1. Introduction
The first human radiograph ever taken was an image of Anna Bertha Roentgen’s left hand [
1] (
Figure 1). It showed remarkable contrast between bone, soft tissue and the metal rings on her ring finger. Contrast on medical images is the difference in brightness (or signal) between two tissues, fluids or materials. The contrast first demonstrated by Wilhelm Roentgen on his wife’s hand in 1895 had never been seen before and it became central to the practice of radiology.
In plain radiographs of the head, contrast could be seen between bone of different thickness, but the soft tissue of the brain itself could not specifically be seen (
Figure 2). If air was substituted for CSF, contrast could be created between the ventricular system (now containing air rather than CSF) and the surrounding brain. This technique, which was developed by Walter Dandy in 1918–1919, was used diagnostically to recognize disease of the brain by displacement or distortion of the ventricular system and/or subarachnoid space in air studies of the brain [
2,
3] as shown in
Figure 3.
Another method of creating contrast in radiographs of the brain was to inject sodium iodinate into the cerebral arteries and so create contrast between large vessels and the surrounding tissues as demonstrated by Egas Moniz in 1927 [
4] and shown in
Figure 4. The technique was called angiography. Obstruction of vessels, displacement of these and development of new vessels could be seen and were used in the diagnosis of disease.
Both air studies and angiography required the introduction of a contrast agent to see disease and were only indirect methods of seeing the brain. The brain itself was not visualized with either technique. This was a problem with diseases such as multiple sclerosis (MS), where abnormal tissue could be present in the brain but not displace the ventricular system or produce vascular abnormalities. As a result, radiological diagnosis of the disease was not possible.
The first major revolution in soft tissue contrast in medical imaging came in 1971 with the introduction of brain computed tomography (CT) [
5,
6]. The technique not only produced slices of the brain that were much more useful than conventional tomography, but it directly displayed brain tissue. Within the brain, there was high intrinsic contrast between normal and abnormal tissues so that lesions such as the glioma shown in
Figure 5 could readily be seen because its signal was lower than that of normal brain. The intrinsic contrast was present without requiring the use of a contrast agent (such as air and iodinated compounds). When intravenous iodinated contrast agents were used with CT, additional information was obtained. CT transformed the practice of neuroradiology from 1971 onwards, and body imaging from 1975 onwards as well (
Figure 6).
The second major revolution in soft tissue contrast imaging came in 1981. While CT had shown high soft tissue contrast in the brain and elsewhere relative to plain radiographs, it was possible to obtain even higher soft tissue contrast with magnetic resonance imaging (MRI). This was seen in ten cases of MS in which 19 lesions were depicted with CT but 112 more lesions were seen with MRI using T
1 dependent inversion recovery (IR) sequences [
7] (
Figure 7). The additional lesions seen with MRI were smaller than those seen with CT and were demonstrated in white matter that appeared normal on CT. Even though MRI was much slower and of lower spatial resolution than CT, its superior soft tissue contrast resulted in a decisive advantage from a clinical point of view.
The contrast advantage of MRI was extended to included T
2 dependent spin echo sequences in 1982 [
8,
9] (
Figure 8) and further improvements in the demonstration of soft tissue contrast with MRI came with short inversion time IR (STIR) (
Figure 9), diffusion-weighted, susceptibility-weighting (
Figure 10) and T
2-fluid attenuated inversion recovery (T
2-FLAIR) (
Figure 11) sequences [
10,
11,
12,
13]. In addition, from 1984 onwards Gadolinium based contrast agents (GBCAs) were used to create additional contrast in particular clinical situations [
14,
15] as shown in
Figure 12. Imaging protocols using multiple sequences of these and other types have been established for different clinical applications and form the basis for modern MRI examinations of the brain and other organs of the body.
