Abstract
Objective
The objective of this article is to evaluate advanced techniques of diffusion-weighted imaging (DWI) and apparent diffusion coefficient (ADC) measurements of the optic nerve in patients with optic neuritis.
Methods
In this prospective and institutional review board-approved trial, we examined 15 patients with acute visual loss and clinical signs of optic neuritis including thin-slice multi-shot segmented readout of long variable echo trains (rs-EPI, RESOLVE) DWI and reduced field-of view DWI using a parallel transmit system (rFOV-EPI). Conventional single-shot echo-planar DWI (ss-EPI) of the whole brain was available in 13 patients. Subjective image quality was compared using a four-point scale and objective ADC measurements were performed in comparison with the non-affected side.
Results
In the intraorbital segment, subjective image quality was significantly higher in rFOV-EPI (score 3.3 ± 0.8) compared with rs-EPI (score 2.1 ± 0.8) and ss-EPI (score 0.9 ± 0.8). Diagnosis was hampered in the canalicular segment (n = 3) and the intracranial segment (n = 1) in all applied DWI techniques. ADC measurements of the affected side differed significantly in all DWI sequences ss-EPI (sensitivity 54%, accuracy 77%), rs-EPI (sensitivity 71%, accuracy 86%), and rFOV-EPI (sensitivity 73%, accuracy 87%).
Conclusion
Optic neuritis in the intraorbital segment can be detected with high sensitivity without the need for contrast application. Using rFOV-EPI improves subjective image quality compared with rs-EPI and ss-EPI. Due to its higher spatial resolution, rFOV-EPI was the preferred technique in our study and can ensure the diagnosis in the intraorbital segment. However, artefacts occur in the canalicular and intracranial segment of the optic nerve, therefore contrast-enhanced T1-weighted images must still be considered as the gold standard.
Keywords: DWI, optic neuritis, parallel transmit, pTX, zoomed EPI, RESOLVE
Introduction
Diffusion-weighted imaging (DWI) provides important information about ischaemic and inflammatory diseases in neuroradiology. Among others, it is increasingly used in patients with acute visual loss in order to assess abnormalities of the optic nerve, including optic neuritis (ON).1,2 ON is characterised by acute and progressive loss of vision and generally affects young, otherwise healthy patients (incidence five cases per 100,000 people per year, mean age 36 years, more than 70% of patients are women).3,4 The visual disturbance might be accompanied by painful eye movements and relative afferent pupillary reflex. ON is frequently related to multiple sclerosis and might be the first manifestation, although there are a wide range of underlying causes.5 A reliable imaging technique to obtain high-resolution data is of great clinical interest. Several studies report decreased molecular diffusion of the optic nerve in patients with ON1,6,7 using conventional single-shot echo planar imaging (ss-EPI). However, DWI of the optic nerve is very challenging and high spatial resolution as well as high signal-to-noise ratio is needed because the optic nerve is a relatively small structure. In theory, further technical improvements are expected to facilitate the detection of structural changes, localisation of the abnormality, and damage to the nerve.
In the recent literature, two technical approaches seem to overcome the limitations of ss-EPI. First, readout segmentation echo planar imaging (rs-EPI) is a promising technique in the assessment of small intracranial structures and has already been reported for diagnosis of ON in a recent study.8 The second promising technology is reduced field-of-view DWI (rFOV-EPI), also called ‘inner volume imaging’ or ‘zoomed DWI’. Zoomed DWI has been reported with improvement of image quality in several organ systems, amongst others the prostate,9 pancreas,10 liver,11 spinal cord12 and neck.13
The aim of this study was to evaluate the role of advanced DWI imaging in ON using a high-resolution rs-EPI and rFOV-EPI sequence.
Materials and methods
The study was approved by the institutional review board of the University Hospital of Tübingen. Fifteen consecutive patients (nine female, six male; median age 32 years, range 18–47 years) where enrolled. Inclusion criteria were acute visual deficits with clinical and ophthalmological signs of ON and the clinical indication for a magnetic resonance imaging (MRI) examination. MRI was performed on a clinically approved 3T Scanner (Magnetom Skyra, Siemens Healthcare, Erlangen, Germany) using a 12-channel head matrix coil.
