty et al., 1994).
The MRI criteria for MS focus on evidence for dissemination of lesions in time and space (McDonald et al., 2001), but several factors can affect the number and volume of MS lesions that can be identified on serial MRI scans. These include: the choice of pulse sequence (Filippi et al., 1996, Patola et al., 2001, Yousry et al., 1997), slice thickness (Dolezal et al., 2007, Filippi et al., 1998b, Molyneux et al., 1998), repositioning errors (Filippi et al., 1997), spatial resolution (Molyneux et al., 1998), differences among types of scanners (Filippi et al., 1999) and magnetic field strengths (Schima et al., 1993, Sicotte et al., 2003).
As high-field imaging becomes increasingly available in clinical routine care, it is more important than ever to understand the impact on SA detection rate of a change to higher field, including commonly associated changes in sequence parameters and/or resolution.
Higher field strength has been shown to detect more SA in clinically defined MS (Fischbach and Bruhn, 2008, Keiper et al., 1998, Lee et al., 1995, Sicotte et al., 2003) and in clinically isolated syndrome (Wattjes et al., 2006a, 2008).
Recent studies have used voxel-wise techniques to compare lesion distributions across populations (Charil et al., 2003, DeCarli et al., 2005, Di Perri et al., 2008, Enzinger et al., 2006, Ghassemi et al., 2008, Lee et al., 1998, Narayanan et al., 1997, Wen and Sachdev, 2004), and the resulting statistical parametric and non-parametric SA probability maps (SAPM) have been shown to be powerful tools for studying SA distribution in vivo.
To the best of our knowledge, SA characteristics between clinically common 1.5 T and 3 T imaging protocols have never been evaluated using fully automated and technically advanced comparison procedures. In particular, no MR studies have focused on possible morphological and spatial differences in brain SA distribution between the two different imaging protocols in MS patients and normal controls (NC).
On this basis, we aimed to investigate the effect of changing to a higher-field scanner on the number, volume and spatial distribution of SA on brain MRI in a sample of MS patients and NC, using SA paired- and voxel-wise fully automated comparison procedures.
Section snippets
Study population
We studied 41 MS patients and 38 NC. All patients were consecutively selected from patients referred to the MS outpatient clinic who fulfilled the criteria for definite MS (Polman et al., 2005). The inclusion criteria were: MRI examination performed at the time of their clinical visit (+/− 24 h), age 18–80 years, Expanded Disability Status Scale (EDSS) (Kurtzke, 1983) 0–8.5, and relapsing-remitting (RR) or secondary-progressive (SP) disease type (Lublin and Reingold, 1996). Exclusion criteria
Demographic characteristics
The study included 41 MS patients (sex: 32 females; disease course: 32 RR, 9 SP; age in years: mean 45.5 ± 8.7, median 47, range 22–58; age at onset in years: mean 30.8 ± 8.3, median 30.5, range 18–50; age at diagnosis in years: mean 34.7 ± 9.2, median 35, range 19–50; disease duration in years: mean 14.9 ± 7.9, median 14.9, range 2–40; EDSS: mean 3.5 ± 2, median 3.7, range 0–7) and 38 NC (sex: 26 females; age in years: mean 37.76 ± 12.9, median 35.5, range 19–60). DMT status was distributed as follows: 27
Discussion
In the present study, we assessed morphological and spatial SA differences between scans acquired on 1.5 T and 3 T scanners using fully automated pair- and voxel-wise quantitative comparison methods.
For pair-wise per SA number and individual volume analyses in MS patients, 3 T showed 16.9% more T2H by number and 35.7% higher T2H volume compared to 1.5 T, and for T1 lesion-wise paired analysis, 3 T showed 25% more T1H by number and 20% higher T1H volume compared to 1.5 T (Table 3, Table 4). In
Conclusions
This lesion- and voxel-wise comparison study between 1.5 T and 3 T scanners in MS patients and NC provides important information regarding morphological and spatial differences between the two scanning approaches.
Use of high-field MRI systems (3 T) increases the likelihood of detecting T2H and T1H, especially those of smaller individual volume. The spatial differences are not homogeneous, i.e., it is more likely to detect more SA in specific brain areas. Further studies are warranted to
Acknowledgments
This study was performed as part of a research collaboration agreement among General Electric Company, The State University of New York at Buffalo, Kaleida Health Systems, and University Neurology, Inc. Dr. Robert Zivadinov served as PI on this research. The authors have no financial relationship to disclose.
Dr. Carol Di Perri was supported by the Dr. Larry D. Jacobs Jog-for-the-Jake Fellowship. We also thank Eve Salczynski for technical support in the preparation of this manuscript.
References (51)
- et al.
Why voxel-based morphometry should be used
Neuroimage
(2001) - et al.
Statistical mapping analysis of lesion location and neurological disability in multiple sclerosis: application to 452 patient data sets
Neuroimage
(2003) - et al.
Detection of cortical lesions is dependent on choice of slice thickness in patients with multiple sclerosis
Int. Rev. Neurobiol.
(2007) - et al.
Threshold-free cluster enhancement: addressing problems of smoothing, threshold dependence and localisation in cluster inference
Neuroimage
(2009) - et al.
Advances in functional and structural MR image analysis and implementation as FSL
Neuroimage
(2004) - et al.