However, there is evidence from postmortem studies that normal appearing tissues seen with MRI using present day protocols may actually be abnormal, and efforts have been made with techniques such as magnetization transfer and magnetic resonance spectroscopy (MRS) to demonstrate these abnormalities. These methods have not been successful enough for them to be included in most contemporary clinical imaging protocols.
At the present time, a third revolution in soft tissue contrast imaging is taking place in which normal appearing tissues seen with state-of-the-art MRI sequences show abnormalities with very high contrast when imaged with newer sequences such as divided subtracted inversion recovery (dSIR). In modelling studies, the dSIR sequence can show ten times more contrast than conventional IR sequences when imaging small changes in T1 due to disease. The small changes in T1 may be insufficient to generate contrast with conventional sequences but the much greater contrast amplification of the dSIR sequence can show obvious abnormalities.
The purpose of this paper is to describe the theory underlying the use of the dSIR sequence, and show how this sequence can make obvious a pattern of brain injury not previously described in MRI (the whiteout sign). This sign can be seen after different insults to the brain when little or no abnormality is apparent with conventional state-of-the-art sequences.
5. Discussion
5.1. Ultra-High Contrast MRI
In general terms, it is possible to describe the soft tissue contrast produced by CT as high in relation to the soft tissue contrast seen on plain radiographs. The increased soft tissue contrast of MRI can then be described as very high, and that of recently developed MRI sequences such as dSIR as ultra-high (
Table 3). This follows the naming of the radiofrequency spectral bands as high frequency (3–30 MHz), very high frequency (30–300 MHz) and ultra-high frequency (300–3000 MHz). Since the speed of radiowaves is 3 × 10
8 m/s, the radiofrequency bands can also be described in terms of wavelengths, i.e., high frequency (100 m–10 m), very high frequency (10 m–1 m) and ultra-high frequency (1 m to 1 mm).
This division also generally follows the classification of static field strengths used in clinical MRI which at the present time is low (less than 1 T with ultra-low < 0.1T), high (1–3 T) and ultra-high (7–11.7 T now, and possibly 14 T in the future). At 7 T, the proton resonance frequency is 300 MHz, which is approximately the operating frequency of a 7 T MRI system. As a result, the boundary between high frequency and ultra-high frequency radiowaves is essentially the same as that between high field and ultra-high field strength MRI systems.
The transition in contrast from high to very high contrast (CT to MRI) and from very high to ultra-high contrast (conventional MRI to dSIR and other forms of MRI) is typically manifest as normal appearing tissue seen with the first imaging modality showing unmistakable high contrast abnormalities with the second imaging modality. This is shown for the transition from conventional MRI to dSIR imaging in
Figure 18 and
Figure 20, where no abnormality is seen in white matter on the T
2-FLAIR images, but extensive high signal changes are seen in white matter on the dSIR images.
A principal mechanism for increases in intrinsic contrast in MRI is synergistic contrast, in which: (i) a single tissue property such as T1 is used twice or more in the same sequence to increase contrast, (ii) two or more different tissue properties such as T1 and T2 are used together to increase contrast, and (iii) both techniques (i) and (ii) are used in a single sequence.
An example of the use of T1 and T2 to create synergistic contrast is the STIR sequence, where increases in both T1 and T2 result in additive contrast. Another example is diffusion-weighted (D*-weighted) imaging with a long TE sequence. For an increase in T2 and a decrease in D*, the T2 and D* contrast is synergistic and overall contrast is increased. If, as is common, both T2 and D* are increased in disease then the contrast produced by each of these tissue properties is opposed and the overall result is a reduction in contrast, which is not usually clinically helpful.
With a single tissue property such as T
1, it is possible to approximately double the contrast produced by small changes in T
1 compared to a conventional IR sequence using subtraction of one image from the other (
Figure 13). However, more dramatic improvement in contrast comes with division of the subtracted image by the sum of the two images (
Figure 14). With this approach, the increase in contrast relative to the conventional IR sequences may be ten or more times. Normalization by the local signal intensity removes coil shading. Mobile proton density and T
2-weightings are also removed by normalization leaving only magnified T
1 contrast. The size of the increase in contrast is eventually limited by signal-to-noise (SNR) and artefact constraints.