After informed written consent, standard conventional MRI of the brain and orbits was performed including high-resolution rs-EPI (readout segmentation of long variable echo trains, RESOLVE, Siemens Healthcare) and rFOV-EPI (syngo ZOOMit, Siemens Healthcare) before contrast application.
The DWI sequences use the property that the diffusion motion causes a reduction of the T2 signal. A bipolar pair of strong gradients is inserted for dephasing and rephasing. The degree of signal loss depends on the molecular movement. The strength of diffusion weighting is controlled by the so-called b value. The ADC value can be determined by measuring different b values. Imaging parameters of the applied DWI sequences were as follows:
Conventional ss-EPI: Repetition time (TR) 4700 ms, echo time (TE) 98 ms, FOV 220 × 200 mm, voxel size 1.2 × 1.2 × 3 mm3, 20 slices, diffusion weightings two (b1 value 0 s/mm2, two averages, b2 value 1000 s/mm2, seven averages), time of acquisition (TA) 2:04 minutes.
rs-EPI: Details of the technique have already been described elsewhere.14 In brief, the k-space trajectory is divided into multiple segments in the readout direction to shorten the duration of the echo reading. The reduced TE and reading the signal in each segment allows a motion correction via a two-dimensional (2D) navigator to correct phase errors.14 The so-called EPI factor describes the number of echoes in the EPI readout.
The following imaging parameters were used: TR 5000 ms, TE1 67 ms and TE2 109 ms.
FOV 200 × 200 mm, voxel size 1.0 × 1.0 × 3 mm3, readout segments seven (number of shots), diffusion weightings two (b1 value 0 s/mm2, one average, b2 value 1000 s/mm2, one average), TA 3:07 minutes.
rFOV-EPI: Details of the technique have already been described elsewhere.15 In brief, the MR system contains two independent phase coherent radiofrequency (RF) channels which enable fully independent parallel transmission (pTX). As a result, totally free waveforms allow excitation of a multidimensional spatially selective small volume (which is called ‘inner volume’, ‘reduced FOV’, or ‘zoomed FOV’) in phase direction. As a result, no signal from the non-excited regions will occur. Using the pTX system therefore prevents the appearance of foldover artefacts that normally occur when the FOV is smaller than the object. The system is commercially known as syngo ZOOMit™ (Siemens Healthcare, Erlangen, Germany).
Imaging parameters were as follows: TR 1800 ms, TE 78 ms, FOV 130 × 42 mm, phase oversampling 30%, voxel size 0.7 × 0.7 × 3 mm3, diffusion weightings two (b1 value 0 s/mm2, five averages, b2 value 1000 s/mm2, 22 averages), TA 2:45 minutes.
Image evaluation
Qualitative image evaluation was performed by two experienced neuroradiologists in consensus. The level of confidence to diagnose a diffusion restriction and image quality were rated on a four-level scale (1 = poor image quality; 2 = fair image quality, anatomic borders moderately blurred; 3 = good, almost no blurring; 4 = excellent, sharply defined borders).
ADC measurements were performed in regions of interest (ROIs) that were placed within the optic nerve with the local picture archiving and communication system (PACS) system (Centricity PACS, GE Healthcare, Barrington, IL, USA). All ROIs were at least four voxels in size and were placed in the area of lowest ADC signal present. Areas affected by susceptibility artefacts or partial volume were excluded. The affected side was compared to a control region in the identical localisation of the healthy, non-affected optic nerve.
Standard of reference to confirm the diagnosis were the findings of fundoscopy, T2-weighted imaging (T2w), contrast enhancement as well as the clinical findings and the course of the patients. Patient subgroups were built based on the localisation of contrast enhancement within the orbital segment (globe to optic canal), the canalicular segment (within the optic canal), and the intracranial segment (optic canal to optic chiasm) in order to further analyse the degree of susceptibility artefacts due to magnetic field distortion near air-tissue interfaces.
Statistical analysis
Continuous data are presented as mean ± standard deviation. DWI was rated true positive if diffusion restriction was located in the same region as T2w hyperintensity and contrast enhancement and correlated to the symptoms of the patients. Visual assessment was compared using the Wilcoxon signed-rank test for comparing non-parametric matched samples. Quantitative ADC variables were tested for statistical significance by using a student t-test. Statistical significance was defined by p < 0.05; all p values were two tailed. For statistical analyses IBM SPSS Software (version 23; Armonk, NY, USA) was used.