Pathologic findings of silent hyperintense white matter lesions on MRI
J. Neurol. Sci.
(1999) - et al.
The topography of white matter hyperintensities on brain MRI in healthy 60- to 64-year-old individuals
Neuroimage
(2004) - et al.
Short-term brain atrophy changes in relapsing-remitting multiple sclerosis
J. Neurol. Sci.
(2004) - et al.
FLAIR imaging for multiple sclerosis: a comparative MR study at 1.5 and 3.0 Tesla
Eur. Radiol.
(2006) - et al.
Improving interobserver variation in reporting gadolinium-enhanced MRI lesions in multiple sclerosis
Neurology
(1997)
Anatomical mapping of white matter hyperintensities (WMH): exploring the relationships between periventricular WMH, deep WMH, and total WMH burden
Stroke
Voxel-based assessment of differences in damage and distribution of white matter lesions between patients with primary progressive and relapsing-remitting multiple sclerosis
Arch. Neurol.
Lesion probability maps of white matter hyperintensities in elderly individuals: results of the Austrian stroke prevention study
J. Neurol.
Pathologic correlates of incidental MRI white matter signal hyperintensities
Neurology
Quantitative assessment of MRI lesion load in multiple sclerosis. A comparison of conventional spin-echo with fast fluid-attenuated inversion recovery
Brain
The effect of imprecise repositioning on lesion volume measurements in patients with multiple sclerosis
Neurology
Interscanner variation in brain MR lesion load measurements in multiple sclerosis using conventional spin-echo, rapid relaxation-enhanced, and fast-FLAIR sequences
AJNR Am. J. Neuroradiol.
Effect of training and different measurement strategies on the reproducibility of brain MRI lesion load measurements in multiple sclerosis
Neurology
Intraobserver and interobserver variability in measuring changes in lesion volume on serial brain MR images in multiple sclerosis
AJNR Am. J. Neuroradiol.
Assessment of in vivo 1H magnetic resonance spectroscopy in the liver: a review
Liver Int.
Lesion distribution in children with clinically isolated syndromes
Ann. Neurol.
The cognitive correlates of white matter abnormalities in normal aging: a quantitative review
Neuropsychology
Prevalence of white matter hyperintensities in a young healthy population
J. Neuroimaging
MR identification of white matter abnormalities in multiple sclerosis: a comparison between 1.5 T and 4 T
AJNR Am. J. Neuroradiol.
Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (EDSS)
Neurology
Cited by (24)
MAGNIMS recommendations for harmonization of MRI data in MS multicenter studies
2022, NeuroImage: ClinicalCitation Excerpt :Harmonization of the MRI acquisition Lesion number and volume should be assessed on images acquired on scanners with the same magnetic field strength (Di Perri et al., 2009; Stankiewicz et al., 2011). The automated detection of lesions can be improved by correcting for B1 inhomogeneity (Sajja et al., 2006).
DeepGRAI (Deep Gray Rating via Artificial Intelligence): Fast, feasible, and clinically relevant thalamic atrophy measurement on clinical quality T2-FLAIR MRI in multiple sclerosis
2021, NeuroImage: ClinicalCitation Excerpt :An internal dataset of MS (n = 3) and healthy control (n = 2) subjects scanned, repositioned, and then re-scanned over one week was employed. ( Di Perri et al., 2009; Dwyer et al., 2017) These subjects had substantial positional changes between imaging sessions, reflecting real-world levels of consistency in positioning for clinical scans. Both 3D T1w images and 2D low-resolution T2-FLAIR images were acquired in each scanning session.
Role of Neuroimaging in Multiple Sclerosis
2016, Multiple Sclerosis: A Mechanistic ViewComparison of brain 3.0-T with 1.5-T MRI in patients with multiple sclerosis: A 6-month follow-up study
2014, Clinical Neurology and NeurosurgeryCitation Excerpt :Additional spinal cord imaging had no effect on these results. MRI studies comparing two different MR scans to attempt to show the superiority of high-field strength for the diagnosis of MS have reported positive results, but the high-field strength did not lead to an earlier diagnosis of MS, and the impact on DIS was minor [2–7]. One study in which 3.0-T and 1.5-T MR scans were performed three times similarly to our study reported that only 1 of 40 patients with clinically isolated syndromes suggestive of MS showed DIT and DIS on 3.0-T MRI but not 1.5-T, leading to a diagnosis of definite MS [8].
Association of amygdala volumes with cortisol secretion in unipolar depressed patients
2012, Psychiatry Research - NeuroimagingCitation Excerpt :Both rMDD and fMDD cases were more frequently measured on 1.5 Tesla and controls more frequently on 3 Tesla instruments (χ² = 33.15; P = 0.000). Besides the apparent argument that a higher field strength reduces the signal to noise ratio and therefore enhances the image quality, studies comparing 1.5 Tesla with 3 Tesla also revealed an influence of field strength on brain morphologic measurement (Levy-Reis et al., 2000; Di Perri et al., 2009). Due to these findings and the fact that scanner type was unequally distributed in the patient versus control groups, field strength was taken into consideration as a factor in the multivariate analysis.
Neuroimaging of Multiple Sclerosis: An Update
2011, Neuroinflammation