The effect of GBCAs in shortening T1 can also be amplified using divided reverse subtracted (drSIR) sequences. This allows ultra-high contrast to be generated by small reductions in T1 produced by the GBCAs which are insufficient to produce contrast with conventional T1-weighted IR sequences.
dSIR sequences can usually be implemented on existing MR machines using IR pulse sequences that already exist on MRI systems. Ultra-high contrast can often be achieved at the same spatial resolution as conventional techniques. This may require some increase in time but typical dSIR acquisitions can readily be performed in less than five minutes on most MRI systems. The software required for the IR image manipulation necessary to produce dSIR sequences only involves basic arithmetic and is easily written in MATLAB (Natick, MA, USA).
5.2. The Whiteout Sign
The whiteout sign shows a bilateral, symmetrical and generally uniform increase in signal in white matter of the cerebral and cerebellar hemispheres. This increase in signal is relative to the normal signals seen in white matter. The normal anterior and posterior central corpus callosum has a very low signal. The corticospinal tracts and their surrounds, as well as the superior longitudinal fasciculi have higher signal than white matter elsewhere in the hemispheres. The increase in signal with the whiteout sign is superimposed on this normal pattern.
The whiteout sign usually shows sparing or relative sparing of the anterior and posterior central corpus callosum (i.e., the genu and splenum of the corpus callosum) as well as adjacent white matter (forceps minor and forceps major). There is usually also sparing or relative sparing of the peripheral white matter of the cerebral hemispheres.
The recovery or reversal phase of the whiteout sign (i.e., return towards normal) proceeds from the anterior and posterior central corpus callosum outwards, the periphery of the cerebral hemisphere white matter inwards, and the central regions of the cerebellar hemisphere white matter outwards.
There may be small focal or multifocal high signal changes present with other features of the whiteout sign. These are made more obvious by having low signal (more normal) white matter around them in, for example, the peripheral white matter of the cerebral hemispheres.
There are differences in the degree of the whiteout sign both in the increase in signal and the extent of the changes within white matter. Qualitative assessment may be by both by the degree of increase in the signal and the extent of the abnormality as a proportion of the total white matter seen in a slice in the cerebral or cerebellar hemispheres.
A grading system can be used to describe these changes (
Table 2). The system used to date has five grades 1 (normal) to 5 (maximum abnormal). The normal grade is low signal (dark appearance). The central corpus callosum and peripheral white matter in the cerebral hemispheres are usually lowest signal within the generally low signal pattern. The centrum semiovale is slightly higher signal as are the cerebellar hemispheres (
Table 3). The maximum abnormal grade is characterized by high signal (light appearance) with partial or complete loss of the high signal boundary between white and gray matter. Use of the grading system is illustrated in
Figure 16,
Figure 17,
Figure 18,
Figure 19,
Figure 20 and
Figure 21.
In preliminary studies, a cutoff point between grades 1 and 2 as normal and grades 3,4,5 as abnormal correlated very closely with clinical assessment of cognitive impairment in patients with mTBI and appears likely to be useful in clinical practice.
Quantitation of T1. dSIR images are T1 maps and in the mDs signal is nearly linearly proportional to T1 (Equation (2)). It is therefore possible to calibrate the display gray scale in terms of T1. This has the advantages over conventional T1 mapping in that it requires no extra acquisition, it is at the same spatial resolution as the clinical image (since it is the clinical image) not the lower spatial resolution generally used for T1 maps, and no registration procedure is required to align the clinical images and the T1 maps.