Results
MRI could be performed in all patients; there were no adverse events. As a result of patient discomfort in the MR scanner, ss-EPI was not performed in two patients and rs-EPI was not performed in one patient. Susceptibility artefacts caused by the air-tissue interface near the sphenoid bone and sella hampered image evaluation and diagnostic accuracy in patients with ON in the canalicular segment (n = 3) and the intracranial segment (n = 1) in all applied DWI sequences. ADC measurements were therefore performed in the patient subgroup with manifestation of ON in the intraorbital segment (n = 11). The optic canal and intracranial segment is known to contain magnetic inhomogeneity and all techniques missed diffusion restriction in these segments compared with contrast-enhanced MRI.
In the orbital segment DWI enabled a distinct delineation of anatomic borders. Visibility of diffusion restriction was rated significantly higher in rFOV-EPI (subjective image quality score 3.3 ± 0.8) compared with rs-EPI (subjective image quality score 2.1 ± 0.8) and ss-EPI (subjective image quality score 0.9 ± 0.8), see Table 1. Figures 1 and 2 show representative examples of image quality in the orbital segment in a patient with ON.
Table 1.
Results of the qualitative image evaluation regarding visibility of diffusion restriction and image quality rated on a four-point scale.
| No patients | 15 |
| Median age (range in years) | 32 (18–47) |
| Sex (male/female) | 6/9 |
| Laterality (right/left) | 7/8 |
| Hyperintensity on T2w (all cases) | 15/15 (100%) |
| Contrast enhancement on T1w (all cases) | 14/15 (93%) |
| Hyperintensity on ss-EPI (all cases) | 7/13 (54%) |
| Hyperintensity on rs-EPI (all cases) | 10/14 (71%) |
| Hyperintensity on rFOV-EPI (all cases) | 11/15 (73%) |
| Hyperintensity on ss-EPI (intraorbital) | 7/10 (70%) |
| Hyperintensity on rs-EPI (intraorbital) | 10/11 (91%) |
| Hyperintensity on rFOV-EPI (intraorbital) | 11/11 (100%) |
| Visibility of diffusion restriction ss-EPI (intraorbital)a | 0.9 ± 0.8 |
| Visibility of diffusion restriction rs-EPI (intraorbital)a | 2.1 ± 0.8 |
| Visibility of diffusion restriction rFOV-EPI (intraorbital)a | 3.3 ± 0.8 |
Subjective image quality scores (mean ± SD) of ss-EPI, rs-EPI, and rFOV-EPI. T2w: T2-weighted imaging; T1w: T1-weighted imaging; ss-EPI: single-shot echo-planar diffusion-weighted imaging; rs-EPI: readout segmentation echo planar imaging; rFOV-EPI: reduced field of view echo planar imaging.
Figure 1.
A 36-year-old male with optic neuritis in the orbital segment of the right eye. Comparison of single-shot echo-planar diffusion-weighted imaging (ss-EPI) (a) and (b) and readout segmentation echo planar imaging (rs-EPI) (c) and (d) with reduced field of view echo planar imaging (rFOV-EPI) (e) and (f). Note the improved image quality and markedly reduced susceptibility distortions in the rs-EPI and rFOV-EPI sequences compared with ss-EPI.
Figure 2.
Diffusion-weighted imaging using single-shot echo-planar diffusion-weighted imaging (a)–(c), readout segmentation echo planar imaging (d)–(f), reduced field of view echo planar imaging (g)–(i) and corresponding T1-weighted post-contrast T1-weighted (j) and (k) images as well as T2-weighted images (l) in a 34-year-old patient with acute optic neuritis of the left eye.
In the subjective analysis seven included patients showed distinct diffusion restriction in ss-EPI (sensitivity 54%, accuracy 77%), 10 patients in rs-EPI (sensitivity 71%, accuracy 86%) and 11 patients using rFOV-EPI (sensitivity 73%, accuracy 87%). Regarding the orbital segment, sensitivity was high in all sequences, especially in rFOV-EPI all manifestations of ON could be detected (sensitivities ss-EPI 70%, rs-EPI 91%, rFOV-EPI 100%). Subjective image was superior in both rs-EPI compared with ss-EPI (p < 0.01) and rFOV-EPI compared with ss-EPI (p < 0.01). All MR diagnoses could be confirmed in the clinical course, and there were no disagreements between MRI findings and clinical symptoms.
Quantitative ADC measurements in the orbital segment were significantly different between the affected and the non-affected side in all sequence techniques. The average ADC of the optic nerve on the healthy side compared with the affected side was: ss-EPI 904 ± 251 versus 1246 ± 190; rs-EPI 801 ± 149 versus 1153 ± 222; rFOV-EPI 696 ± 169 versus 1090 ± 180. The level of significance was high in all applied sequences (p < 0.01) with the best rating for rs-EPI and rFOV-EPI, which were the preferred imaging techniques in our study.
Discussion
Despite standard T2w sequences and post-contrast T1-weighted (T1w) sequences, DWI has been reported to add additional information in evaluating the extent of damage of the optic nerve and predict changes of recovery. Fatima et al.16 reported significant differences in ADC maps in a retrospective analysis of patients with acute ON, patients with chronic ON, and controls using conventional single-shot EPI DWI. The ADC values of patients in the acute ON group were 900 ± 120, which is in concordance with our results of the ss-EPI sequence (904 ± 251) and even lower ADC values using the advanced high-resolution techniques rFOV-EPI and rs-EPI. Fatima et al. concluded that contrast enhancement is not always necessary in patients with acute ON. Lu et al. also reported positive results of the visualisation of ON in a recently published study using coronary slice planning.17
However, DWI of the optic nerve is known to be challenging due to the small diameter of the nerve, the signal of the surrounding cerebrospinal fluid (CSF), the motion of the nerve, and the magnetic susceptibility changes caused by the bony cavity of the optic canal and the adjacent sinuses. Conventional ss-EPI sequences sample the entire k-space in a single TE after an RF excitation pulse. Due to its long readout, it is prone to susceptibility artefacts at tissue interfaces, B0-inhomogeneities, and chemical shift between fat and water. Application of ss-EPI in orbital imaging has therefore been limited in the past. In this study, we evaluated new technical approaches of 3-mm thin-section DWI of the optic nerve in patients with ON.
Combining several sub-sampled k-space acquisitions (multi-shot) enables reducing echo spacing in order for faster image acquisition and reduction of susceptibility artefacts.18 The rs-EPI sequence divides the k-space trajectory into multiple segments in the readout direction. As a result, TE and encoding times can be reduced and motion correction can be performed using a 2D navigator correcting motion-induced, non-linear phase errors.14,18 A disadvantage is that the overall acquisition time is longer because multiple TR intervals are required for data acquisition. The feasibility of assessment pathologies in the optic nerve has already been shown.8,19 In our study, rs-EPI showed superior image quality compared with ss-EPI in patients with ON in the intraorbital segment. The high diagnostic performance of rs-EPI is in concordance with a recent study by Wan et al., who even reported a similar sensitivity and specificity of rs-EPI compared with contrast-enhanced T1w imaging.8
Another promising innovation is reducing the FOV in order to increase spatial resolution. The idea of acquiring spatially 2D-selective pulses reducing the FOV was implemented in DWI of the optic nerve in 2002.20 However, due to the hardware-conditions at that time, the duration of zoomed multislice EPI along the phase-encoding direction was >9 minutes for four slices and consequently >28 minutes for the acquisition of the whole orbit. The recently introduced independent parallel radiofrequency transmit coils allow free waveforms switching of arbitrary gradient shapes with extended dynamic RF excitation schemes for excitation of shaped volumes with reduced FOV15 (syngo ZOOMit, Siemens, Erlangen, Germany). This results in the possibility of excitation small FOVs (called ‘inner-volume’ or ‘zoomed FOV’) in the phase-encoding direction without aliasing-artefacts that normally occur when the FOV is smaller than the object. The reduced FOV approach can be applied in any exam that is focused on a specific anatomic area. Future work of applying multiple gradients may allow calculation of full diffusion tensor in the optic nerve. In the recent literature, an increasing number of reduced FOV DWI applications with reduced blurring and higher spatial resolution than ss-EPI have been reported, amongst others of the prostate,9 pancreas,10 spinal cord,12 neck13 and liver imaging.11 The required number of k-space lines is decreased by reducing the FOV. The significant improvement of spatial resolution in combination with thin slices (3 mm) that are aligned parallel to the optic nerve show a definite improvement of both subjective and objective image quality. Diffusion restriction that might have been missed due to the small size of the optic nerve in the past can now be detected. It overcomes the major problem of low specific absorption rate and low spatial resolution facing DWI on the optic nerve. The high spatial resolution is necessary because even a small amount of partial volume effects of surrounding CSF will increase the measured diffusion coefficient. For this reason, the rFOV-EPI could contribute to a more accurate ADC measurement without extending the measuring time.