It is possible both to directly read values of T1 and to measure differences in T1 between regions on dSIR images and determine how these correlate with visible contrast. This applies both to normal and abnormal white matter. The values of T1 do not apply outside of the mD, so values for gray matter are in error when white matter is targeted with dSIR sequences. Also, the values of T1 assume full relaxation with a long TR. If TR is less than that needed to allow full relaxation of a tissue, or if a shorter duration TR is desired, a more general expression for T1 recovery than those used in Equations (1) and (2) may be used to calculate suitable TIs. Quantitation of both increase in T1 and the extent of the abnormality can be combined.
The whiteout sign may be seen in acute, recurrent and chronic disease with natural histories including complete reversal as well as persistence. It may be seen together with signs of other diseases such as infarction, hemorrhage and small vessel disease.
When the whiteout sign is seen, there is usually little or no abnormality present in the corresponding regions in the brain using conventional T2-weighted images such as T2-wFSE and T2-FLAIR sequences. To date, the whiteout sign has only been shown with dSIR sequences and these are described in more detail in the next section.
5.3. dSIR Sequences
The dSIR sequence is one of the class of Multiplied Added Subtracted and/or Divided Inversion Recovery (MASDIR) sequences. It employs a subtraction of signal from a longer TI image from a shorter TI image. This is divided by the sum of the two images. As a consequence, the dSIR sequence is very largely independent of mobile proton density (rm) and T2. It is essentially a pure measure of T1. The relationship between signal and T1 is very closely linear within the mD. Thus, images can be regarded as quite accurate T1 maps in the mD while outside the mD signals fade to mid-gray (zero) as T1 is decreased or increased. The sequences can be used in 2D and 3D (isotropic or anisotropic) forms. The data collection may be FSE or gradient echo (GE), both usually with short TEs to increase SNR. The value of TE/TR is not critical since the T1 information is determined by the difference in TI.
The dSIR sequence is used in targeted form. The first TI is typically chosen to null the shortest T
1 in the normal white matter of interest. The second TI is chosen to accommodate small increases in the T
1 in normal white matter and keep these within the mD as illustrated in
Figure 14C. The sequence is highly sensitive to small increases in T
1 and produces much greater contrast from those changes in T
1 than conventional MP-RAGE sequences. For changes in T
1 outside the mD, the sensitivity to change is generally much less, and the contrast may “overshoot” and be reduced when there are increases in T
1 that take the change in T
1 outside of the mD.
There are high signal lines which have an “etched” appearance between white matter and gray matter as well as between white matter and CSF. These high signal lines show the extent of white matter in the cerebral and cerebellar hemispheres.
The boundaries arise because partial volume effects between two tissues (or a tissue and a fluid) with different T
1s in mixed voxels produce T
1s which correspond to the values of T
1 required to produce the maximum signal of the T
1-bipolar filters shown on
Figure 14C and
Figure 15. If the T
1 and signal of white matter is increased, then the contrast between the abnormal white matter and the high signal may be reduced so the boundary appears less obvious.
It is possible to create synthetic dSIR images from T1 maps using T1-BipoLAr fIlteRs (T1-BLAIRs). These synthetic dSIR images may have essentially any chosen values of TI and may be targeted at particular tissues and specific changes in the T1 of these tissues in disease. It is also possible to create synthetic narrower mD dSIR images from wider mD dSIR images.
Synthetic TP-bipolar filters may also be applied to other TP maps in addition to T1 maps such as T2, T2*, D* and χ to create synthetic T2-BLAIR, T2*-BLAIR, D*-BLAIR and χ-BLAIR images. Two or more synthetic TP-BLAIR images may also be multiplied together to create synthetic multi-TP contrast, e.g., T1, T2-BLAIR images. The term T1-BLAIR can be used generally to include both directly acquired dSIR and drSIR images as well as synthetic versions of these images, both with T1 as the tissue property. Synthetic TP-BLAIR with other TPs such as T2, T2* and D* are also included in the generic category of TP-BLAIR images.