The values of ADC we obtained in patients with ON were significantly lower than the contralateral side in the ROI analysis, with the best results using rFOV-EPI. Limitations of all applied techniques were artefacts due to the air-filled sinus and bone, which has already been reported in other DWI techniques and in reduced FOV T1w and T2w Sampling Perfection with Application optimised Contrasts using different flip angle Evolution (SPACE) imaging of the orbits.21 No false-positive results were found in DWI. We therefore state rFOV-EPI as the best available DWI technique in detecting diffusion restriction of the optic nerve in patients with suspected ON. The small FOV of the rFOV-EPI allows a high resolution. rs-EPI can also add to this diagnosis and has advantages due to its availability and the coverage of the brainstem and optic tract. However, the accuracy of contrast-enhanced T1w imaging was higher than that of all DWI techniques. Other applications of the DWI of the orbit can also benefit from the use of rFOV-EPI. To date, DWI has been used to differentiate between benign and malignant changes as well as in therapy monitoring, e.g. in the case of radiotherapy.22–24
Limitations of the study are the small number of samples and the fact that the readers were not blinded with regard to clinical diagnosis in the absence of a control group. The placement and size of the ROI was also a subjective assessment, and coronary image data were not available. Further studies could investigate the value of a threshold-based analysis. Nevertheless, the procedure used here is established in everyday clinical practice and represents the classical diagnostic procedure.
In conclusion, the improved imaging techniques rs-EPI and rFOV-EPI enable high-quality imaging in the orbital segment of the optic nerve. Both rs-EPI and rFOV-EPI are suitable techniques for assessment of diffusion restriction and provide significant image quality improvement compared with ss-EPI. Due to its higher spatial resolution, rFOV-EPI is the preferred technique in our study and can ensure the right diagnosis without the need of contrast application. However, artefacts occur in the canalicular and intracranial segment of the optic nerve. Therefore contrast-enhanced T1w images must still be considered as the gold standard in this location in patients with suspected ON.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Conflict of interest
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
ORCID iD
Achim Seeger http://orcid.org/0000-0002-4585-649X.
References
- 1.Hickman SJ, Wheeler-Kingshott CA, Jones SJ, et al. Optic nerve diffusion measurement from diffusion-weighted imaging in optic neuritis. AJNR Am J Neuroradiol 2005; 26: 951–956. [PMC free article] [PubMed] [Google Scholar]
- 2.Bender B, Heine C, Danz S, et al. Diffusion restriction of the optic nerve in patients with acute visual deficit. J Magn Reson Imaging 2014; 40: 334–340. [DOI] [PubMed] [Google Scholar]
- 3.Rodriguez M, Siva A, Cross SA, et al. Optic neuritis: a population-based study in Olmsted County, Minnesota. Neurology 1995; 45: 244–250. [DOI] [PubMed] [Google Scholar]
- 4.Wilhelm H, Schabet M. The diagnosis and treatment of optic neuritis. Dtsch Arztebl Int 2015; 112: 616–625. quiz 626. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Atkins EJ, Biousse V, Newman NJ. Optic neuritis. Semin Neurol 2007; 27: 211–220. [DOI] [PubMed] [Google Scholar]
- 6.Iwasawa T, Matoba H, Ogi A, et al. Diffusion-weighted imaging of the human optic nerve: a new approach to evaluate optic neuritis in multiple sclerosis. Magn Reson Med 1997; 38: 484–491. [DOI] [PubMed] [Google Scholar]