5.4. Post-Insult Leukoencephalopathy Syndromes (PILS)
Post-insult leukoencephalopathy is the name given to syndromes in which, following an insult to the brain, white matter changes are seen in a whiteout sign as shown in Cases 1 and 2 in this paper.
The causes of the syndrome observed to date have been methamphetamine use disorder in this paper and reference [
17], mTBI (acute in this paper, and recurrent in reference [
18]) as well as Grinker’s myelinopathy (delayed post-hypoxic leukoencephalopathy [
19]. In spite of their disparate etiologies, a stereotypical pattern of change in the white matter has been seen (i.e., the whiteout sign) in these conditions [
20]. Other possible causes of the syndrome include other drugs, e.g., opiates, and post-viral infections, e.g., long COVID.
The severity, forms of injury and time course may all be important in inducing the whiteout sign. There may also be individual differences in patient susceptibility to insults that may result in the whiteout sign.
The symptoms and signs of the PILS may vary with the cause and severity of the insult. These may also affect the time course of symptoms and signs. Cognitive impairment is a feature and its time course may parallel the onset and remission of the whiteout sign.
5.5. Pathophysiology and Pathology
The bilateral symmetrical features with mainly uniform signal favor generalized pathological processes as the origin of the whiteout sign. These processes include neuroinflammation, particularly in the acute phase, but also chronically. Demyelination and degeneration may also be important. The rapid reversibility in acute cases favors an edematous and/or inflammatory process. It is likely that more than one pathophysiological process is involved. Histological validation of the appearances of the whiteout sign is likely to require animal studies, as is the study of the evolution of the sign over time.
5.6. Normal Appearing White Matter
This term is applied to white matter which appears normal with conventional MRI sequences, but there may be suspicions that this white matter is actually abnormal. Methods of trying to demonstrate this include magnetization transfer, diffusion and proton spectroscopy, none of which have achieved an established role in routine clinical diagnosis of suspected white matter abnormalities. The dSIR approach differs in that changes in a tissue property used in routine clinical diagnosis, i.e., T1 (as with MP-RAGE sequences) are used to demonstrate abnormalities in normal appearing white matter not a different tissue property.
5.7. Validation
In the absence of pathological verification, validation of the imaging findings is indirect and is based on:
The theory is summarized in
Figure 13 and
Figure 14. This provides a consistent account of the contrast seen on dSIR images.
Normal controls which show low signal in a characteristic pattern in normal white matter with dSIR sequences as illustrated in the two normal controls included in this paper. Also, the patient can in effect act as her/his own control when the abnormal changes revert to normal or near normal as shown in the two cases included in this paper.
Boundaries between white matter and gray matter, and between white matter and CSF. These are a distinct feature of dSIR images and can be explained using the model in
Figure 15. They are always seen in normal subjects and patients when the dSIR sequence is correctly performed. Boundaries of this type have not been seen with any other sequence. This supports the validity of the model described using
Figure 13 and
Figure 14.
There is also consistency between lesions with large changes in T2 (as well as T1) seen on T2-FLAIR images in the same location as changes on dSIR images. The features include high signal boundaries and the lesions on dSIR images as predicted by them.
The similar appearances of whiteout signs despite the different insults causing them suggests that there is a common pathophysiological process underlying them.
A phantom containing solutions with known T
1s was imaged with IR FSE sequences with TIs ranging from 24 to 1024 ms. Measurements of signal taken with dSIR sequences with TIs of 124 and 524 ms are shown (
Figure 22). These values showed very close agreement with the theoretical dSIR curves shown in
Figure 14C and
Figure 15 with correspondence to the values of T
1 of the solutions [
21]. These quantitative results support the validity of the numerical simulations shown in
Figure 13 and
Figure 14.
There has been a similarity of appearances on dSIR images obtained on GE, Philips and Siemens images at both 1.5 and 3.0 T, so the findings are not machine or field strength specific.