- 7.He M, Cestari D, Cunnane MB, et al. The use of diffusion MRI in ischemic optic neuropathy and optic neuritis. Semin Ophthalmol 2010; 25: 225–232. [DOI] [PubMed] [Google Scholar]
- 8.Wan H, Sha Y, Zhang F, et al. Diffusion-weighted imaging using readout-segmented echo-planar imaging, parallel imaging, and two-dimensional navigator-based reacquisition in detecting acute optic neuritis. J Magn Reson Imag 2016; 43: 655–660. [DOI] [PubMed] [Google Scholar]
- 9.Rosenkrantz AB, Chandarana H, Pfeuffer J, et al. Zoomed echo-planar imaging using parallel transmission: impact on image quality of diffusion-weighted imaging of the prostate at 3T. Abdom Imaging 2015; 40: 120–126. [DOI] [PubMed] [Google Scholar]
- 10.Thierfelder KM, Sommer WH, Dietrich O, et al. Parallel-transmit-accelerated spatially-selective excitation MRI for reduced-FOV diffusion-weighted-imaging of the pancreas. Eur J Radiol 2014; 83: 1709–1714. [DOI] [PubMed] [Google Scholar]
- 11.Rosenkrantz AB, Geppert C, Kiritsy M, et al. Diffusion-weighted imaging of the liver: Comparison of image quality between monopolar and bipolar acquisition schemes at 3T. Abdom Imaging 2015; 40: 289–298. [DOI] [PubMed] [Google Scholar]
- 12.Seeger A, Klose U, Bischof F, et al. Zoomed EPI DWI of acute spinal ischemia using a parallel transmission system. Clin Neuroradiol 2016; 26: 177–182. [DOI] [PubMed] [Google Scholar]
- 13.Riffel P, Michaely HJ, Morelli JN, et al. Zoomed EPI-DWI of the head and neck with two-dimensional, spatially-selective radiofrequency excitation pulses. Eur Radiol 2014; 24: 2507–2512. [DOI] [PubMed] [Google Scholar]
- 14.Porter DA, Heidemann RM. High resolution diffusion-weighted imaging using readout-segmented echo-planar imaging, parallel imaging and a two-dimensional navigator-based reacquisition. Magn Reson Med 2009; 62: 468–475. [DOI] [PubMed] [Google Scholar]
- 15.Blasche MR, Lichy M, Riffel P. TimTX TrueShape and syngo ZOOMit technical and practical aspects. Magnetom Flash 2012; 1: 74–84. [Google Scholar]
- 16.Fatima Z, Motosugi U, Muhi A, et al. Diffusion-weighted imaging in optic neuritis. Can Assoc Radiol J 2013; 64: 51–55. [DOI] [PubMed] [Google Scholar]
- 17.Lu P, Sha Y, Wan H, et al. Role of coronal high-resolution diffusion-weighted imaging in acute optic neuritis: A comparison with axial orientation. Neuroradiology 2017; 59: 737–745. [DOI] [PubMed] [Google Scholar]
- 18.Cohen-Adad J. High-resolution DWI in brain and spinal cord with syngo RESOLVE. Magnetom Flash 2012; 2: 16–23. [Google Scholar]
- 19.Däubler BF, Kamli A, Runge VM. Readout-segmented diffusion-weighted imaging in a critical anatomic area – diagnosing posterior ischemic optic neuropathy (PION). RoFo 2014; 186: 1151–1152. [DOI] [PubMed] [Google Scholar]
- 20.Wheeler-Kingshott CA, Parker GJ, Symms MR, et al. ADC mapping of the human optic nerve: increased resolution, coverage, and reliability with CSF-suppressed ZOOM-EPI. Magn Reson Med 2002; 47: 24–31. [DOI] [PubMed] [Google Scholar]
- 21.Seeger A, Schulze M, Schuettauf F, et al. Feasibility and evaluation of dual-source transmit 3D imaging of the orbits: comparison to high-resolution conventional MRI at 3T. Eur J Radiol 2015; 84: 1150–1158. [DOI] [PubMed] [Google Scholar]
- 22.Razek AA. Diffusion-weighted magnetic resonance imaging of head and neck. J Comput Assist Tomogr 2010; 34: 808–815. [DOI] [PubMed] [Google Scholar]
- 23.Bhatt N, Gupta N, Soni N, et al. Role of diffusion-weighted imaging in head and neck lesions: pictorial review. Neuroradiol J 2017; 30: 356–369. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Soni N, Gupta N, Kumar Y, et al. Role of diffusion-weighted imaging in skull base lesions: a pictorial review. Neuroradiol J 2017; 30: 370–384. [DOI] [PMC free article] [PubMed] [Google Scholar]