5.8. Multiple Sclerosis (MS)
In multiple sclerosis (MS), both focal features and generalized abnormalities have been observed [
16] but, to date, these do not usually have the bilateral symmetry and generally uniform signal with sparing in a specific pattern that is seen with the whiteout sign.
It is possible that MS may have been precipitated by an immune response following an initial Epstein–Barr viral infection. However, significant time may have elapsed between this event and the MR examination. During this time, there may also have been exacerbations and remissions at different times and in different parts of the brain breaking up the uniform pattern seen with the whiteout sign.
The focal and multifocal features are usually more prominent in MS and may follow other particular distributions, e.g., along cerebral veins in the form of Dawson’s fingers. The lesions seen in MS tend to be focal, signal overshoot is common and apparent “outpouching” of the ventricular system is seen when the high signal boundary around the ventricular system is lost at the site of an adjacent MS lesion.
5.9. Diffuse Disease in White Matter Seen with Conventional MRI Sequences
The whiteout sign is usually not associated with abnormalities seen on T2-wFSE or T2-FLAIR, but there are situations where bilateral and often symmetrical changes may be seen with conventional sequences. The relation of these to dSIR imaging of the same patients is a topic of considerable interest.
Grinker’s myelinopathy or delayed post-hypoxic leukoencephalopathy may produce extensive white matter changes that are obvious on T
2-wFSE and/or T
2-FLAIR images. The condition is thought to be rare. The whiteout sign seen in post-hypoxic patients on dSIR images may actually be a less severe form of the changes seen in Grinker’s myelinopathy in which changes are not seen with T
2-wFSE and T
2-FLAIR images [
19]. It is therefore possible that Grinker’s myelinopathy is much more common than usually thought, but is not recognized because the changes in white matter are usually insufficient to produce diagnostic contrast with conventional sequences.
Posterior reversible encephalopathy syndrome (PRES) is a condition induced by a variety of different insults [
22]. There are usually bilateral symmetrical changes in the posterior white matter of both hemispheres but changes may be seen elsewhere in the brain. dSIR imaging could show more extensive changes.
Diffusely abnormal white matter (DAWM) in MS. This is a condition in which, using conventional sequences, diffusely abnormal increase in signal is seen in up to 25% of cases of MS [
23]. These patients may show a more generalized pattern of abnormality than the MS patients examined to date with dSIR sequences.
The leukodystrophies. These often show diffuse abnormal high signal in white matter. The distribution in metachromatic leukodystrophy is similar to that seen with the whiteout sign but these are not reversible. Normal appearing white matter in the leukodystrophies may show abnormalities with dSIR sequences.
5.10. Clinical Value
Showing the whiteout sign as evidence of brain disease, using dSIR sequences may be of importance in distinguishing organic and psychological origins of disease. The demonstration of brain changes may also be of prognostic value and importance in monitoring the effects of treatment in a variety of different conditions.
5.11. Amplified MRI (aMRI)
Amplified MRI in which a video of brain motion with arterial pulsation is recorded and the displacement is amplified ten or more times is another example of ultra-high contrast MRI [
24]. The tissue motion is not apparent with conventional imaging, but is obvious with aMRI.
5.12. Advances in Ultra-High Contrast MRI
Short term technical developments include implementation of 3D forms of dSIR which are likely to require different TIs than the 2D dSIR sequence for this study. A 1 mm3 or smaller isotropic 3D imaging will facilitate imaging of the cerebral cortex as well as the central gray matter and brainstem. It will also facilitate GBCA studies and other serial examinations in which rigid body registration can be used.
There are also technical advances in AI, image noise reduction and registration which may improve the quality of dSIR images.
Additional imaging dSIR images can be created synthetically from T1 maps. It is also possible to synthetically create narrower mD images from wider mD images. The T1 maps may be produced by MR fingerprinting and other methods such as actual flip angle imaging which do not involve an IR sequence so that the more general term T1-BipoLAr fIlteR (T1-BLAIR) imaging may be preferred to describe imaging utilizing a synthetic T1-bipolar filter. The term T1-BLAIR includes direct acquisitions (such as with dSIR) as well as synthetic imaging utilizing T1 maps generated in different ways.
5.13. Ultra-High Spatial Resolution MRI
David Feinberg has supervised the construction of a head only 7T systems which produces ten times the spatial resolution (i.e., voxel size 0.2 × 0.2 × 1 mm) of conventional 3T clinical MRI systems [
25]. This NexGen system incorporates three layer gradients which have performance specifications of 200 mT/m gradient strength and 900 T/m/s slew rate. The cost to date has been USD 22 M. The system is particularly designed for high spatial resolution imaging of the cerebral cortex.
5.14. Ultra-High Field MRI
There are now an estimated 120 ultra-high field MRI systems in operation with the majority 7 T systems. This is approximately 0.2% of the 60,000 MRI systems installed Worldwide.
Notable installations at field strengths above 7 T are whole-body systems operating at 10.5 T [
26,
27] and 11.7 T [
28]. The latter system cost approximately USD 75M. Planning is underway for systems operating at 14T in Germany [
29] and the Netherlands [
30]. SNR ratio increases linearly with field strength or better (e.g., to the power of 1.65) and this results in a potential gain of approximately an order of magnitude in SNR when the static field strength is increased from 3 T to 11.7 T. This can be used to increase the speed of scanning, spatial resolution and/or contrast.
5.15. Summary
Another revolution in imaging soft tissue contrast is now in progress. The first major revolution was from plain radiographs to CT, the second major revolution was from CT to conventional MRI and the current one is from conventional MRI to ultra-high contrast MRI. The revolutions are characterized by change from invisible lesions with one modality to lesions seen with high contrast with the other modality.
With dSIR the spatial resolution is essentially the same as conventional images and the acquisition times are similar. This is in order to make the technique clinically realizable. The increase in contrast may be ten times greater than that with conventional IR sequences. This increase in contrast is typically targeted at normal appearing tissues, where there are only small changes in T1 and/or T2 present and these are insufficient to produce useful contrast with conventional state-of-the-art sequences. Thus, dSIR sequences are targeted to produce amplified contrast where it is most needed, i.e., to provide clinically useful visualization of the abnormalities which are not otherwise seen. Contrast is not increased in areas where there is already high contrast available on conventional images due to large changes in T1 and/or T2. Contrast amplification is not needed in this situation.
The changes in T1 seen with the whiteout sign are only small, but they are widespread and so high spatial resolution is not necessary to see them.
The whiteout sign may be part of a generalized neuroinflammatory response to insults to the brain of various types. It may be initially due to edema and acute inflammation and may regress over two days or less. It may also become persistent for years and its character may change to include demyelination and degeneration. Although the changes in T1 associated with the whiteout sign are small in size, they are widespread within white matter and this may result in them having significant clinical impact.
The dSIR technique is ideally used as complementary to T2-FLAIR. The T2-FLAIR sequence shows high contrast from larger changes in T2 against a generally bland background and has the advantage of clarity in demonstrating abnormality in comparison with sequences such as T2-wFSE and conventional IR sequences. The dSIR sequence typically shows abnormalities due to small changes in T1 and these may be present in the normal appearing white matter seen on positionally matched T2-FLAIR images. The sequences are thus complementary.
Ultra-high contrast dSIR sequences are easy to implement on existing MRI systems using sequences already on the systems, and the cost of this is a tiny fraction of that of building and installing ultra-high spatial resolution MRI or ultra-high field MRI systems. In addition, dSIR sequences can readily be implemented and successfully tested on volunteers in a day or two. Another major advantage of ultra-high contrast MRI is that it can, in principle, be implemented on MRI systems operating at any field strength. This is not the case with ultra-high spatial resolution MRI or ultra-high field MRI which are performed on the 0.2% of MRI systems that operate at 7T or greater field strengths.